Diagnostics

9 pages, 218 KB

Open AccessArticle

Retrospective Multicenter Analysis of Withdrawal Syndrome in Parkinson’s Disease Patients After Cessation of Deep Brain Stimulation

by Hatice Ömercikoğlu Özden, Fatma Nazlı Durmaz Çelik, Fatma Şeyda Üstüner, Galip Yardımcı, Orhan Abdullah Omar Tbh Bash, Serhat Özkan, Murat Vural, Fatih Bayraklı and Dilek Günal

Diagnostics 2026, 16(4), 644; https://doi.org/10.3390/diagnostics16040644 - 23 Feb 2026

Viewed by 442

Abstract

Background: Abrupt cessation of deep brain stimulation (DBS) in Parkinson’s disease (PD), most commonly due to implantable pulse generator (IPG) battery depletion, may lead to DBS withdrawal syndrome (DBS-WDS). However, withdrawal syndrome does not occur in all patients following stimulation cessation. Methods: We [...] Read more.

Background: Abrupt cessation of deep brain stimulation (DBS) in Parkinson’s disease (PD), most commonly due to implantable pulse generator (IPG) battery depletion, may lead to DBS withdrawal syndrome (DBS-WDS). However, withdrawal syndrome does not occur in all patients following stimulation cessation. Methods: We retrospectively analyzed 210 PD patients treated with DBS. Patients with documented stimulation cessation were evaluated for the presence of withdrawal syndrome based on established clinical criteria. Demographic, disease-related, and treatment characteristics were assessed, and descriptive analysis was conducted on severe cases requiring intensive care. Results: DBS battery shutdown occurred in 28 patients (13.3%). Most patients did not develop withdrawal syndrome and experienced only transient motor worsening. Severe DBS-WDS requiring intensive care was rare, occurring in only three patients (1.4%). Battery shutdown alone did not predict withdrawal, nor was preoperative levodopa equivalent daily dose associated with withdrawal risk. Conclusions: DBS battery shutdown is usually not accompanied by withdrawal syndrome, and severe DBS-WDS is uncommon. Proactive battery management may help to prevent this rare but serious complication. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

12 pages, 468 KB

Open AccessArticle

The Wrist Circumference-to-Body Mass Index Ratio for Preprocedural Risk Stratification of Radial Artery Spasm in Transradial Coronary Angiography and Percutaneous Coronary Intervention

by Ahmet Can Çakmak, Betül Sarıbıyık Çakmak and Muhammed Necati Murat Aksoy

Diagnostics 2026, 16(4), 643; https://doi.org/10.3390/diagnostics16040643 - 23 Feb 2026

Viewed by 375

Abstract

Objectives: Radial artery spasm (RAS) is a common complication of transradial coronary angiography that may adversely affect procedural success and patient comfort. This study aimed to evaluate clinical, procedural, and anthropometric factors associated with RAS in patients undergoing elective transradial coronary angiography, [...] Read more.

Objectives: Radial artery spasm (RAS) is a common complication of transradial coronary angiography that may adversely affect procedural success and patient comfort. This study aimed to evaluate clinical, procedural, and anthropometric factors associated with RAS in patients undergoing elective transradial coronary angiography, with a particular focus on the wrist circumference-to-body mass index (WC/BMI) ratio as a novel predictor. Methods: A total of 466 patients who underwent elective coronary angiography via the right radial artery between January 2024 and December 2024 were included. All procedures were performed using a 6 Fr introducer sheath according to a standardized protocol. Radial artery spasm was clinically defined as operator resistance during catheter manipulation accompanied by patient-reported pain or marked discomfort in the accessed arm. Wrist circumference and body mass index were measured before the procedure, and the WC/BMI ratio was calculated. Radial artery diameter was assessed using ultrasonography. Variables associated with RAS were evaluated using univariable and multivariable logistic regression analyses. Due to collinearity between WC/BMI and radial artery diameter, two separate multivariable models were constructed. Discriminative performance was assessed using receiver operating characteristic (ROC) curve analysis. Results: Radial artery spasm occurred in 51 patients (10.9%). Patients who developed RAS had significantly lower WC/BMI ratios and smaller radial artery diameters compared with those without spasm (both p ≤ 0.001). In multivariable analysis, a lower WC/BMI ratio was independently associated with an increased risk of RAS (odds ratio [OR] 0.51 per 0.1-unit increase; 95% confidence interval [CI] 0.34–0.78; p = 0.002). Similarly, smaller radial artery diameter remained an independent predictor of RAS (OR 0.83 per 0.1 mm increase; 95% CI 0.75–0.92; p < 0.001). The area under the curve (AUC) was 0.651 for WC/BMI and 0.636 for radial artery diameter. The combined model demonstrated improved discriminative ability (AUC 0.713). Conclusions: The WC/BMI ratio is a simple, practical, and readily obtainable anthropometric parameter that can predict the risk of radial artery spasm before transradial coronary angiography. When combined with radial artery diameter, it provides improved discrimination for identifying patients at higher risk of RAS. Full article

(This article belongs to the Section Medical Imaging and Theranostics)

► Show Figures

Figure 1

19 pages, 3734 KB

Open AccessProtocol

Beyond the Image Frame: An Art-Based Pedagogical Framework for Teaching Diagnostic Reasoning in Breast Ultrasound to Medical Students

by Marcin Śniadecki, Maria Morawska, Patrycja Kijańska, Olga Kondratowicz, Julia Nowakowska, Oliwia Musielak, Abhishek Singla, Ritu Amit Chhabria, Hanaf Alvi, Amelia Banaszak, Lena Grono, Diana Akhmed, Klaudia Kokot, Maksymilian Grzelak, Konrad Duszyński, Katsiaryna Marozik, Patrycja Jaworska, Jakub Majchrzak, Natallia Krupovich, Zuzanna Boyke, Julia Respondek, Weronika Ciećko, Ewa Bandurska, Jakub Szałek, Agata Rutkowska, Martyna Danielkiewicz, Patryk Poniewierza, Ewelina Klimik, Jarosław Meyer-Szary, Cynthia Aristei, Anna Malitowska and on behalf of Senological Gynecology Working Group Show full author list Hide full author list

Diagnostics 2026, 16(4), 642; https://doi.org/10.3390/diagnostics16040642 - 23 Feb 2026

Viewed by 631

Abstract

Breast ultrasound is a key diagnostic method for breast cancer and relies heavily on the interpretation of visual cues. At the same time, medical education is increasingly being driven by time constraints, which favors rapid pattern recognition, limiting the scope for reflective image [...] Read more.

Breast ultrasound is a key diagnostic method for breast cancer and relies heavily on the interpretation of visual cues. At the same time, medical education is increasingly being driven by time constraints, which favors rapid pattern recognition, limiting the scope for reflective image analysis and the diagnostic process. Therefore, the aim of this study was to propose and evaluate an artistic and pedagogical teaching model, inspired by the interpretive practices of Italian High Renaissance painting, as a tool to support the development of diagnostic reasoning in breast ultrasound. This model focuses on careful observation, analysis of the relationship between detail and the overall image, and the conscious transformation of visual cues into clinical meaning. This study was conducted during the four-day ARSA Think Tank Meeting (ARSATTM). Medical students worked in four groups; two groups received methodological training based on visual cue analysis, and two did not. All groups performed identical tasks involving the interpretation of breast ultrasound images and ultrasound examinations on real patients. The results indicate that an artistic–pedagogical teaching model to promote more coherent and reflective diagnostic reasoning in breast ultrasound is feasible. Therefore, integrating this approach may be a valuable addition to medical students’ ultrasound education in the realities of limited clinical time. Full article

(This article belongs to the Special Issue Frontline of Breast Imaging)

► Show Figures

Graphical abstract

20 pages, 6491 KB

Open AccessArticle

Comprehensive Sonographic Paradigm and Trend Pattern of Median Nerve Indices in Carpal Tunnel Syndrome from Wrist to Forearm: What We Need to Know

by Adeena Khan, Fawaz T. Salamah, Syed S. Habib, Waleed Fawzy, Fawzia AlRouq, Huthayfah T. Alkhliwi, Mamoona Sultan and Ahmed O. Alsabih

Diagnostics 2026, 16(4), 641; https://doi.org/10.3390/diagnostics16040641 - 23 Feb 2026

Viewed by 454

Abstract

Objective: The study aim was panoramic sonographic inspection of the median nerve (MN) from the wrist to the forearm in cases and controls. Additionally, integration of comparisons at various levels may aid in identifying principal ultrasound parameters of carpal tunnel syndrome (CTS). Methods: [...] Read more.

Objective: The study aim was panoramic sonographic inspection of the median nerve (MN) from the wrist to the forearm in cases and controls. Additionally, integration of comparisons at various levels may aid in identifying principal ultrasound parameters of carpal tunnel syndrome (CTS). Methods: Dynamic, static, and panoramic sonographies of 65 healthy and 83 CTS hands were performed. Multileveled qualitative (MN and flexor retinaculum morphology) and quantitative (cross-sectional area CSA, differences, and ratios) MN variables for CTS, followed by comparative statistical analysis to predict values and patterns of MN, were derived. Results: Subjectively, hypoechoic, vascular, compressed, hypomobile MN and bowed thickened flexor retinaculum were significantly more prevalent in cases (28.9–66.3%) than in controls (0–7.7%). Objectively, the proximal to inlet (pi) and the forearm at 12 cm (12) were the most representative sites. The area under curve (AUC) values for the MN dimensions, in decreasing order, were 0.9, 0.89, 0.86, and ≤0.80 for the CSA difference ‘pi’ and ‘12’ (Cpi-C12), the CSA proximal to inlet (Cpi), the ratio of CSA at pi and 12 (Cpi/C12), and the CSA at inlet (Ci), respectively. Their cut-off values were 3.7, 9.1, 1.8, and 7.2 mm², respectively. Differences and ratios between ‘Cpi’ and ‘Ci’ were less reliable (AUC ≤ 0.74, sensitivity ≤ 61.4%). Flexor retinaculum bowing, thickening, and MN flattening ratios were unreliable. Conclusions: Sensitivity, specificity, and precision of MN sonographic parameters in CTS increase by utilizing differences and ratios between wrist and forearm rather than isolated values. The recommended site in wrist is proximal to the inlet, and in the forearm, the best site to determine ratios and differences is at 12 cm from the distal wrist crease. Full article

(This article belongs to the Section Medical Imaging and Theranostics)

► Show Figures

Figure 1

11 pages, 697 KB

Open AccessArticle

Contrast-Enhanced Mammography vs. Breast MRI for Assessing Neoadjuvant Chemotherapy Response: A Prospective Clinical Comparison Study

by Omer Acar, Çağdaş Rıza Açar, İhsan Sebnem Orguc, Ferhat Ekinci, Mustafa Sahbazlar and Atike Pınar Erdoğan

Diagnostics 2026, 16(4), 640; https://doi.org/10.3390/diagnostics16040640 - 23 Feb 2026

Viewed by 551

Abstract

Objective: To compare contrast-enhanced mammography (CEM) with breast magnetic resonance imaging (MRI) in evaluating residual tumor size and pathological complete response after neoadjuvant chemotherapy (NAC) in breast cancer patients. Methods: This prospective study included patients with histopathologically confirmed breast cancer who [...] Read more.

