Bioengineering, Volume 10, Issue 8 (August 2023) – 110 articles

Robotic rehabilitation is one of the most advanced treatments helping people with stroke to faster recovery from motor deficits. The clinical impact of this type of treatment has been widely defined and established using clinical scales. The neurofunctional indicators of motor recovery following conventional rehabilitation treatments have already been identified by previous meta-analytic investigations. However, a clear definition of the neural correlates associated with robotic neurorehabilitation treatment has never been performed. This systematic review assesses the neurofunctional correlates (fMRI, fNIRS) of cutting-edge robotic therapies in enhancing motor recovery of stroke populations in accordance with PRISMA standards. A total of 7, of the initial yield of 150 articles, have been included in this review. Lessons from these studies suggest that neural plasticity within the ipsilateral primary motor cortex, the contralateral sensorimotor cortex, and the premotor cortices are more sensitive to compensation strategies reflecting upper and lower limbs’ motor recovery despite the high heterogeneity in robotic devices, clinical status, and neuroimaging procedures. Unfortunately, the paucity of RCT studies prevents us from understanding the neurobiological differences induced by robotic devices with respect to traditional rehabilitation approaches. Despite this technology dating to the early 1990s, there is a need to translate more functional neuroimaging markers in clinical settings since they provide a unique opportunity to examine, in-depth, the brain plasticity changes induced by robotic rehabilitation. Full article

Three-dimensional reconstruction of the corneal surface provides a powerful tool for managing corneal diseases. This study proposes a novel method for reconstructing the corneal surface from elevation point clouds, using modal schemes capable of reproducing corneal shapes using surface polynomial functions. The multivariable polynomial fitting was performed using a non-dominated sorting multivariable genetic algorithm (NS-MVGA). Standard reconstruction methods using least-squares discrete fitting (LSQ) and sequential quadratic programming (SQP) were compared with the evolutionary algorithm-based approach. The study included 270 corneal surfaces of 135 eyes of 102 patients (ages 11–63) sorted in two groups: control (66 eyes of 33 patients) and keratoconus (KC) (69 eyes of 69 patients). Tomographic information (Sirius, Costruzione Strumenti Oftalmici, Italy) was processed using Matlab. The goodness of fit for each method was evaluated using mean squared error (MSE), measured at the same nodes where the elevation data were collected. Polynomial fitting based on NS-MVGA improves MSE values by 86% compared to LSQ-based methods in healthy patients. Moreover, this new method improves aberrated surface reconstruction by an average value of 56% if compared with LSQ-based methods in keratoconus patients. Finally, significant improvements were also found in morpho-geometric parameters, such as asphericity and corneal curvature radii. Full article

Pain is considered an unpleasant perceptual experience associated with actual or potential somatic and visceral harm. Human subjects have different sensitivity to painful stimulation, which may be related to different painful response pattern. Excellent studies using functional magnetic resonance imaging (fMRI) have found the effect of the functional organization of white matter (WM) on the descending pain modulatory system, which suggests that WM function is feasible during pain modulation. In this study, 26 pain sensitive (PS) subjects and 27 pain insensitive (PIS) subjects were recruited based on cold pressor test. Then, all subjects underwent the cold bottle test (CBT) in normal (26 degrees temperature stimulating) and cold (8 degrees temperature stimulating) conditions during fMRI scan, respectively. WM functional networks were obtained using K-means clustering, and the functional connectivity (FC) was assessed among WM networks, as well as gray matter (GM)–WM networks. Through repeated measures ANOVA, decreased FC was observed between the GM–cerebellum network and the WM–superior temporal network, as well as the WM–sensorimotor network in the PS group under the cold condition, while this difference was not found in PIS group. Importantly, the changed FC was positively correlated with the state and trait anxiety scores, respectively. This study highlighted that the WM functional network might play an integral part in pain processing, and an altered FC may be related to the descending pain modulatory system. Full article

Aging and age-related changes impact the quality of life (QOL) in elderly with a decline in movement, cognitive abilities and increased vulnerability towards age-related diseases (ARDs). One of the key contributing factors is cellular senescence, which is triggered majorly by DNA damage response (DDR). Accumulated senescent cells (SCs) release senescence-associated secretory phenotype (SASP), which includes pro-inflammatory cytokines, matrix metalloproteinases (MMPs), lipids and chemokines that are detrimental to the surrounding tissues. Chronic low-grade inflammation in the elderly or inflammaging is also associated with cellular senescence and contributes to ARDs. The literature from the last decade has recorded the use of platelet rich plasma (PRP) to combat senescence and inflammation, alleviate pain as an analgesic, promote tissue regeneration and repair via angiogenesis—all of which are essential in anti-aging and tissue regeneration strategies. In the last few decades, platelet-rich plasma (PRP) has been used as an anti-aging treatment option for dermatological applications and with great interest in tissue regeneration for orthopaedic applications, especially in osteoarthritis (OA). In this exploration, we connect the intricate relationship between aging, ARDs, senescence and inflammation and delve into PRP’s properties and potential benefits. We conduct a comparative review of the current literature on PRP treatment strategies, paying particular attention to the instances strongly linked to ARDs. Finally, upon careful consideration of this interconnected information in the context of aging, we suggest a prospective role for PRP in developing anti-aging therapeutic strategies. Full article