Objective: To compare contrast-enhanced mammography (CEM) with breast magnetic resonance imaging (MRI) in evaluating residual tumor size and pathological complete response after neoadjuvant chemotherapy (NAC) in breast cancer patients. Methods: This prospective study included patients with histopathologically confirmed breast cancer who were scheduled to receive NAC followed by surgery. All patients underwent both CEM and breast MRI before initiation of NAC and within seven days after completion of treatment. Surgery was performed at a median of 14 days after post-treatment imaging. Residual tumor size measurements obtained by both imaging modalities were compared with histopathological findings, which served as the reference standard. Pathological complete response was defined as the absence of residual invasive carcinoma in the surgical specimen. Results: A total of 74 female patients were included. CEM estimated residual tumor size within ±1 cm of histopathology in 84.7% of cases, whereas MRI achieved this accuracy in 76.4%. Agreement with histopathology was higher for CEM than for MRI. In predicting pathological complete response, CEM demonstrated higher sensitivity (91.3%) and negative predictive value compared with MRI; however, this difference did not reach statistical significance (p = 0.24). MRI showed slightly higher specificity. Pathological complete response was observed in 31.1% of patients. Conclusion: Contrast-enhanced mammography demonstrated performance comparable to breast MRI in assessing response to neoadjuvant chemotherapy, with numerically higher sensitivity for predicting pathological complete response. CEM may represent a practical and accessible alternative to MRI, particularly in settings where MRI is unavailable or contraindicated. These findings support the clinical use of CEM as a reliable alternative imaging modality for response assessment. Full article

(This article belongs to the Special Issue Recent Advances in Breast Cancer Imaging 2026)

► Show Figures

Figure 1

13 pages, 534 KB

Open AccessArticle

Psychological Morbidity After Ocular Trauma: Association Between Initial Visual Loss and PTSD

by Gamze Ucan Gunduz, Oguzhan Kilincel, Sema Nizam Tekcan, Cengiz Akkaya and Ozgur Yalcinbayir

Diagnostics 2026, 16(4), 639; https://doi.org/10.3390/diagnostics16040639 - 23 Feb 2026

Viewed by 437

Abstract

Background: Ocular trauma is a significant cause of monocular visual impairment and potential psychological morbidity. This study aimed to determine the prevalence of anxiety, depression, and post-traumatic stress disorder (PTSD) in patients with mechanical ocular trauma and to investigate the predictive value of [...] Read more.

Background: Ocular trauma is a significant cause of monocular visual impairment and potential psychological morbidity. This study aimed to determine the prevalence of anxiety, depression, and post-traumatic stress disorder (PTSD) in patients with mechanical ocular trauma and to investigate the predictive value of baseline clinical characteristics, specifically initial visual acuity. Methods: This retrospective study included 58 adult patients treated for mechanical ocular trauma. Sociodemographic data, injury mechanisms, and clinical variables, including initial visual acuity (LogMAR), ocular trauma score, and number of ocular surgeries, were analyzed. Psychological status was assessed using the Beck Depression Inventory, Beck Anxiety Inventory, and a PTSD checklist. Multivariate logistic regression and correlation analyses were performed to identify predictors of severe PTSD. Results: The cohort was predominantly male (86.2%) with a mean age of 42.5 years. Severe or very severe PTSD symptoms were identified in 35.1% of patients. Analysis revealed a significant positive correlation between initial visual acuity and PTSD scores (r = 0.273, p = 0.038). In the logistic regression model, initial visual acuity (LogMAR) demonstrated the highest odds ratio for severe PTSD in the multivariable model; however, this association did not reach statistical significance (OR = 2.164, 95% CI: 0.720–6.508, p = 0.169) and should therefore be interpreted as an exploratory trend rather than a confirmed predictor. Conclusions: Greater visual loss at the time of injury showed the strongest, although non-significant, association with subsequent PTSD symptom severity. These findings suggest that patients with severe initial visual impairment following ocular trauma may benefit from early psychological screening and timely mental health referral, warranting confirmation in larger prospective studies. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

► Show Figures

Figure 1

16 pages, 1036 KB

Open AccessArticle

Breast Cancer Classification Using Feature Selection via Improved Simulated Annealing and SVM Classifier

by Maedeh Kiani Sarkaleh, Hossein Azgomi and Azadeh Kiani-Sarkaleh

Diagnostics 2026, 16(4), 637; https://doi.org/10.3390/diagnostics16040637 - 23 Feb 2026

Viewed by 407

Abstract

Background: Breast cancer is among the most common cancers in women, and early diagnosis is critical for better treatment outcomes and reduced mortality. Efficient computer-aided diagnostic (CAD) systems play a crucial role in enhancing diagnostic accuracy and facilitating timely clinical decisions. Methods: This [...] Read more.

Background: Breast cancer is among the most common cancers in women, and early diagnosis is critical for better treatment outcomes and reduced mortality. Efficient computer-aided diagnostic (CAD) systems play a crucial role in enhancing diagnostic accuracy and facilitating timely clinical decisions. Methods: This study proposes an automated CAD system for detecting cancerous tumors in mammograms, consisting of four stages: preprocessing, feature extraction, feature selection, and classification. In preprocessing, the region of interest (ROI) is extracted, followed by noise suppression and contrast enhancement to improve image quality. Shape, histogram, and tissue-related features are then computed from each ROI. An Improved Simulated Annealing (ISA) algorithm is employed to adaptively select the most informative features through a flexible process and composite fitness function, effectively reducing dimensionality while preserving high classification accuracy. Finally, classification is performed using a Support Vector Machine (SVM) to distinguish between malignant and benign masses. Results: Evaluation on the CBIS-DDSM and MIAS datasets showed the system achieved accuracies of 99.67% and 98%, sensitivities of 99.33% and 98%, and F1-scores of 99.66% and 97.9%, respectively. These results indicate notable improvements over traditional SA and full-feature approaches. Conclusions: The findings confirm the effectiveness of the ISA algorithm in selecting relevant features, thereby enhancing the performance of breast cancer detection. Full article

(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)

► Show Figures

Figure 1

24 pages, 2038 KB

Open AccessArticle

Evaluating the Managerial Feasibility of an AI-Based Tooth-Percussion Signal Screening Concept for Dental Caries: An In Silico Study

by Stefan Lucian Burlea, Călin Gheorghe Buzea, Irina Nica, Florin Nedeff, Diana Mirila, Valentin Nedeff, Lacramioara Ochiuz, Lucian Dobreci, Maricel Agop and Ioana Rudnic

Diagnostics 2026, 16(4), 638; https://doi.org/10.3390/diagnostics16040638 - 22 Feb 2026

Viewed by 496

Abstract

Background: Early detection of dental caries is essential for effective oral health management. Current diagnostic workflows rely heavily on radiographic imaging, which involves infrastructure requirements, workflow coordination, and resource considerations that may limit frequent use in high-throughput or resource-constrained settings. These contextual factors [...] Read more.

Background: Early detection of dental caries is essential for effective oral health management. Current diagnostic workflows rely heavily on radiographic imaging, which involves infrastructure requirements, workflow coordination, and resource considerations that may limit frequent use in high-throughput or resource-constrained settings. These contextual factors motivate exploration of adjunct screening concepts that could support front-end triage decisions within existing care pathways. This study evaluates, in simulation, whether modeled tooth-percussion response signals contain sufficient discriminative information to justify further translational and managerial investigation. Implementation costs, workflow optimization, and economic outcomes are not evaluated directly; rather, the objective is to assess whether the technical preconditions for a potentially scalable screening concept are satisfied under controlled in silico conditions. Methods: An in silico model of tooth percussion was developed in which enamel, dentin, and pulp/root structures were represented as a simplified layered mechanical system. Impulse responses generated from simulated tapping were used to compute the modeled surface-vibration response (enamel-layer displacement), which served as a proxy for a measurable percussion-related signal (e.g., contact vibration), rather than a recorded acoustic waveform. Carious conditions were simulated through depth-dependent reductions in stiffness and effective mass and increases in damping to represent enamel and dentin demineralization. A synthetic dataset of labeled simulated signals was generated under varying structural parameters and measurement-noise assumptions. Machine-learning models using Mel-frequency cepstral coefficient (MFCC) features were trained to classify healthy teeth, enamel caries, and dentin caries at a screening (triage) level. Results: Under baseline simulation conditions, the classifier achieved an overall accuracy of 0.97 with balanced macro-averaged F1-score (0.97). Misclassifications occurred primarily between healthy and enamel-caries categories, whereas dentin-caries cases were most consistently identified. When measurement noise and structural variability were increased, performance declined gradually, reaching approximately 0.90 accuracy under the most challenging simulated scenario. These results indicate that discriminative information is present within the modeled signals at a screening (triage) level, meaning that higher-risk categories can be distinguished probabilistically rather than with definitive diagnostic certainty. Sensitivity and specificity trade-offs were not optimized in this study, as the objective was to assess separability rather than to define clinical decision thresholds. Conclusions: Within the constraints of the in silico model, simulated tooth-percussion response signals demonstrated discriminative patterns between healthy, enamel caries, and dentin caries categories at a screening (triage) level. These findings establish technical plausibility under controlled simulation conditions and support further investigation of percussion-based screening as a potential adjunct to clinical assessment. From a healthcare management perspective, the present results address a prerequisite question—whether such signals contain sufficient information to justify translational research, rather than demonstrating workflow optimization, cost reduction, or system-level impact. Clinical validation, threshold optimization, and implementation studies are required before managerial or operational benefits can be evaluated. Full article

(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)

► Show Figures

Figure 1

13 pages, 679 KB

Open AccessArticle

Association of Intraoperative Parathyroid Hormone Decline with Early Postoperative Hypocalcemia: A Single-Center Retrospective Study

by Suat Evirgen, Elif Menekse, Ecem Avci, Burak Yasin Avci, Çiğdem Tura Bahadır and Cafer Polat

Diagnostics 2026, 16(4), 636; https://doi.org/10.3390/diagnostics16040636 - 22 Feb 2026

Viewed by 405

Abstract

Background/Objectives: Postoperative early hypocalcemia (PEH) is a key postoperative issue after parathyroidectomy in primary hyperparathyroidism. It often leads to long-lasting hypocalcemia, requiring more calcium and active vitamin D supplements. This study aimed to determine whether the extent of intraoperative parathyroid hormone (PTH) [...] Read more.