Cellulose micro/nanomaterials (CMNMs) are innovative materials with a wide spectrum of industrial and biomedical applications. Although cellulose has been recognized as a safe material, the unique properties of its nanosized forms have raised concerns about their safety for human health. Genotoxicity is an endpoint that must be assessed to ensure that no carcinogenic risks are associated with exposure to nanomaterials. In this study, we evaluated the genotoxicity of two types of cellulose micro/nanofibrils (CMF and CNF) and one sample of cellulose nanocrystals (CNC), obtained from industrial bleached Eucalyptus globulus kraft pulp. For that, we exposed co-cultures of human alveolar epithelial A549 cells and THP-1 monocyte-derived macrophages to a concentration range of each CMNM and used the micronucleus (MN) and comet assays. Our results showed that only the lowest concentrations of the CMF sample were able to induce DNA strand breaks (FPG-comet assay). However, none of the three CMNMs produced significant chromosomal alterations (MN assay). These findings, together with results from previous in vitro studies using monocultures of A549 cells, indicate that the tested CNF and CNC are not genotoxic under the conditions tested, while the CMF display a low genotoxic potential. Full article

The vibration of the heart valves’ closure is an important component of the heart sound and contains important information about the mechanical activity of a heart. Stenosis of the distal pulmonary artery can lead to pulmonary hypertension (PH). Therefore, in this paper, the relationship between the vibration sound of heart valves and the pulmonary artery blood pressure was investigated to contribute to the noninvasive detection of PH. In this paper, a lumped parameter circuit platform of pulmonary circulation was first set to guide the establishment of a mock loop of circulation. By adjusting the distal vascular resistance of the pulmonary artery, six different pulmonary arterial pressure states were achieved. In the experiment, pulmonary artery blood pressure, right ventricular blood pressure, and the vibration sound of the pulmonary valve and tricuspid valve were measured synchronously. Features of the time domain and frequency domain of two valves’ vibration sound were extracted. By conducting a significance analysis of the inter-group features, it was found that the amplitude, energy and frequency features of vibration sounds changed significantly. Finally, the continuously varied pulmonary arterial blood pressure and valves’ vibration sound were obtained by continuously adjusting the resistance of the distal pulmonary artery. A backward propagation neural network and deep learning model were used, respectively, to estimate the features of pulmonary arterial blood pressure, pulmonary artery systolic blood pressure, the maximum rising rate of pulmonary artery blood pressure and the maximum falling rate of pulmonary artery blood pressure by the vibration sound of the pulmonary and tricuspid valves. The results showed that the pulmonary artery pressure parameters can be well estimated by valve vibration sounds. Full article

Recent research has achieved a great classification rate for separating healthy people from those with Parkinson’s disease (PD) using speech and the voice. However, these studies have primarily treated early and advanced stages of PD as equal entities, neglecting the distinctive speech impairments and other symptoms that vary across the different stages of the disease. To address this limitation, and improve diagnostic precision, this study assesses the selected acoustic features of dysphonia, as they relate to PD and the Hoehn and Yahr stages, by combining various preprocessing techniques and multiple classification algorithms, to create a comprehensive and robust solution for classification tasks. The dysphonia features extracted from the three sustained Korean vowels /아/(a), /이/(i), and /우/(u) exhibit diversity and strong correlations. To address this issue, the analysis of variance F-Value feature selection classifier from scikit-learn was employed, to identify the topmost relevant features. Additionally, to overcome the class imbalance problem, the synthetic minority over-sampling technique was utilized. To ensure fair comparisons, and mitigate the influence of individual classifiers, four commonly used machine learning classifiers, namely random forest (RF), support vector machine (SVM), k-nearest neighbor (kNN), and multi-layer perceptron (MLP), were employed. This approach enables a comprehensive evaluation of the feature extraction methods, and minimizes the variance in the final classification models. The proposed hybrid machine learning pipeline using the acoustic features of sustained vowels efficiently detects the early and mid-advanced stages of PD with a detection accuracy of 95.48%, and with a detection accuracy of 86.62% for the 4-stage, and a detection accuracy of 89.48% for the 3-stage classification of PD. This study successfully demonstrates the significance of utilizing the diverse acoustic features of dysphonia in the classification of PD and its stages. Full article

Ophthalmic disorders consist of a broad spectrum of ailments that impact the structures and functions of the eye. Due to the crucial function of the retina in the vision process, the management of eye ailments is of the utmost importance, but several unmet needs have been identified in terms of the outcome measures in clinical trials, more proven minimally invasive glaucoma surgery, and a lack of comprehensive bibliometric assessments, among others. The current evaluation seeks to fulfill several of these unmet needs via a dual approach consisting of a molecular docking analysis based on the potential of ripasudil and fasudil to inhibit Rho-associated protein kinases (ROCKs), virtual screening of ligands, and pharmacokinetic predictions, emphasizing the identification of new compounds potentially active in the management of glaucoma, and a comprehensive bibliometric analysis of the most recent publications indexed in the Web of Science evaluating the management of several of the most common eye conditions. This method resulted in the finding of ligands (i.e., ZINC000000022706 with the most elevated binding potential for ROCK1 and ZINC000034800307 in the case of ROCK2) that are not presently utilized in any therapeutic regimen but may represent a future option to be successfully applied in the therapeutic scheme of glaucoma following further comprehensive testing validations. In addition, this research also analyzed multiple papers listed in the Web of Science collection of databases via the VOSviewer application to deliver, through descriptive analysis of the results, an in-depth overview of publications contributing to the present level of comprehension in therapeutic approaches to ocular diseases in terms of scientific impact, citation analyses, most productive authors, journals, and countries, as well as collaborative networks. Based on the molecular docking study’s preliminary findings, the most promising candidates must be thoroughly studied to determine their efficacy and risk profiles. Bibliometric analysis may also help researchers set targets to improve ocular disease outcomes. Full article

Monolateral pin-to-bar-clamp fixators are commonly used to stabilize acute extremity injuries. Certain rules regarding frame geometry have been established that affect construct stability. The influence of sagittal pin angulation on construct stiffness and strength has not been investigated. The purpose of this biomechanical study was to demonstrate the effect of a pin angulation in the monolateral fixator using a composite cylinder model. Three groups of composite cylinder models with a fracture gap were loaded with different mounting variants of monolateral pin-to-bar-clamp fixators. In the first group, the pins were set parallel to each other and perpendicular to the specimen. In the second group, both pins were set convergent each in an angle of 15° to the specimen. In the third group, the pins were set each 15° divergent. The strength of the constructions was tested using a mechanical testing machine. This was followed by a cyclic loading test to produce pin loosening. A pull-out test was then performed to evaluate the strength of each construct at the pin–bone interface. Initial stiffness analyses showed that the converging configuration was the stiffest, while the diverging configuration was the least stiff. The parallel mounting showed an intermediate stiffness. There was a significantly higher resistance to pull-out force in the diverging pin configuration compared to the converging pin configuration. There was no significant difference in the pull-out strength of the parallel pins compared to the angled pin pairs. Convergent mounting of pin pairs increases the stiffness of a monolateral fixator, whereas a divergent mounting weakens it. Regarding the strength of the pin–bone interface, the divergent pin configuration appears to provide greater resistance to pull-out force than the convergent one. The results of this pilot study should be important for the doctrine of fixator mounting as well as for fixator component design. Full article

Background: Lung cancer is one of the most fatal cancers worldwide, and malignant tumors are characterized by the growth of abnormal cells in the tissues of lungs. Usually, symptoms of lung cancer do not appear until it is already at an advanced stage. The proper segmentation of cancerous lesions in CT images is the primary method of detection towards achieving a completely automated diagnostic system. Method: In this work, we developed an improved hybrid neural network via the fusion of two architectures, MobileNetV2 and UNET, for the semantic segmentation of malignant lung tumors from CT images. The transfer learning technique was employed and the pre-trained MobileNetV2 was utilized as an encoder of a conventional UNET model for feature extraction. The proposed network is an efficient segmentation approach that performs lightweight filtering to reduce computation and pointwise convolution for building more features. Skip connections were established with the Relu activation function for improving model convergence to connect the encoder layers of MobileNetv2 to decoder layers in UNET that allow the concatenation of feature maps with different resolutions from the encoder to decoder. Furthermore, the model was trained and fine-tuned on the training dataset acquired from the Medical Segmentation Decathlon (MSD) 2018 Challenge. Results: The proposed network was tested and evaluated on 25% of the dataset obtained from the MSD, and it achieved a dice score of 0.8793, recall of 0.8602 and precision of 0.93. It is pertinent to mention that our technique outperforms the current available networks, which have several phases of training and testing. Full article

Background: The purpose of this study was to determine the force production during self-selected speed normal gait by muscle–tendon units that cross the knee. The force of a single knee muscle is not directly measurable without invasive methods, yet invasive techniques are not appropriate for clinical use. Thus, an EMG-to-force processing (EFP) model was developed which scaled muscle–tendon unit (MTU) force output to gait EMG. Methods: An EMG-to-force processing (EFP) model was developed which scaled muscle–tendon unit (MTU) force output to gait EMG. Active muscle force power was defined as the product of MTU forces (derived from EFP) and that muscle’s contraction velocity. Net knee EFP moment was determined by summing individual active knee muscle moments. Net knee moments were also calculated for these study participants via inverse dynamics (kinetics plus kinematics, KIN). The inverse dynamics technique used are well accepted and the KIN net moment was used to validate or reject this model. Closeness of fit of the moment power curves for the two methods (during active muscle forces) was used to validate the model. Results: The correlation between the EFP and KIN methods was sufficiently close, suggesting validation of the model’s ability to provide reasonable estimates of knee muscle forces. Conclusions: The EMG-to-force processing approach provides reasonable estimates of active individual knee muscle forces in self-selected speed walking in neurologically intact adults. Full article