Background/Objectives: Postoperative early hypocalcemia (PEH) is a key postoperative issue after parathyroidectomy in primary hyperparathyroidism. It often leads to long-lasting hypocalcemia, requiring more calcium and active vitamin D supplements. This study aimed to determine whether the extent of intraoperative parathyroid hormone (PTH) decline, measured 15 min after parathyroid tumor excision, could serve as a reliable intraoperative rule-out marker for PEH. Methods: We conducted a retrospective review of 88 adult patients who underwent surgical intervention for a solitary parathyroid tumor at a single institution. Postoperative early hypocalcemia (PEH) was defined as a total serum calcium level <8.5 mg/dL within the postoperative 6th hour or on postoperative day 1, requiring clinical calcium supplementation (oral and/or intravenous), with active vitamin D when appropriate. The percentage decrease in PTH at 15 min post-excision was calculated using morning-of-surgery preoperative PTH values alongside the 15-min post-excision levels. Additional variables assessed included preoperative alkaline phosphatase (ALP), parathyroid tumor weight, and serum concentrations of calcium, phosphate, magnesium, and 25-hydroxyvitamin D. Predictive factors were identified by logistic regression, and the diagnostic accuracy of the 15-min PTH decline was evaluated using receiver operating characteristic (ROC) curve analysis, optimizing cutoff selection with Youden’s index. Odds ratios were standardized per 10-unit increments for ALP and parathyroid tumor weight for interpretability. Results: Of the studied cohort, 10 patients (11.4%) developed PEH. The intraoperative 15-min PTH decline was notably greater in those who developed PEH compared to those who did not (81.2 ± 4.4% vs. 69.9 ± 8.3%; p < 0.001). Univariate logistic regression showed a significant association between the 15-min PTH decline and PEH (OR 1.22 per 1% increment; 95% CI 1.08–1.38). That said, when we added ALP and parathyroid tumor weight to the multivariate models, PTH decline no longer predicted independently. In contrast, ALP (OR 3.11 per 10 U/L; 95% CI 1.34–7.93; p = 0.011) and parathyroid tumor weight (OR 1.22 per 10 mg; 95% CI 1.10–1.48; p = 0.004) stayed significant. Thus, the incremental prognostic contribution of the 15-min PTH decline beyond ALP and parathyroid tumor weight appears limited. The ROC curve for the 15-min PTH decline produced an AUC of 0.883, with an optimal cutoff of 75% providing 100% sensitivity and 74.4% specificity. No patients with a PTH decline below 75% developed PEH. Conclusions: Preoperative ALP and parathyroid tumor weight showed the strongest independent associations with PEH following parathyroid tumor surgery. An intraoperative PTH decline of less than 75% at 15 min may serve as a practical rule-out tool for PEH, although further validation in larger patient populations is warranted. Full article

(This article belongs to the Special Issue State of the Art in the Diagnosis and Management of Endocrine Tumors)

► Show Figures

Graphical abstract

38 pages, 536 KB

Open AccessReview

Toward Smart Salivary Diagnostics: A Comprehensive Review of Heavy Metal Biomarkers and Digital Risk Modeling

by Claudia Florina Bogdan-Andreescu, Lucia Bubulac, Cristina-Crenguţa Albu, Dan Alexandru Slăvescu, Andreea Mariana Bănăţeanu, Oana Botoacă, Gabriela-Cornelia Muşat, Viorica Tudor, Emin Cadar and Mariana Păcurar

Diagnostics 2026, 16(4), 635; https://doi.org/10.3390/diagnostics16040635 - 22 Feb 2026

Viewed by 650

Abstract

Background: Saliva has been identified as a valuable diagnostic biofluid due to its non-invasive collection and its capacity to reflect oral and systemic biological processes. Advances in analytical chemistry, biosensing technologies, and artificial intelligence (AI)-assisted data integration have broadened the applications of [...] Read more.

Background: Saliva has been identified as a valuable diagnostic biofluid due to its non-invasive collection and its capacity to reflect oral and systemic biological processes. Advances in analytical chemistry, biosensing technologies, and artificial intelligence (AI)-assisted data integration have broadened the applications of salivary diagnostics. Among salivary exposome components, heavy metals such as lead, cadmium, mercury, nickel, chromium, arsenic, and aluminum serve as biologically and clinically relevant indicators of environmental exposure, toxic burden, and disease-associated molecular disorders. Methods: This structured review integrates clinical, experimental, and translational studies published between January 2020 and January 2026 that examined salivary heavy metal profiling in relation to oral health. Evidence was identified using systematic searches of PubMed/MEDLINE and supplementary sources. Studies were qualitatively assessed regarding analytical methodologies, reported concentration ranges, biological mechanisms, disease associations, and the development of digital and AI-assisted diagnostic applications. Results: Thirteen human clinical studies and six animal or in vivo investigations met the inclusion criteria. Across these studies, altered salivary metal profiles were linked to oxidative stress, inflammatory signaling, immune dysregulation, microbiome disturbances, and genotoxic markers relevant to periodontal disease, oral mucosal pathology, and the risk of oral squamous cell carcinoma. Inductively coupled plasma mass spectrometry was the predominant analytical platform, while emerging biosensor technologies showed potential for rapid detection and monitoring. Digital and AI-based approaches were identified as promising tools for integrating metallomic data with clinical and molecular biomarkers to support exposure-informed risk stratification. Conclusions: Salivary heavy metal profiling represents a biologically informative, non-invasive method for exposure-aware risk assessment in oral health. Although current clinical translation is limited by methodological variability, small cohort sizes, and the lack of standardized reference ranges, integration with digital biosensing platforms and explainable AI frameworks might facilitate scalable, precision-oriented salivary diagnostics. Full article

(This article belongs to the Special Issue Digital Innovation in Salivary Diagnostics for Oral and Systemic Health)

13 pages, 1638 KB

Open AccessArticle

Evaluation of Root Angulations Through Panoramic Films Using Artificial Intelligence

by Deniz Şevik, Nurullah Akkaya, Ulas Oz and Beste Kamiloglu

Diagnostics 2026, 16(4), 634; https://doi.org/10.3390/diagnostics16040634 - 22 Feb 2026

Viewed by 393

Abstract

Background/Objectives: Accurate evaluation of root angulation is essential for assessing root parallelism and orthodontic treatment outcomes. In routine clinical practice, this assessment is often performed by visual inspection of panoramic radiographs, which is subjective and prone to observer variability. The objective of [...] Read more.

Background/Objectives: Accurate evaluation of root angulation is essential for assessing root parallelism and orthodontic treatment outcomes. In routine clinical practice, this assessment is often performed by visual inspection of panoramic radiographs, which is subjective and prone to observer variability. The objective of this study was to develop and validate an artificial intelligence (AI)–based algorithm for automated, quantitative assessment of mesiodistal root angulations on panoramic radiographs and to evaluate its accuracy relative to conventional manual measurements. Methods: A total of 214 panoramic radiographs (orthopantomograms), comprising 4280 posterior teeth, were retrospectively selected after applying strict inclusion and exclusion criteria. Individual teeth were automatically segmented using a U²-Net–based deep learning architecture. Tooth long-axis orientation was calculated using principal component analysis, with exclusion of the apical third to minimize the influence of root curvature. Angular deviation was measured relative to fixed horizontal reference lines. Manual measurements performed by experienced examiners using 3D Slicer software served as the reference standard. Intra- and inter-examiner reliability, agreement between AI-based and manual measurements, intraclass correlation coefficients (ICC), and Bland–Altman analyses were calculated. Results: Manual measurements demonstrated excellent reliability, with intra-examiner and inter-examiner ICC values of 0.972 and 0.963, respectively. Agreement between the AI-based algorithm and manual measurements was also excellent (ICC = 0.941). Bland–Altman analysis showed a mean difference of −0.10°, with 95% limits of agreement ranging from −1.60° to 1.41°, indicating minimal bias and no proportional error. Conclusions: The proposed AI-based algorithm provides accurate, objective, and reproducible measurements of posterior tooth root angulations on panoramic radiographs. This approach may support clinical decision-making, reduce observer-related variability, and facilitate efficient assessment of root parallelism in orthodontic practice. Full article

(This article belongs to the Special Issue Advances in Dental Imaging)

► Show Figures

Figure 1

20 pages, 1278 KB

Open AccessSystematic Review

Performance and Clinical Utility of Deep Learning for Detecting Referable Age-Related Macular Degeneration on Fundus Photographs: A Systematic Review and Meta-Analysis

by Wei-Ting Luo and Ting-Wei Wang

Diagnostics 2026, 16(4), 633; https://doi.org/10.3390/diagnostics16040633 - 22 Feb 2026

Viewed by 413

Abstract

Background/Objectives: Age-related macular degeneration (AMD) is a leading cause of irreversible central vision loss in older adults. Detection of referable AMD—typically intermediate or advanced disease requiring specialist evaluation—is critical for timely intervention. Deep learning (DL) applied to color fundus photographs has emerged as [...] Read more.