One of the most promising research initiatives in the healthcare field is focused on the rising incidence of skin cancer worldwide and improving early discovery methods for the disease. The most significant factor in the fatalities caused by skin cancer is the late identification of the disease. The likelihood of human survival may be significantly improved by performing an early diagnosis followed by appropriate therapy. It is not a simple process to extract the elements from the photographs of the tumors that may be used for the prospective identification of skin cancer. Several deep learning models are widely used to extract efficient features for a skin cancer diagnosis; nevertheless, the literature demonstrates that there is still room for additional improvements in various performance metrics. This study proposes a hybrid deep convolutional neural network architecture for identifying skin cancer by adding two main heuristics. These include Xception and MobileNetV2 models. Data augmentation was introduced to balance the dataset, and the transfer learning technique was utilized to resolve the challenges of the absence of labeled datasets. It has been detected that the suggested method of employing Xception in conjunction with MobileNetV2 attains the most excellent performance, particularly concerning the dataset that was evaluated: specifically, it produced 97.56% accuracy, 97.00% area under the curve, 100% sensitivity, 93.33% precision, 96.55% F1 score, and 0.0370 false favorable rates. This research has implications for clinical practice and public health, offering a valuable tool for dermatologists and healthcare professionals in their fight against skin cancer. Full article

Retinopathy, a prevalent disease causing visual impairment and sometimes blindness, affects many individuals in the population. Early detection and treatment of the disease can be facilitated by monitoring the retina using fundus imaging. Nonetheless, the limited availability of fundus images and the imbalanced datasets warrant the development of more precise and efficient algorithms to enhance diagnostic performance. This study presents a novel online knowledge distillation framework, called CLRD, which employs a collaborative learning approach for detecting retinopathy. By combining student models with varying scales and architectures, the CLRD framework extracts crucial pathological information from fundus images. The transfer of knowledge is accomplished by developing distortion information particular to fundus images, thereby enhancing model invariance. Our selection of student models includes the Transformer-based BEiT and the CNN-based ConvNeXt, which achieve accuracies of 98.77% and 96.88%, respectively. Furthermore, the proposed method has 5.69–23.13%, 5.37–23.73%, 5.74–23.17%, 11.24–45.21%, and 5.87–24.96% higher accuracy, precision, recall, specificity, and F1 score, respectively, compared to the advanced visual model. The results of our study indicate that the CLRD framework can effectively minimize generalization errors without compromising independent predictions made by student models, offering novel directions for further investigations into detecting retinopathy. Full article

At present, the double-enzyme catalyzed method using maltooligosyltrehalose synthase (MTSase) and maltooligosyltrehalose trehalohydrolase (MTHase) is the mainstream technology for industrial trehalose production. However, MTSase and MTHase are prepared mainly using the heterologous expression in the engineered Escherichia coli strains so far. In this study, we first proved that the addition of 3 U/g neutral pullulanase PulA could enhance the trehalose conversion rate by 2.46 times in the double-enzyme catalyzed system. Then, a CBM68 domain was used to successfully assist the secretory expression of MTSase and MTHase from Arthrobacter ramosus S34 in Bacillus subtilis SCK6. At the basis, an engineered strain B. subtilis PSH02 (amyE::pulA/pHT43-C68-ARS/pMC68-ARH), which co-expressed MTSase, MTHase, and PulA, was constructed. After the 24 h fermentation of B. subtilis PSH02, the optimum ratio of the extracellular multi-enzymes was obtained to make the highest trehalose conversion rate of 80% from 100 g/L maltodextrin. The high passage stability and multi-enzyme preservation stability made B. subtilis PSH02 an excellent industrial production strain. Moreover, trehalose production using these extracellular enzymes produced via the one-step fermentation of B. subtilis PSH02 would greatly simplify the procedure for multi-enzyme preparation and be expected to reduce production costs. Full article

Rehabilitation robots are helpful in poststroke telerehabilitation; however, their feasibility and rehabilitation effectiveness in clinical settings have not been sufficiently investigated. A non-randomized controlled trial was conducted to investigate the feasibility of translating a telerehabilitation program assisted by a mobile wrist/hand exoneuromusculoskeleton (WH-ENMS) into routine clinical services and to compare the rehabilitative effects achieved in the hospital-service-based group (n = 12, clinic group) with the laboratory-research-based group (n = 12, lab group). Both groups showed significant improvements (p ≤ 0.05) in clinical assessments of behavioral motor functions and in muscular coordination and kinematic evaluations after the training and at the 3-month follow-up, with the lab group demonstrating better motor gains than the clinic group (p ≤ 0.05). The results indicated that the WH-ENMS-assisted tele-program was feasible and effective for upper limb rehabilitation when integrated into routine practice, and the quality of patient–operator interactions physically and remotely affected the rehabilitative outcomes. Full article

The primary aim of this study was to analyze studies that use electrochemotherapy (ECT) in “deep-seated” tumors in solid organs (liver, kidney, bone metastasis, pancreas, and abdomen) and understand the similarities between patient selection, oncologic selection, and use of new procedures and technology across the organ systems to assess response rates. A literature search was conducted using the term “Electrochemotherapy” in the title field using publications from 2017 to 2023. After factoring in inclusion and exclusion criteria, 29 studies were analyzed and graded based on quality in full. The authors determined key patient and oncologic selection characteristics and ECT technology employed across organ systems that yielded overall responses, complete responses, and partial responses of the treated tumor. It was determined that key selection factors included: the ability to be administered bleomycin, life expectancy greater than three months, unrespectability of the lesion being treated, and a later stage, more advanced cancer. Regarding oncologic selection, all patient cohorts had received chemotherapy or surgery previously but had disease recurrence, making ECT the only option for further treatment. Lastly, in terms of the use of technology, the authors found that studies with better response rates used the ClinporatorTM and updated procedural guidelines by SOP. Thus, by considering patient, oncologic, and technology selection, ECT can be further improved in treating lesions in solid organs. Full article

Background: The recent development of deep neural network models for the analysis of breast images has been a breakthrough in computer-aided diagnostics (CAD). Contrast-enhanced mammography (CEM) is a recent mammography modality providing anatomical and functional imaging of the breast. Despite the clinical benefits it could bring, only a few research studies have been conducted around deep-learning (DL) based CAD for CEM, especially because the access to large databases is still limited. This study presents the development and evaluation of a CEM-CAD for enhancing lesion detection and breast classification. Materials & Methods: A deep learning enhanced cancer detection model based on a YOLO architecture has been optimized and trained on a large CEM dataset of 1673 patients (7443 images) with biopsy-proven lesions from various hospitals and acquisition systems. The evaluation was conducted using metrics derived from the free receiver operating characteristic (FROC) for the lesion detection and the receiver operating characteristic (ROC) to evaluate the overall breast classification performance. The performances were evaluated for different types of image input and for each patient background parenchymal enhancement (BPE) level. Results: The optimized model achieved an area under the curve (AUROC) of 0.964 for breast classification. Using both low-energy and recombined image as inputs for the DL model shows greater performance than using only the recombined image. For the lesion detection, the model was able to detect 90% of all cancers with a false positive (non-cancer) rate of 0.128 per image. This study demonstrates a high impact of BPE on classification and detection performance. Conclusion: The developed CEM CAD outperforms previously published papers and its performance is comparable to radiologist-reported classification and detection capability. Full article

Deep-learning-assisted medical diagnosis has brought revolutionary innovations to medicine. Breast cancer is a great threat to women’s health, and deep-learning-assisted diagnosis of breast cancer pathology images can save manpower and improve diagnostic accuracy. However, researchers have found that deep learning systems based on natural images are vulnerable to attacks that can lead to errors in recognition and classification, raising security concerns about deep systems based on medical images. We used the adversarial attack algorithm FGSM to reveal that breast cancer deep learning systems are vulnerable to attacks and thus misclassify breast cancer pathology images. To address this problem, we built a deep learning system for breast cancer pathology image recognition with better defense performance. Accurate diagnosis of medical images is related to the health status of patients. Therefore, it is very important and meaningful to improve the security and reliability of medical deep learning systems before they are actually deployed. Full article

Abdominal computed tomography (CT) is a frequently used imaging modality for evaluating gastrointestinal diseases. The detection of colorectal cancer is often realized using CT before a more invasive colonoscopy. When a CT exam is performed for indications other than colorectal evaluation, the tortuous structure of the long, tubular colon makes it difficult to analyze the colon carefully and thoroughly. In addition, the sensitivity of CT in detecting colorectal cancer is greatly dependent on the size of the tumor. Missed incidental colon cancers using CT are an emerging problem for clinicians and radiologists; consequently, the automatic localization of lesions in the CT images of unprepared bowels is needed. Therefore, this study used artificial intelligence (AI) to localize colorectal cancer in CT images. We enrolled 190 colorectal cancer patients to obtain 1558 tumor slices annotated by radiologists and colorectal surgeons. The tumor sites were double-confirmed via colonoscopy or other related examinations, including physical examination or image study, and the final tumor sites were obtained from the operation records if available. The localization and training models used were RetinaNet, YOLOv3, and YOLOv8. We achieved an F1 score of 0.97 (±0.002), a mAP of 0.984 when performing slice-wise testing, 0.83 (±0.29) sensitivity, 0.97 (±0.01) specificity, and 0.96 (±0.01) accuracy when performing patient-wise testing using our derived model YOLOv8 with hyperparameter tuning. Full article