Background/Objectives: Age-related macular degeneration (AMD) is a leading cause of irreversible central vision loss in older adults. Detection of referable AMD—typically intermediate or advanced disease requiring specialist evaluation—is critical for timely intervention. Deep learning (DL) applied to color fundus photographs has emerged as a potential tool to support large-scale AMD screening. This systematic review and meta-analysis evaluated the diagnostic accuracy of DL algorithms for detecting referable AMD and compared their performance with human graders. Methods: We systematically searched PubMed, Embase, Web of Science, and IEEE Xplore through 18 December 2025. Diagnostic accuracy studies assessing DL algorithms on color fundus photographs for referable AMD in adults were included. Two reviewers independently screened studies, extracted data, and assessed risk of bias using an AI-adapted PROBAST framework. Pooled sensitivity and specificity were estimated using a bivariate random-effects model. Clinical utility was evaluated using likelihood ratios, and paired head-to-head comparisons were synthesized using a contrast-based meta-analysis. Results: Fourteen studies were included. DL algorithms achieved a pooled sensitivity of 0.91 (95% CI: 0.86–0.94) and specificity of 0.93 (95% CI: 0.86–0.96), with substantial heterogeneity. The pooled positive and negative likelihood ratios were 12.22 and 0.10, respectively, indicating strong diagnostic utility. In direct comparisons, DL systems showed slightly lower sensitivity but higher specificity than human graders. Conclusions: Deep learning demonstrates high diagnostic accuracy for detecting referable AMD from fundus photographs and may support screening and referral workflows. Further prospective validation and standardized evaluation are needed before widespread clinical implementation. Full article

(This article belongs to the Special Issue AI-Powered Ophthalmic Diagnostics: From Image Segmentation to Disease Classification)

► Show Figures

Figure 1

17 pages, 1139 KB

Open AccessArticle

Gestational Diabetes Mellitus in Singleton and Twin Pregnancies: A Comparison of Fetomaternal Outcomes

by Selina Balke, Izabela A. Kotzott, Annette Aigner, Petra Weid, Wolfgang Henrich, Joachim W. Dudenhausen and Josefine T. Königbauer

Diagnostics 2026, 16(4), 632; https://doi.org/10.3390/diagnostics16040632 - 22 Feb 2026

Viewed by 486

Abstract

Background: Gestational diabetes mellitus (GDM) complicates a significant number of pregnancies and is associated with both short- and long-term risks for the mother and child. Twin pregnancies are inherently high risk, and the coexistence of GDM may amplify these risks. While the effects [...] Read more.

Background: Gestational diabetes mellitus (GDM) complicates a significant number of pregnancies and is associated with both short- and long-term risks for the mother and child. Twin pregnancies are inherently high risk, and the coexistence of GDM may amplify these risks. While the effects of GDM in singleton pregnancies have been widely studied, data on its impact in twin gestations remain limited. The aim of this study was to determine differences regarding metabolic characteristics, treatment requirements, and maternal as well as fetal outcomes between twin and singleton pregnancies with GDM to contribute to improved perinatal care. Methods: This retrospective study included obstetric data from 73 twin pregnancies (146 neonates) and 1664 singleton pregnancies with a GDM diagnosis at a tertiary perinatal center in Berlin, Germany, between 2015 and 2022. Baseline characteristics and perinatal outcomes were assessed. Adjusted multiple linear and logistic regression analyses were used for group comparisons. Results: Women with GDM in twin and singleton pregnancies exhibited comparable glucose values in the 75 g oral glucose tolerance test (OGTT) (median fasting: 95 vs. 96 mg/dL; 1 h: 183 vs. 183 mg/dL; 2 h: 144 vs. 139 mg/dL). Despite this, insulin therapy was required significantly less often in twin (5.5%) compared to singleton pregnancies (22.3%) (OR = 0.86; 95% CI: 0.78–0.96). Among insulin-treated women, combined insulin therapy was most common in twins (75%), while singleton mothers most frequently received long-acting insulin alone (61.7%), followed by combined therapy (31.3%) and short-acting insulin alone (7%). Birthweight was significantly lower in twins (β = –0.83 kg; 95% CI: –0.98 to –0.69), and when evaluated using twin-based growth standards, twins were more likely to be classified as having intrauterine growth restriction (IUGR, <3rd percentile) (OR = 3.37; 95% CI: 0.96–9.11), being small for gestational age (SGA, <10th percentile) (OR = 2.50; 95% CI: 1.23–4.76), or having a birthweight below the 30th percentile (OR = 6.11; 95% CI: 3.49–11.12). No large-for-gestational-age (LGA, >90th percentile) neonates were observed in the twin group. Conclusions: GDM manifests differently in twin and singleton pregnancies. Despite similar OGTT values, twin mothers require insulin less frequently. Growth-related complications such as IUGR and SGA are significantly more frequent in twins, likely reflecting the physiological constraints of multiple gestations rather than GDM itself. Conversely, LGA is predominantly a concern in singleton pregnancies. These findings underscore the need for individualized diagnostic criteria and management strategies for GDM in twin pregnancies. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

► Show Figures

Figure 1

16 pages, 2796 KB

Open AccessArticle

MiMics-Net: A Multimodal Interaction Network for Blastocyst Component Segmentation

by Adnan Haider, Muhammad Arsalan and Kyungeun Cho

Diagnostics 2026, 16(4), 631; https://doi.org/10.3390/diagnostics16040631 - 21 Feb 2026

Viewed by 443

Abstract

Objectives: Global infertility rates are rapidly increasing. Assisted reproductive technologies combined with artificial intelligence are the next hope for overcoming infertility. In vitro fertilization (IVF) is gaining popularity owing to its increasing success rates. The success rate of IVF essentially depends on the [...] Read more.

Objectives: Global infertility rates are rapidly increasing. Assisted reproductive technologies combined with artificial intelligence are the next hope for overcoming infertility. In vitro fertilization (IVF) is gaining popularity owing to its increasing success rates. The success rate of IVF essentially depends on the assessment and inspection of blastocysts. Blastocysts can be segmented into several important compartments, and advanced and precise assessment of these compartments is strongly associated with successful pregnancies. However, currently, embryologists must manually analyze blastocysts, which is a time-consuming, subjective, and error-prone process. Several AI-based techniques, including segmentation, have been recently proposed to fill this gap. However, most existing methods rely only on raw grayscale intensity and do not perform well under challenging blastocyst image conditions, such as low contrast, similarity in textures, shape variability, and class imbalance. Methods: To overcome this limitation, we developed a novel and lightweight architecture, the microscopic multimodal interaction segmentation network (MiMics-Net), to accurately segment blastocyst components. MiMics-Net employs a multimodal blastocyst stem to decompose and process each frame into three modalities (photometric intensity, local textures, and directional orientation), followed by feature fusion to enhance segmentation performance. Moreover, MiMic dual-path grouped blocks have been designed, in which parallel-grouped convolutional paths are fused through point-wise convolutional layers to increase diverse learning. A lightweight refinement decoder is employed to refine and restore the spatial features while maintaining computational efficiency. Finally, semantic skip pathways are induced to transfer low- and mid-level spatial features after passing through the grouped and point-wise convolutional layers. Results/Conclusions: MiMics-Net was evaluated using a publicly available human blastocyst dataset and achieved a Jaccard index score of 87.9% while requiring only 0.65 million trainable parameters. Full article

(This article belongs to the Special Issue Artificial Intelligence in Diagnostics: From Algorithms to Clinical Impact)

► Show Figures

Figure 1

23 pages, 26789 KB

Open AccessArticle

DermaCalibra: A Robust and Explainable Multimodal Framework for Skin Lesion Diagnosis via Bayesian Uncertainty and Dynamic Modulation

by Ben Wang, Qingjun Niu, Chengying She, Jialu Zhang, Wei Gao and Lizhuang Liu

Diagnostics 2026, 16(4), 630; https://doi.org/10.3390/diagnostics16040630 - 21 Feb 2026

Viewed by 414

Abstract

Background: Accurate and timely diagnosis of skin lesions, including Melanoma (MEL), Basal Cell Carcinoma (BCC), Squamous Cell Carcinoma (SCC), Actinic Keratosis (ACK), Seborrheic Keratosis (SEK), and Nevus (NEV), is often hindered by the severe class imbalance and high morphological similarity among pathologies in [...] Read more.

Background: Accurate and timely diagnosis of skin lesions, including Melanoma (MEL), Basal Cell Carcinoma (BCC), Squamous Cell Carcinoma (SCC), Actinic Keratosis (ACK), Seborrheic Keratosis (SEK), and Nevus (NEV), is often hindered by the severe class imbalance and high morphological similarity among pathologies in clinical practice. Although multimodal learning has shown potential in resolving these issues, existing approaches often fail to address predictive uncertainty or effectively integrate heterogeneous clinical metadata. Therefore, this study proposes DermaCalibra, a robust and explainable multimodal framework optimized for small-scale, imbalanced clinical datasets. Methods: The proposed framework integrates three essential modules: First, the Attention-Based Multimodal Channel Recalibration (AMCR) module introduces a probabilistic Bayesian uncertainty estimation mechanism via Monte Carlo dropout to adjust focal loss weights, prioritizing features from underrepresented classes. Second, the Metadata-Driven Dynamic Feature Modulation and Cross-Attention Fusion (MDFM-CAF) module, designed to resolve inter-class visual ambiguity, dynamically rescales dermoscopic feature maps using non-linear clinical context transformations. Lastly, the Gradient Feature Attribution (GFA) module is implemented to provide pixel-level diagnostic heatmaps and metadata importance scores. Results: Evaluated on the PAD-UFES-20 dataset, DermaCalibra achieves a balanced accuracy (BACC) of 84.2%, outperforming current state-of-the-art (SOTA) methods by 3.6%, and a Macro Area Under the Receiver Operating Characteristic Curve (Macro AUC) of 96.9%. Extensive external validation on unseen hospital and synthetic datasets confirms robust generalizability across diverse clinical settings without the need for retraining. Conclusions: DermaCalibra effectively bridges the gap between deep learning complexity and clinical intuition through uncertainty-aware reasoning and transparent interpretability. The framework provides a reliable and scalable computer-aided diagnostic tool for early skin lesion detection, particularly in resource-limited clinical environments. Full article

(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)

► Show Figures

Figure 1

24 pages, 545 KB

Open AccessReview

The Effect of Human Papillomavirus Infection on Pregnancy Outcomes: A Scoping Review

by Borek Sehnal, Jan Zapletal, Martin Hruda, Vit Drochytek, Katerina Maxova, Michael J. Halaska, Lukas Rob and Ruth Tachezy

Diagnostics 2026, 16(4), 629; https://doi.org/10.3390/diagnostics16040629 - 21 Feb 2026

Viewed by 600

Abstract

Background: Human papillomavirus (HPV) is the most common sexually transmitted viral infection worldwide. Moreover, the prevalence of HPV infection is twice as high in pregnant women as in non-pregnant individuals. The aim of this review was to examine adverse pregnancy outcomes associated with [...] Read more.