One of the early manifestations of systemic atherosclerosis, which leads to blood circulation issues, is the enhanced arterial light reflex (EALR). Fundus images are commonly used for regular screening purposes to intervene and assess the severity of systemic atherosclerosis in a timely manner. However, there is a lack of automated methods that can meet the demands of large-scale population screening. Therefore, this study introduces a novel cross-scale transformer-based multi-instance learning method, named MIL-CT, for the detection of early arterial lesions (e.g., EALR) in fundus images. MIL-CT utilizes the cross-scale vision transformer to extract retinal features in a multi-granularity perceptual domain. It incorporates a multi-head cross-scale attention fusion module to enhance global perceptual capability and feature representation. By integrating information from different scales and minimizing information loss, the method significantly improves the performance of the EALR detection task. Furthermore, a multi-instance learning module is implemented to enable the model to better comprehend local details and features in fundus images, facilitating the classification of patch tokens related to retinal lesions. To effectively learn the features associated with retinal lesions, we utilize weights pre-trained on a large fundus image Kaggle dataset. Our validation and comparison experiments conducted on our collected EALR dataset demonstrate the effectiveness of the MIL-CT method in reducing generalization errors while maintaining efficient attention to retinal vascular details. Moreover, the method surpasses existing models in EALR detection, achieving an accuracy, precision, sensitivity, specificity, and F1 score of 97.62%, 97.63%, 97.05%, 96.48%, and 97.62%, respectively. These results exhibit the significant enhancement in diagnostic accuracy of fundus images brought about by the MIL-CT method. Thus, it holds potential for various applications, particularly in the early screening of cardiovascular diseases such as hypertension and atherosclerosis. Full article

Kidney–ureter–bladder (KUB) imaging is used as a frontline investigation for patients with suspected renal stones. In this study, we designed a computer-aided diagnostic system for KUB imaging to assist clinicians in accurately diagnosing urinary tract stones. The image dataset used for training and testing the model comprised 485 images provided by Kaohsiung Chang Gung Memorial Hospital. The proposed system was divided into two subsystems, 1 and 2. Subsystem 1 used Inception-ResNetV2 to train a deep learning model on preprocessed KUB images to verify the improvement in diagnostic accuracy with image preprocessing. Subsystem 2 trained an image segmentation model using the ResNet hybrid, U-net, to accurately identify the contours of renal stones. The performance was evaluated using a confusion matrix for the classification model. We conclude that the model can assist clinicians in accurately diagnosing renal stones via KUB imaging. Therefore, the proposed system can assist doctors in diagnosis, reduce patients’ waiting time for CT scans, and minimize the radiation dose absorbed by the body. Full article

We designed a photo-ECMO device to speed up the rate of carbon monoxide (CO) removal by using visible light to dissociate CO from hemoglobin (Hb). Using computational fluid dynamics, fillets of different radii (5 cm and 10 cm) were applied to the square shape of a photo-ECMO device to reduce stagnant blood flow regions and increase the treated blood volume while being constrained by full light penetration. The blood flow at different flow rates and the thermal load imposed by forty external light sources at 623 nm were modeled using the Navier-Stokes and convection–diffusion equations. The particle residence times were also analyzed to determine the time the blood remained in the device. There was a reduction in the blood flow stagnation as the fillet radii increased. The maximum temperature change for all the geometries was below 4 °C. The optimized device with a fillet radius of 5 cm and a blood priming volume of up to 208 cm³ should decrease the time needed to treat CO poisoning without exceeding the critical threshold for protein denaturation. This technology has the potential to decrease the time for CO removal when treating patients with CO poisoning and pulmonary gas exchange inhibition. Full article

Background: Our objective was to conduct a comprehensive analysis of the reproducibility of foot and ankle anthropometric measurements with a three-dimensional (3D) optical scanner. Methods: We evaluated thirty-nine different anthropometric parameters obtained with a 3D Laser UPOD-S Full-Foot Scanner in a healthy population of twenty subjects. We determined the variance of the measurements for each foot/ankle, and the average variance among different subjects. Results: For 40 feet and ankles (15 women and 5 men; mean age 35.62 +/− 9.54 years, range 9–75 years), the average variance was 1.4 ± 2 (range 0.1 to 8). Overall, the mean absolute measurement error was <1 mm, with a maximum variance percentage of 8.3%. Forefoot and midfoot circumferences had a low variance <2.5, with variance percentages <1%. Hindfoot circumferences, malleolar heights, and the length of the first and fifth metatarsal to the ground contact points showed the highest variance (range 1 to 7). Conclusions: The UPOD-S Full-Foot optical Scanner achieved a good reproducibility in a large set of foot and ankle anthropometric measurements. It is a valuable tool for clinical and research purposes. Full article

Subarachnoid hemorrhage (SAH) denotes a serious type of hemorrhagic stroke that often leads to a poor prognosis and poses a significant socioeconomic burden. Timely assessment of the prognosis of SAH patients is of paramount clinical importance for medical decision making. Currently, clinical prognosis evaluation heavily relies on patients’ clinical information, which suffers from limited accuracy. Non-contrast computed tomography (NCCT) is the primary diagnostic tool for SAH. Radiomics, an emerging technology, involves extracting quantitative radiomics features from medical images to serve as diagnostic markers. However, there is a scarcity of studies exploring the prognostic prediction of SAH using NCCT radiomics features. The objective of this study is to utilize machine learning (ML) algorithms that leverage NCCT radiomics features for the prognostic prediction of SAH. Retrospectively, we collected NCCT and clinical data of SAH patients treated at Beijing Hospital between May 2012 and November 2022. The modified Rankin Scale (mRS) was utilized to assess the prognosis of patients with SAH at the 3-month mark after the SAH event. Based on follow-up data, patients were classified into two groups: good outcome (mRS ≤ 2) and poor outcome (mRS > 2) groups. The region of interest in NCCT images was delineated using 3D Slicer software, and radiomic features were extracted. The most stable and significant radiomic features were identified using the intraclass correlation coefficient, t-test, and least absolute shrinkage and selection operator (LASSO) regression. The data were randomly divided into training and testing cohorts in a 7:3 ratio. Various ML algorithms were utilized to construct predictive models, encompassing logistic regression (LR), support vector machine (SVM), random forest (RF), light gradient boosting machine (LGBM), adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost), and multi-layer perceptron (MLP). Seven prediction models based on radiomic features related to the outcome of SAH patients were constructed using the training cohort. Internal validation was performed using five-fold cross-validation in the entire training cohort. The receiver operating characteristic curve, accuracy, precision, recall, and f-1 score evaluation metrics were employed to assess the performance of the classifier in the overall dataset. Furthermore, decision curve analysis was conducted to evaluate model effectiveness. The study included 105 SAH patients. A comprehensive set of 1316 radiomics characteristics were initially derived, from which 13 distinct features were chosen for the construction of the ML model. Significant differences in age were observed between patients with good and poor outcomes. Among the seven constructed models, model_SVM exhibited optimal outcomes during a five-fold cross-validation assessment, with an average area under the curve (AUC) of 0.98 (standard deviation: 0.01) and 0.88 (standard deviation: 0.08) on the training and testing cohorts, respectively. In the overall dataset, model_SVM achieved an accuracy, precision, recall, f-1 score, and AUC of 0.88, 0.84, 0.87, 0.84, and 0.82, respectively, in the testing cohort. Radiomics features associated with the outcome of SAH patients were successfully obtained, and seven ML models were constructed. Model_SVM exhibited the best predictive performance. The radiomics model has the potential to provide guidance for SAH prognosis prediction and treatment guidance. Full article

Automatic medical report generation based on deep learning can improve the efficiency of diagnosis and reduce costs. Although several automatic report generation algorithms have been proposed, there are still two main challenges in generating more detailed and accurate diagnostic reports: using multi-view images reasonably and integrating visual and semantic features of key lesions effectively. To overcome these challenges, we propose a novel automatic report generation approach. We first propose the Cross-View Attention Module to process and strengthen the multi-perspective features of medical images, using mean square error loss to unify the learning effect of fusing single-view and multi-view images. Then, we design the module Medical Visual-Semantic Long Short Term Memorys to integrate and record the visual and semantic temporal information of each diagnostic sentence, which enhances the multi-modal features to generate more accurate diagnostic sentences. Applied to the open-source Indiana University X-ray dataset, our model achieved an average improvement of $0.8\%$ over the state-of-the-art (SOTA) model on six evaluation metrics. This demonstrates that our model is capable of generating more detailed and accurate diagnostic reports. Full article

Taking COVID-19 as an example, we know that a pandemic can have a huge impact on normal human life and the economy. Meanwhile, the population flow between countries and regions is the main factor affecting the changes in a pandemic, which is determined by the airline network. Therefore, realizing the overall control of airports is an effective way to control a pandemic. However, this is restricted by the differences in prevention and control policies in different areas and privacy issues, such as how a patient’s personal data from a medical center cannot be effectively combined with their passenger personal data. This prevents more precise airport control decisions from being made. To address this, this paper designed a novel data-sharing framework (i.e., PPChain) based on blockchain and federated learning. The experiment uses a CPU i7-12800HX and uses Docker to simulate multiple virtual nodes. The model is deployed to run on an NVIDIA GeForce GTX 3090Ti GPU. The experiment shows that the relationship between a pandemic and aircraft transport can be effectively explored by PPChain without sharing raw data. This approach does not require centralized trust and improves the security of the sharing process. The scheme can help formulate more scientific and rational prevention and control policies for the control of airports. Additionally, it can use aerial data to predict pandemics more accurately. Full article