Background: Human papillomavirus (HPV) is the most common sexually transmitted viral infection worldwide. Moreover, the prevalence of HPV infection is twice as high in pregnant women as in non-pregnant individuals. The aim of this review was to examine adverse pregnancy outcomes associated with cervicovaginal or placental HPV infection confirmed by a sensitive molecular method. Methods: We conducted searches on major medical databases including PubMed, EMBASE, Global Health, and the Cochrane Library to identify all studies examining HPV infection during pregnancy. Additionally, other online sources were consulted for relevant studies. Thirty-four records out of the initial 1868 identified were included in this review for thematic synthesis. The PRISMA-ScR guidelines were followed. Results: This scoping review included a total of 28 original observational studies, 1 systematic review, and 5 meta-analyses. Active HPV infection appears to significantly increase the risk of preterm premature rupture of membranes and preterm birth, as indicated by findings from published meta-analyses and systematic reviews. Determining the association of HPV infection with certain adverse pregnancy outcomes is challenging due to their frequency (such as miscarriage) or rarity (such as intrauterine fetal death). For conditions like preeclampsia and intrauterine fetal growth restriction, the limited number of heterogeneous studies precludes definitive conclusions. Moreover, the causes of these outcomes are typically multifactorial. The presence of HPV in trophoblasts and placental tissue is considered crucial for potential adverse pregnancy outcomes. There appears to be a strong correlation between cervicovaginal or urinary HPV infections and placental HPV infections in pregnant women. Conclusions: Persistent HPV infection seems to elevate the risk of preterm premature rupture of membranes and preterm birth. However, the currently available observational evidence does not allow for definitive conclusions regarding causality, and the reported findings should be interpreted as associations rather than proof of a causal relationship. The changes in frequency of certain perinatal complications in populations of women with high HPV vaccination rates may shed more light on this connection. Full article

(This article belongs to the Special Issue Infectious Diseases in Obstetrics and Gynecology: Diagnosis and Management)

► Show Figures

Figure 1

16 pages, 820 KB

Open AccessArticle

Left Atrial Size Modification After Catheter Ablation Predicts Late Atrial Fibrillation Recurrence

by Yung-Lung Chen, Bang-Yan Hou, Hsiang-Hsuan Chen, Pei-Ting Lin and Hui-Ting Wang

Diagnostics 2026, 16(4), 628; https://doi.org/10.3390/diagnostics16040628 - 21 Feb 2026

Viewed by 451

Abstract

Background: Radiofrequency catheter ablation for atrial fibrillation (AF) restores sinus rhythm, but late recurrence is common. Left atrial (LA) size is a known predictor of AF recurrence, but the prognostic value of early post-ablation LA remodeling remains underexplored. Objective: We aimed [...] Read more.

Background: Radiofrequency catheter ablation for atrial fibrillation (AF) restores sinus rhythm, but late recurrence is common. Left atrial (LA) size is a known predictor of AF recurrence, but the prognostic value of early post-ablation LA remodeling remains underexplored. Objective: We aimed to evaluate whether pre-ablation and early post-ablation LA volume index (LAVI) predict late atrial tachyarrhythmia recurrence after AF ablation. Methods: This is a retrospective single-center study of adults undergoing their first radiofrequency ablation for AF between January 2013 and December 2021. LA volume was measured by transthoracic echocardiography and indexed to body surface area to derive LAVI within one week before ablation and at 6 and 12 months after the procedure. The 6-month echocardiographic assessment was prespecified as the primary early post-ablation time point because it occurs beyond the 3-month blanking period and captures early structural remodeling during routine follow-up. Early recurrence was defined as atrial tachyarrhythmia occurring within 3 months after ablation, and late recurrence (LR) as any atrial tachyarrhythmic event thereafter. Multivariable Cox proportional hazards models were used to identify independent predictors of LR. Results: Among 408 patients with at least one year of follow-up, 157 (38.5%) experienced LR. Age and sex were similar between recurrence and non-recurrence groups (60.7 ± 9.8 vs. 59.9 ± 0.8 years; 56.1% vs. 64.1% male). Recurrence was associated with a higher prevalence of atrial flutter and persistent AF, higher pre-ablation and post-ablation LAVI, and lower post-ablation left ventricular ejection fraction. In multivariable analysis, atrial flutter, persistent AF, and LAVI, measured both before and after ablation, were independent predictors of LR. In receiver operating characteristic analysis, pre-ablation LAVI demonstrated modest discrimination for LR (AUC = 0.622; 95% CI 0.563–0.681; p < 0.001), with an optimal cut-off of 41.6 mL/m², while post-ablation LAVI showed similar performance (AUC = 0.597; 95% CI 0.532–0.662; p = 0.003), with a cut-off of 38.6 mL/m². Overall, discrimination was modest (AUC < 0.65), limiting LAVI as a standalone predictor. Conclusions: Elevated LAVI measured before and early after AF ablation independently predicts LR. Limited post-ablation LA reverse remodeling, reflected by persistently increased LAVI, is associated with unfavorable long-term rhythm outcomes. Serial assessment of LAVI may enhance post-ablation risk stratification. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

► Show Figures

Figure 1

15 pages, 721 KB

Open AccessArticle

Increased Risk of Incident Uveitis Among Patients with Psoriasis: A Nationwide Population-Based Cohort Study

by Scott Ehrenberg, Yoav Elizur, Niv Ben-Shabat, Paula David, Kassem Sharif, Yossef S. Bernstein, Ibrahim Abu Hilwe, Arnon D. Cohen, Abdulla Watad, Howard Amital and Yonatan Shneor Patt

Diagnostics 2026, 16(4), 627; https://doi.org/10.3390/diagnostics16040627 - 21 Feb 2026

Viewed by 438

Abstract

Background: Psoriasis is a chronic systemic inflammatory disease with established extra-cutaneous manifestations. While the association between uveitis and spondyloarthritis (SpA)-related disorders is well recognized, the incident risk of uveitis among broader psoriasis populations remains inadequately defined due to methodological limitations and inconsistent findings [...] Read more.

Background: Psoriasis is a chronic systemic inflammatory disease with established extra-cutaneous manifestations. While the association between uveitis and spondyloarthritis (SpA)-related disorders is well recognized, the incident risk of uveitis among broader psoriasis populations remains inadequately defined due to methodological limitations and inconsistent findings across previous studies. We aimed to estimate the incidence of uveitis in a large, nationwide population-based cohort and identify specific clinical and treatment-related predictors of ocular inflammation. Methods: This retrospective cohort study utilised electronic health records from Clalit Health Services, Israel’s largest health maintenance organization (2002–2024). We identified 157,360 patients with dermatologist-confirmed psoriasis and 156,927 age- and sex-matched controls. The primary outcome was incident uveitis, with risk estimated using Cox proportional hazards models. Within the psoriasis cohort, multivariable logistic regression was employed to identify predictors of uveitis, ensuring appropriate temporal sequencing between psoriasis treatment exposure and outcome. Results: Over a median follow-up of 12.6 years, psoriasis was associated with a significantly higher risk of incident uveitis (adjusted Hazard Ratio [aHR] 1.80; 95% CI, 1.50–2.15). Stratified analysis revealed a graded risk pattern: mild psoriasis showed no increased risk (aHR 1.01; 95% CI, 0.91–1.13), whereas severe disease (aHR 1.59; 95% CI, 1.25–2.03) and concomitant SpA (aHR 2.21; 95% CI, 1.87–2.61) demonstrated markedly elevated risks. Within the psoriasis cohort, independent predictors included SpA, diabetes mellitus, systemic lupus erythematosus, and sarcoidosis. Exposure to biologics, particularly etanercept (OR 3.37; 95% CI, 2.42–4.54), was associated with higher odds of uveitis, potentially reflecting higher disease severity. Conclusions: Incident uveitis risk in psoriasis is primarily driven by the magnitude of systemic inflammatory burden, with the highest risk observed in severe disease and those with concomitant SpA. Clinicians should maintain heightened vigilance for ocular symptoms in these high-risk subgroups to ensure timely intervention. Full article

(This article belongs to the Special Issue Trends and Diagnosis of Autoimmune Diseases)

► Show Figures

Figure 1

16 pages, 1991 KB

Open AccessArticle

Machine Learning-Driven Probability Scoring Enhances Diagnostic Certainty and Reduces Costs in Suspected Periprosthetic Joint Infection

by Jim Parr, Van Thai-Paquette, Amy Worden, James Baker, Paul Edwards and Krista O’Shaughnessey Toler

Diagnostics 2026, 16(4), 626; https://doi.org/10.3390/diagnostics16040626 - 20 Feb 2026

Viewed by 518

Abstract

Background: Accurate diagnosis of periprosthetic joint infection (PJI) remains challenging, particularly in culture-negative and borderline cases where current practices lead to high diagnostic uncertainty. SynTuition™, a machine-learning-based probability score integrating preoperative biomarkers, was developed to support clinical decision-making. This study compared its [...] Read more.

Background: Accurate diagnosis of periprosthetic joint infection (PJI) remains challenging, particularly in culture-negative and borderline cases where current practices lead to high diagnostic uncertainty. SynTuition™, a machine-learning-based probability score integrating preoperative biomarkers, was developed to support clinical decision-making. This study compared its diagnostic performance and economic impact with standard physician practice. Methods: A total of 12 physicians provided diagnoses of 274 clinical vignettes representing suspected PJI cases. SynTuition probabilities were converted to binary diagnostic classifications using a validated threshold. Diagnostic accuracy, agreement, indecision rates, decision curve analysis, and misdiagnosis-related costs were evaluated. Results: SynTuition achieved an overall percent agreement of 96.0% when compared against the expert adjudicated clinical reference, outperforming the pooled physician group at 90.8%. Physicians showed high indecision (38–48%) in inconclusive 2018 ICM cases, whereas SynTuition generated a definitive diagnosis with an 86.7% agreement against expert adjudication. Decision curve analysis demonstrated a higher net benefit for SynTuition across a broad range of thresholds, reducing projected unnecessary revision by up to 5.8%. Economic modeling showed a reduction in misdiagnosis-related costs from $6.9 million to $2.9 million per 1000 suspected PJI cases, yielding estimated savings of $4000 per suspected case. Conclusions: SynTuition demonstrated high diagnostic accuracy, lower uncertainty, and significant clinical and economic advantages over routine physician practice, supporting its integration into clinical decision-making for suspected PJI, particularly in diagnostically ambiguous cases. Full article

(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)

► Show Figures

Figure 1

14 pages, 562 KB

Open AccessArticle

Assessment of Salivary Parameters—pH, Buffering Capacity and Flow-Associated with Caries Susceptibility

by Alexandru Ștefârță, Mihaela Roxana Brătoiu, Maria Alexandra Rădoi, Veronica Mercuț, Mihaela Ionescu, Monica Scrieciu, Ileana-Cristiana Petcu, Petre-Costin Mărășescu, Marina Olimpia Amărăscu, Adrian Marcel Popescu and Diana-Elena Vlăduțu

Diagnostics 2026, 16(4), 625; https://doi.org/10.3390/diagnostics16040625 - 20 Feb 2026

Viewed by 618

Abstract

Background/Objectives: Saliva plays an essential role in maintaining the oral ecological balance, and its quantitative and qualitative characteristics may influence susceptibility to dental caries. The aim of this study was to determine susceptibility to dental caries based on the DMFT index and to [...] Read more.

Background/Objectives: Saliva plays an essential role in maintaining the oral ecological balance, and its quantitative and qualitative characteristics may influence susceptibility to dental caries. The aim of this study was to determine susceptibility to dental caries based on the DMFT index and to establish a correlation between caries experience and salivary parameters in a group of young adults. Methods: This cross-sectional study was conducted between July and November 2025 on a sample of 87 fourth-year students from the Faculty of Dentistry in Craiova. Each participant underwent an intraoral clinical examination to determine the DMFT index. The salivary parameters assessed included unstimulated salivary flow rate, saliva consistency, salivary pH, stimulated salivary flow rate, and buffering capacity, using the GC Saliva-Check Buffer kit. Statistical analyses were performed using SPSS (Statistical Package for Social Sciences) software, version 26 (SPSS Inc., Armonk, NY, USA). Results: The mean DMFT index value for the entire sample was 8.26 ± 4.481, with higher values observed among female participants. Low salivary pH was significantly associated with higher DMFT values. Participants with low or very low buffering capacity exhibited higher DMFT values compared to those with normal capacity, indicating that a reduced ability to neutralize salivary acidity is associated with increased caries activity. Conclusions: The results indicate that salivary pH and buffering capacity are important factors in dental caries susceptibility among young adults. The integration of salivary testing into the diagnostic assessment of caries risk may contribute to personalized and effective preventive strategies. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

► Show Figures

Figure 1

11 pages, 600 KB

Open AccessArticle

External Evaluation of a Predictive Model of Suboptimal Cytoreduction in Advanced Ovarian Cancer

by Anna Serra Rubert, Maria Victoria Ibañez Gual, Maria Teresa Climent Martí, Vicente Bebia, Antonio Gil-Moreno, Berta Díaz-Feijóo, Nadia Veiga Canuto, Juan Carlos Muruzábal, Gregorio Lopez-Gonzalez, Álvaro Tejerizo and Antoni Llueca

Diagnostics 2026, 16(4), 624; https://doi.org/10.3390/diagnostics16040624 - 20 Feb 2026

Viewed by 373

Abstract

Objective: The aim of this thesis was to externally validate a predictive model of suboptimal surgery in advanced ovarian cancer, developed by doctors Escrig and Llueca. The model classifies patients pre-surgically to estimate the likelihood of incomplete cytoreductive surgery. Methods: A retrospective cohort [...] Read more.

Objective: The aim of this thesis was to externally validate a predictive model of suboptimal surgery in advanced ovarian cancer, developed by doctors Escrig and Llueca. The model classifies patients pre-surgically to estimate the likelihood of incomplete cytoreductive surgery. Methods: A retrospective cohort comparison between two time periods was performed. Validation used a new cohort of 83 patients with advanced ovarian cancer, prospectively collected between 2017 and 2023 across five hospitals (experimental group). This group was compared with the original control cohort (2013–2016), which had served for model development. The predictive models (R3 and R4) are based on the Peritoneal Carcinomatosis Index (PCI) assessed by CT, laparoscopic PCI, and the presence of intestinal sub-obstruction. For model R4, intraoperative PCI was also included. Results: The experimental group had a lower rate of suboptimal cytoreduction compared with the control group (4.8% vs. 13.8%; p = 0.049). Significant differences were observed in ascites (49.4% vs. 27.5%; p = 0.002), and no patient in the experimental group presented intestinal sub-obstruction (0% vs. 8%; p = 0.002). Although at least 13 suboptimal surgeries were expected for validation, only four occurred. The predictive models did not classify any of these four cases as high risk, instead categorizing them as low or intermediate risk. Conclusions: Statistical external validation could not be performed due to event scarcity. This reduced incidence is attributed to selection bias: highly experienced surgical teams from participating centres likely applied criteria similar to those of the model, referring high risk patients (e.g., with intestinal sub-obstruction) to neoadjuvant therapy and thus avoiding suboptimal primary surgeries. Although direct validation was not possible, the findings indirectly suggest that the model is effective in guiding patient selection and improving surgical outcomes. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

► Show Figures

Figure 1

15 pages, 20078 KB

Open AccessArticle

IDH Mutation Assessment in Gliomas from Anatomical MRI Using Deep Learning: A Comparative Analysis of Centralized and Federated Learning Frameworks

by Abdullah Bas and Esin Ozturk-Isik

Diagnostics 2026, 16(4), 623; https://doi.org/10.3390/diagnostics16040623 - 20 Feb 2026

Viewed by 516

Abstract

Background/Objectives: Isocitrate dehydrogenase (IDH) mutation is a key prognostic indicator in diffuse gliomas; however, it is clinically determined from invasive tissue sampling. Non-invasive preoperative identification of IDH mutation from routine anatomical MRI could support treatment decision making. This study evaluated deep learning models [...] Read more.

Background/Objectives: Isocitrate dehydrogenase (IDH) mutation is a key prognostic indicator in diffuse gliomas; however, it is clinically determined from invasive tissue sampling. Non-invasive preoperative identification of IDH mutation from routine anatomical MRI could support treatment decision making. This study evaluated deep learning models for IDH mutation detection using routine anatomical MRI (post-contrast T1-weighted (T1c), T2-weighted, and fluid attenuated inversion recovery (FLAIR) MRI) and quantified how tumor-focused image preprocessing and different training schemes, centralized learning (CL) versus federated learning (FL) with alternative aggregation strategies, affected model performance. Methods: Anatomical MRI from 501 diffuse glioma patients in the UCSF Preoperative Diffuse Glioma MRI (UCSF-PDGM) dataset was analyzed using a deep learning classifier built on a 2D U-Net encoder, with age and sex included as covariates. Two methods of tumor-focused image preprocessing, Naïve Soft Filtering (NSF) and Gradient-Based Soft Filtering (GBSF), were compared. Centralized learning (CL) was benchmarked against federated learning (FL) using Federated Averaging (FA) and Federated Trimmed Mean (FTM) aggregation strategies. Model performance was compared in terms of accuracy, precision, recall, F1 score, specificity, and the area under the receiver operating characteristic curve (ROC-AUC). Results: The CL model with NSF achieved the best test performance (accuracy = 0.949, F1 = 0.951, ROC-AUC = 0.971), with NSF consistently outperforming GBSF. FL’s performance decreased relative to CL’s, but the FA strategy outperformed FTM (FTM accuracy = 0.915 vs. FA accuracy = 0.949), which indicates that the FL aggregation strategy has an influence on model performance. Conclusions: Deep learning applied to routine anatomical MRI could classify IDH mutation status with high accuracy. Context-preserving image preprocessing with NSF substantially improved performance across training schemes. FL provides a privacy-preserving alternative to CL, but incurs a measurable performance degradation that is sensitive to the choice of aggregation strategy. Full article

(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)

► Show Figures

Figure 1

33 pages, 12030 KB

Open AccessArticle

An Interpretable Ensemble Transformer Framework for Breast Cancer Detection in Ultrasound Images

by Riyadh M. Al-Tam, Aymen M. Al-Hejri, Fatma A. Hashim, Sachin M. Narangale, Mugahed A. Al-Antari and Sarah A. Alzakari

Diagnostics 2026, 16(4), 622; https://doi.org/10.3390/diagnostics16040622 - 20 Feb 2026

Viewed by 689

Abstract

Background/Objectives: Early and accurate detection of breast cancer is essential for reducing mortality and improving patient outcomes. However, the manual interpretation of breast ultrasound images is challenging due to image variability, noise, and inter-observer subjectivity. This study aims to address these limitations [...] Read more.