Traumatic injuries are a major cause of morbidity and mortality worldwide; however, there is limited research on microvascular traumatic injuries. To address this gap, this research aims to develop and optimise an in vitro construct for traumatic injury research at the microvascular level. Tissue engineering constructs were created using a range of polymers (collagen, fibrin, and gelatine), solvents (PBS, serum-free endothelial media, and MES/NaCl buffer), and concentrations (1–5% w/v). Constructs created from these hydrogels and HUVECs were evaluated to identify the optimal composition in terms of cell proliferation, adhesion, migration rate, viability, hydrogel consistency and shape retention, and tube formation. Gelatine hydrogels were associated with a lower cell adhesion, whereas fibrin and collagen ones displayed similar or better results than the control, and collagen hydrogels exhibited poor shape retention; fibrin scaffolds, particularly at high concentrations, displayed good hydrogel consistency. Based on the multipronged evaluation, fibrin hydrogels in serum-free media at 3 and 5% w/v were selected for further experimental work and enabled the formation of interconnected capillary-like networks. The networks formed in both hydrogels displayed a similar architecture in terms of the number of segments (10.3 ± 3.21 vs. 9.6 ± 3.51) and diameter (8.6446 ± 3.0792 $\mathsf{\mu }$m vs. 7.8599 ± 2.3794 $\mathsf{\mu }$m). Full article

(1) Background: This study aims to develop a deep learning model based on a 3D Deeplab V3+ network to automatically segment multiple structures from magnetic resonance (MR) images at the L4/5 level. (2) Methods: After data preprocessing, the modified 3D Deeplab V3+ network of the deep learning model was used for the automatic segmentation of multiple structures from MR images at the L4/5 level. We performed five-fold cross-validation to evaluate the performance of the deep learning model. Subsequently, the Dice Similarity Coefficient (DSC), precision, and recall were also used to assess the deep learning model’s performance. Pearson’s correlation coefficient analysis and the Wilcoxon signed-rank test were employed to compare the morphometric measurements of 3D reconstruction models generated by manual and automatic segmentation. (3) Results: The deep learning model obtained an overall average DSC of 0.886, an average precision of 0.899, and an average recall of 0.881 on the test sets. Furthermore, all morphometry-related measurements of 3D reconstruction models revealed no significant difference between ground truth and automatic segmentation. Strong linear relationships and correlations were also obtained in the morphometry-related measurements of 3D reconstruction models between ground truth and automated segmentation. (4) Conclusions: We found it feasible to perform automated segmentation of multiple structures from MR images, which would facilitate lumbar surgical evaluation by establishing 3D reconstruction models at the L4/5 level. Full article

The resazurin reduction test is one of the basic tests for bacterial culture viability and drug resistance endorsed by the World Health Organisation. At the same time, conventional spectrophotometric and spectrofluorimetric methods demand rather bulky and expensive equipment. This induces a challenge for developing simpler approaches to sensor systems that are portable and applicable in resource-limited settings. In this work, we address two such alternative approaches, based on the colour processing of the microbiological plate’s photographic images and single-channel photometry with a recently developed portable microbiological analyser. The key results consist of establishing a sequential linear correspondence between the concentration of resorufin produced due to the reduction of resazurin by viable bacteria as determined by the UV-Vis studies, the intensity of the a* channel of the CIE L*a*b* colour space and the transmitted light intensity registered by a luxmeter under the LED illumination with a yellow colour filter. This route is illustrated with the chemical system “Hydrazine hydrate – resazurin”, isolating the target colour change-inducing reaction and the test of determining the minimal inhibition concentration of the antibacterial first-line drug isoniazid acting on the culture of the H37Rv strain of M. tuberculosis. Full article

SARS-CoV-2 exploits the homotrimer transmembrane Spike glycoproteins (S protein) during host cell invasion. The Omicron XBB subvariant, delta, and prototype SARS-CoV-2 receptor-binding domain show similar binding strength to hACE2 (human Angiotensin-Converting Enzyme 2). Here we utilized multiligand virtual screening to identify small molecule inhibitors for their efficacy against SARS-CoV-2 virus using QPLD, pseudovirus ACE2 Inhibition -Time Resolved Forster/Fluorescence energy transfer (TR-FRET) Assay Screening, and Molecular Dynamics simulations (MDS). Three hundred and fifty thousand compounds were screened against the macrodomain of the nonstructural protein 3 of SARS-CoV-2. Using TR-FRET Assay, we filtered out two of 10 compounds that had no reported activity in in vitro screen against Spike S1: ACE2 binding assay. The percentage inhibition at 30 µM was found to be 79% for “Compound F1877-0839” and 69% for “Compound F0470-0003”. This first of its kind study identified “FILLY” pocket in macrodomains. Our 200 ns MDS revealed stable binding poses of both leads. They can be used for further development of preclinical candidates. Full article