Background/Objectives: Early and accurate detection of breast cancer is essential for reducing mortality and improving patient outcomes. However, the manual interpretation of breast ultrasound images is challenging due to image variability, noise, and inter-observer subjectivity. This study aims to address these limitations by developing an automated and interpretable computer-aided diagnosis (CAD) system. Methods: We propose an automated and interpretable computer-aided diagnosis (CAD) system that integrates ensemble transfer learning with Vision Transformer architectures. The system combines the Data-Efficient Image Transformer (Deit) and Vision Transformer (ViT) through concatenation-based feature fusion to exploit their complementary representations. Preprocessing, normalization, and targeted data augmentation enhance robustness, while Gradient-weighted Class Activation Mapping (Grad-CAM) provides visual explanations to support clinical interpretability. The proposed model is benchmarked against state-of-the-art CNNs (VGG16, ResNet50, DenseNet201) and Transformer models (ViT, DeiT, Swin, Beit) using the Breast Ultrasound Images (BUSI) dataset. Results: The ensemble achieved 96.92% accuracy and 97.10% AUC for binary classification, and 94.27% accuracy with 94.81% AUC for three-class classification. External validation on independent datasets demonstrated strong generalizability, with 87.76%/88.07% accuracy/AUC on BrEaST, 86.77%/85.90% on BUS-BRA, and 86.99%/86.99% on BUSI_WHU. Performance decreased for fine-grained BI-RADS classification—76.68%/84.59% accuracy/AUC on BUS-BRA and 68.75%/81.10% on BrEaST—reflecting the inherent complexity and subjectivity of clinical subclassification. Conclusions: The proposed Vision Transformer-based ensemble demonstrates high diagnostic accuracy, strong cross-dataset generalization, and clinically meaningful explainability. These findings highlight its potential as a reliable second-opinion CAD tool for breast cancer diagnosis, particularly in resource-limited clinical environments. Full article

(This article belongs to the Special Issue Artificial Intelligence in Biomedical Imaging and Signal Processing)

► Show Figures

Figure 1

17 pages, 565 KB

Open AccessArticle

Bridging Perception and Reasoning: An Evidence-Based Agentic System for Diagnosis and Treatment Recommendations of Vascular Anomalies

by Yize Zhang, Yajing Qiu and Xiaoxi Lin

Diagnostics 2026, 16(4), 621; https://doi.org/10.3390/diagnostics16040621 - 20 Feb 2026

Viewed by 518

Abstract

Background: Vascular anomalies (VAs), including hemangiomas and vascular malformations, present a significant diagnostic challenge due to their high prevalence, complex classification (nearly 100 subtypes), and visual mimicry. Current Multimodal Large Language Models (MLLMs) struggle in this specialized domain, often failing to capture fine-grained [...] Read more.

Background: Vascular anomalies (VAs), including hemangiomas and vascular malformations, present a significant diagnostic challenge due to their high prevalence, complex classification (nearly 100 subtypes), and visual mimicry. Current Multimodal Large Language Models (MLLMs) struggle in this specialized domain, often failing to capture fine-grained visual features or lacking evidence-based reasoning. To address these limitations, we introduce HevaDx, an agentic diagnostic system that explicitly decouples visual perception from clinical reasoning. Methods: Leveraging a newly constructed large-scale dataset of VA patients, HevaDx employs a lightweight visual specialist for precise feature extraction and a reasoning specialist equipped with Retrieval-Augmented Generation (RAG) for therapeutic planning. This cooperative architecture mitigates the “reasoning gap” observed in end-to-end models by grounding decisions in up-to-date clinical guidelines. Results: Experimental results demonstrate that HevaDx achieves high performance with a top-3 diagnostic accuracy of 94.8% and a treatment recommendation accuracy of 83.3%. Conclusions: By bridging visual precision with transparent, verifiable logic, HevaDx offers a reliable framework for AI-assisted management of vascular anomalies. Full article

(This article belongs to the Special Issue Deep Learning in Biomedical Image and Signal Processing: Recent Advancements and Applications)

► Show Figures

Figure 1

18 pages, 1914 KB

Open AccessSystematic Review

From Image-Guided Surgery to Computer-Assisted Real-Time Diagnosis with Hyperspectral and Multispectral Imaging: A Systematic Review in Gynecologic Oncology

by Chiara Innocenzi, Matteo Pavone, Barbara Seeliger, Manuel Barberio, Nicolò Bizzarri, Toby Collins, Alexandre Hostettler, Lise Lecointre, Francesco Fanfani, Anna Fagotti, Antonello Forgione, Mariano Eduardo Giménez, Denis Querleu and Jacques Marescaux

Diagnostics 2026, 16(4), 620; https://doi.org/10.3390/diagnostics16040620 - 20 Feb 2026

Viewed by 667

Abstract

Background: There is a need for intraoperative image guidance in gynecologic oncologic surgery to provide accurate identification of malignant tissue and ensure negative resection margins. Emerging imaging technologies can complement standard histopathology and reshape intraoperative decision-making. Spectral imaging can extract information on tissue [...] Read more.

Background: There is a need for intraoperative image guidance in gynecologic oncologic surgery to provide accurate identification of malignant tissue and ensure negative resection margins. Emerging imaging technologies can complement standard histopathology and reshape intraoperative decision-making. Spectral imaging can extract information on tissue composition and physiological status in real time, without the need for tissue contact, contrast agents, staining, or freezing. This systematic review synthesizes its current clinical applications in gynecologic oncology, decision support utility, and diagnostic performance with data processing frameworks for tissue classification. Materials and Methods: This systematic review (PROSPERO: CRD420251032899) adhered to PRISMA guidelines. PubMed, Google Scholar, Embase, ClinicalTrials.gov, and Scopus databases were searched until September 2025. Manuscripts reporting data on spectral imaging in gynecologic oncology were included in the analysis. Results: Twenty-nine studies and two clinical trials met the inclusion criteria. Most of them focused on cervical neoplasia (n = 17, 58.6%) and ovarian cancer (n = 7, 24.1%) detection, followed by assessment of the fallopian tubes (n = 2, 6.9%), endometrium (n = 1, 3.4%), and vulvar skin (n = 2, 6.9%). Using final pathology as the gold standard, overall specificity ranged from 30 to 99%, and overall sensitivity from 75 to 100%, with particularly high sensitivity for cervical lesions (79–100%) and ovarian cancer (81–100%). Among the included studies, thirteen (44.8%) used data interpretation algorithms, of which eleven (84.6%) applied machine learning, one (7.7%) deep learning, and one (7.7%) combined both. Conclusions: Spectral imaging, supported by computational methods, has shown promising results in the diagnostic evaluation of gynecologic disease by providing functional and molecular information beyond the capacities of standard visual assessment. Full article

(This article belongs to the Special Issue Pathology and Diagnosis of Gynecologic Diseases, 3rd Edition)

► Show Figures

Figure 1

16 pages, 4401 KB

Open AccessArticle

Pulmonary Artery and Vein Morphology as an Imaging Biomarker for the Diagnosis of Pulmonary Hypertension

by Nedim Christoph Beste, Alexander Christian Bunck, Jonathan Kottlors, Robert Peter Wawer Matos Reimer, Jan Robert Kröger, Thomas Schömig, Lenhard Pennig, Kenan Kaya, Carsten Gietzen, Nils Große-Hokamp, Martin Urschler, Horst Olschewski, Stephan Rosenkranz, Florian J. Fintelmann, Michael Pienn and Roman Johannes Gertz

Diagnostics 2026, 16(4), 619; https://doi.org/10.3390/diagnostics16040619 - 20 Feb 2026

Cited by 1 | Viewed by 522

Abstract

Background/Objectives: To evaluate whether peripheral pulmonary artery and vein morphology improves image-based diagnosis of pulmonary hypertension (PH), in accordance with the recently updated hemodynamic definition. Methods: 229 patients underwent CT pulmonary angiography (CTPA) within 30 days of RHC. Pulmonary vessels ranging [...] Read more.

Background/Objectives: To evaluate whether peripheral pulmonary artery and vein morphology improves image-based diagnosis of pulmonary hypertension (PH), in accordance with the recently updated hemodynamic definition. Methods: 229 patients underwent CT pulmonary angiography (CTPA) within 30 days of RHC. Pulmonary vessels ranging between 2 and 10 mm in diameter were extracted and labeled as either arteries or veins by an independently validated fully automated algorithm. Segmentation labels were validated by a radiologist. Results: The segmentation algorithm reached a median accuracy of 90%, aligning with the radiologist’s assessments. Vessel density of pulmonary arteries with diameters between 6 and 10 mm was higher in patients with versus those without PH (median [inter-quartile range]: 8.9 [6.1–10.8] 1/L vs. 6.2 [3.1–7.0] 1/L; p = 0.007). Artery-to-vein ratio was higher in PH (1.32 [0.93–2.06] vs. 0.88 [0.48–1.17], p = 0.004). The artery-to-vein ratio for vessels with diameters between 6 and 10 mm identified PH with an AUC of 0.73 (95% CI: 0.60–0.87). Combining this readout with the DMPA resulted in a numerically higher AUC (sole DMPA AUC: 0.79 (95% CI: 0.68–0.90)) vs. DMPA + artery-to-vein ratio for vessels with diameters of 6–10 mm: 0.81 (95% CI: 0.71–0.92); however, this improvement was not statistically significant (p = 0.4). Conclusions: PH is associated with an increased ratio of peripheral pulmonary arteries to veins within the 6–10 mm diameter range. Pulmonary vascular morphology may complement the established morphological criterion of MPA diameter and improve the diagnostic accuracy of PH on CT. Full article

(This article belongs to the Special Issue Medical Images Segmentation and Diagnosis)

► Show Figures

Figure 1

16 pages, 600 KB

Open AccessArticle

Prevalence and Distribution of Apical Periodontitis in Root Canal-Treated Teeth: A Cone-Beam Computed Tomography Study in a Saudi Subpopulation

by Obadah Austah, Lama Alghamdi, Amjad Alshamrani, Taggreed Wazzan, Mohammed Barayan, Mohammed A. Alharbi, Abdullah Bokhary and Loai Alsofi

Diagnostics 2026, 16(4), 618; https://doi.org/10.3390/diagnostics16040618 - 20 Feb 2026

Viewed by 478

Abstract

Background: Apical periodontitis (AP) is a common inflammatory condition of the periapical tissues, most often associated with persistent endodontic infection. Conventional two-dimensional radiography may underestimate AP because of anatomical superimposition and limited sensitivity. Cone-beam computed tomography (CBCT) allows three-dimensional visualization of periapical structures [...] Read more.

Background: Apical periodontitis (AP) is a common inflammatory condition of the periapical tissues, most often associated with persistent endodontic infection. Conventional two-dimensional radiography may underestimate AP because of anatomical superimposition and limited sensitivity. Cone-beam computed tomography (CBCT) allows three-dimensional visualization of periapical structures and has been increasingly used in epidemiological research. Objective: This study aimed to evaluate the prevalence and distribution of apical periodontitis, with particular emphasis on apical periodontitis associated with root canal-treated teeth (AP-RCT), in a Saudi subpopulation using CBCT imaging. Methods: This retrospective cross-sectional study analyzed CBCT scans of Saudi patients obtained for routine diagnostic purposes between 2017 and 2021. Apical periodontitis was identified using standardized radiographic criteria requiring the presence of periapical radiolucency in more than one imaging plane. Demographic and clinical variables were recorded. Descriptive statistics were used to estimate prevalence. Associations between demographic factors and AP-RCT counts were evaluated using multivariable negative binomial regression. Regional tooth distribution was analyzed using generalized estimating equation models accounting for within-participant clustering. Results: A total of 320 CBCT scans were analyzed. Apical periodontitis was detected in 231 participants (72.2%) and in 667 teeth (8.3% of examined teeth). Of the affected teeth, 457 (68.5%) were associated with root canal treatment. The mean number of AP-RCT per participant was 1.36 ± 1.81 (median: 1; IQR: 0–2). Multivariable analysis identified age as the only significant predictor of AP-RCT. Compared with individuals aged 21–30 years, higher AP-RCT rates were observed in the 31–40-year and 41–50-year age groups, while participants ≤20 years showed lower rates. Tooth-level analysis demonstrated higher AP-RCT prevalence in maxillary premolars, maxillary molars, and mandibular molars, whereas mandibular anterior teeth showed the lowest prevalence. Conclusions: Apical periodontitis, particularly AP-RCT, was frequently observed in this Saudi subpopulation when assessed using CBCT. Age and tooth location were the primary determinants of disease distribution. These findings provide population-level epidemiological data on the prevalence and anatomical distribution of apical periodontitis in root canal-treated teeth. Clinical Significance: CBCT-based epidemiological assessment enables detailed evaluation of the distribution of apical periodontitis in dentate populations and may assist in characterizing disease patterns in anatomically complex regions, without implying comparative diagnostic accuracy or treatment outcome assessment. Full article

(This article belongs to the Special Issue Advances in Dental Diagnostics)

► Show Figures

Figure 1

12 pages, 1034 KB

Open AccessCase Report

Cervical Adenocarcinoma In Situ in Young Nulliparous Patient with Persistent ASC-US and Multiple-Type HPV Infections Without HPV 16 and 18 Types—Case Report

by Nikola Milic, Marija Varnicic Lojanica, Stefan Ivanovic, Milica Ivanovic, Katarina Ivanovic and Nikola Jovic

Diagnostics 2026, 16(4), 617; https://doi.org/10.3390/diagnostics16040617 - 20 Feb 2026

Viewed by 374

Abstract

The most severe premalignant lesion of glandular epithelium of the cervix is adenocarcinoma in situ (AIS). In most cases it is associated with persistent human papillomavirus (HPV) infection and most often occurs in women in the fourth decade of life. In most high-income [...] Read more.

The most severe premalignant lesion of glandular epithelium of the cervix is adenocarcinoma in situ (AIS). In most cases it is associated with persistent human papillomavirus (HPV) infection and most often occurs in women in the fourth decade of life. In most high-income countries, primary screening has shifted to HPV testing, while cytology is used for patient triage. Even with current robust screening protocols, their sensitivity for glandular lesions remains limited. Diagnosis of AIS obtained by biopsy, brushing or curettage is confirmed by excisional methods and pathohistological verification. Therapy depends on the patient’s lifestyle and reproductive age. In our case, we present a nulliparous patient with persistent ASC-US, multiple-type HPV infection without HPV 16 and 18 types, and AIS which was diagnosed after conization, follow-up and two biopsies with curettage of cervical canal. Our case report highlights limitations in detection of glandular lesions and need for caution in patients with persistent and seemingly low-grade cytological abnormalities, notably in young patients with high-risk HPV types. Full article

(This article belongs to the Section Diagnostic Microbiology and Infectious Disease)

► Show Figures

Figure 1

14 pages, 499 KB

Open AccessArticle

Decoding AI Competence: Benchmarking Large Language Models (LLMs) in Ovarian Cancer Diagnosis and Treatment—A Systematic Evaluation of Generative AI Accuracy and Completeness

by Haojie Cai, Chao Wang, Yue Zhang, Hui Ding, Wei Hong, Yaqian Zhao, Shanshan Cheng and Yu Wang

Diagnostics 2026, 16(4), 616; https://doi.org/10.3390/diagnostics16040616 - 20 Feb 2026

Viewed by 498

Abstract

Objective: To evaluate the practical value of DeepSeek-R1 and Doubao-1.5-pro in the context of ovarian cancer management by examining their diagnostic and treatment-related competencies. Methods: 20 key ovarian cancer diagnosis and treatment issues were identified, divided into 4 domains with 5 [...] Read more.

Objective: To evaluate the practical value of DeepSeek-R1 and Doubao-1.5-pro in the context of ovarian cancer management by examining their diagnostic and treatment-related competencies. Methods: 20 key ovarian cancer diagnosis and treatment issues were identified, divided into 4 domains with 5 questions each. Two large language models answered these questions, and 5 gynecologic oncology chief physicians evaluated the answers on a 1–10 scale for completeness and accuracy. For each score and the mean score for each question, if it surpassed 7, it is evaluated as “Excellent.” The Kruskal–Wallis test compared scores within each LLM across 4 categories, and the Mann–Whitney-Wilcoxon test compared scores between the two LLMs in each category. Results: 200 scores were collected (100 per model). DeepSeek-R1 got 98 “Excellent” ratings, while Doubao-1.5-pro got 41. All 20 DeepSeek-R1 responses had “Excellent” average scores, compared to 9 for Doubao-1.5-pro. DeepSeek-R1 had less variability. Tests revealed significant differences between the models and showed DeepSeek-R1 outperformed Doubao-1.5-pro, and charts showed Doubao-1.5-pro scored lower in all aspects, especially “Medical”. Conclusions: DeepSeek-R1 shows potential in ovarian cancer diagnosis and treatment but has limitations like inaccuracies and overly technical responses due to outdated data and lack of humanistic elements. LLMs like DeepSeek-R1 are useful for medical education and assistive diagnosis, but they require ongoing updates and refinement for broader clinical use. Selecting the appropriate LLM for medical tasks and improving their clarity and accuracy is crucial for their future effectiveness. Full article

(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)

► Show Figures

Figure 1

31 pages, 2801 KB

Open AccessArticle

Intelligent Neurovascular Imaging Engine (INIE): Topology-Aware Compressed Sensing and Multimodal Super-Resolution for Real-Time Guidance in Clinically Relevant Porcine Stroke Recanalization

by Krzysztof Malczewski, Ryszard Kozera, Zdzislaw Gajewski and Maria Sady

Diagnostics 2026, 16(4), 615; https://doi.org/10.3390/diagnostics16040615 - 20 Feb 2026

Viewed by 448

Abstract

Introduction: Rapid and reliable neurovascular imaging is critical for time-sensitive diagnosis in acute cerebrovascular disorders, yet conventional magnetic resonance imaging (MRI) workflows remain constrained by acquisition speed, motion sensitivity, and limited integration of physiological context. We introduce the Intelligent Neurovascular Imaging Engine (INIE), [...] Read more.

Introduction: Rapid and reliable neurovascular imaging is critical for time-sensitive diagnosis in acute cerebrovascular disorders, yet conventional magnetic resonance imaging (MRI) workflows remain constrained by acquisition speed, motion sensitivity, and limited integration of physiological context. We introduce the Intelligent Neurovascular Imaging Engine (INIE), a sensor-informed, topology-aware framework that jointly optimizes accelerated data acquisition, physics-grounded reconstruction, and cross-scale physiological consistency. Methods: INIE combines adaptive sampling, structured low-rank (Hankel) priors, and topology-preserving objectives with multimodal physiological sensors and scanner telemetry, enabling phase-consistent gating and confidence-weighted reconstruction under realistic operating conditions. The framework was evaluated using synthetic phantoms, a translational porcine stroke recanalization model with repeated measures, and retrospective human datasets. Across

N_{runs} = 120

acquisition–reconstruction runs derived from

N_{animals} = 18

pigs with animal-level train/validation/test separation, performance was assessed using image quality, topological fidelity, and cross-modal consistency metrics. Multiple-comparison control was performed using Bonferroni/Holm–Bonferroni procedures. Results: INIE achieved acquisition acceleration exceeding 70% while maintaining high reconstruction fidelity (PSNR

\approx 35

–36 dB, SSIM

\approx 0.90

–0.92). Topology-aware analysis showed an approximately twofold reduction in Betti number deviation relative to baseline accelerated methods. Cross-modal validation in a PET subset demonstrated strong agreement between MRI-derived perfusion parameters and metabolic markers (Pearson

r \approx 0.9

). INIE improved large-vessel occlusion detection accuracy to approximately 93% and reduced automated time-to-decision to under three minutes. Conclusions: These results indicate that sensor-informed, topology-aware, closed-loop imaging improves the reliability and physiological consistency of accelerated neurovascular MRI and supports faster, more robust decision-making in acute cerebrovascular imaging workflows. Full article

(This article belongs to the Section Medical Imaging and Theranostics)

► Show Figures

Graphical abstract

Journal Menu

Journal Browser

Diagnostics, Volume 16, Issue 4 (February-2 2026) – 140 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI