Personalized Medicine in Ophthalmic Diseases: Challenges and Opportunities

A special issue of Journal of Personalized Medicine (ISSN 2075-4426). This special issue belongs to the section "Methodology, Drug and Device Discovery".

Deadline for manuscript submissions: closed (10 May 2023) | Viewed by 19827

Printed Edition Available!
A printed edition of this Special Issue is available here.

Special Issue Editors

Department of Ophthalmology, the Second Affiliated Hospital of Zhejiang University, School of Medicine, Hangzhou, China
Interests: retina; diabetic retinopathy; macular degeneration; imaging; artificial intelligence
Special Issues, Collections and Topics in MDPI journals
Department of Ophthalmology, Peking University Third Hospital, Beijing, China
Interests: glaucoma; optic nerve atrophy; laser trabeculoplasty; filtering surgery; MIGS

Special Issue Information

Dear Colleagues,

Personalized medicine is a broadly used term encompassing approaches for tailoring healthcare to the needs of individual patients. It was adopted early in ophthalmology and was mainly achieved through disease stratification and individualization. Therefore, diagnostic techniques capable of realizing comprehensive individual assessments are very important. Previous studies put forward techniques such as next-generation sequencing and translational research. Gene-therapy-based treatment trials have been presented for ophthalmic diseases, such as retinitis pigmentosa and age-related macular degeneration. Recently, with the rapid development of artificial intelligence and interdisciplinary collaboration, concepts such as machine learning and wearable devices have been frequently raised in ophthalmic research, potentially proving to be promising novel methods for the realization of personalized ophthalmology. In this Special Issue, we would like to explore "Personalized Medicine in Ophthalmic Diseases: Challenges and Opportunities", which often involve methods with the aim of achieving personalized diagnoses and treatments of ophthalmic diseases. Studies following the above criteria with potential impacts on personalized medicine in ophthalmic diseases will be considered for publication in this Special Issue of the Journal of Personalized Medicine (IF: 3.508, ISSN 2075-4426). We appreciate your contributions to the field and look forward to hearing from you.

Dr. Kai Jin
Dr. Chun Zhang
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Journal of Personalized Medicine is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • ocular diseases
  • visual science
  • clinical research
  • novel diagnostic techniques
  • novel treatment methods
  • translational research
  • medical imaging
  • artificial intelligence
  • machine learning
  • wearable devices

Published Papers (11 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Editorial

Jump to: Research

2 pages, 172 KiB  
Editorial
Personalized Medicine in Ophthalmic Diseases: Challenges and Opportunities
by Kai Jin and Chun Zhang
J. Pers. Med. 2023, 13(6), 893; https://doi.org/10.3390/jpm13060893 - 25 May 2023
Cited by 1 | Viewed by 954
Abstract
Personalized medicine is a broadly used term to encompass approaches used to tailor healthcare to the needs of individual patients [...] Full article

Research

Jump to: Editorial

12 pages, 8672 KiB  
Article
Analysis of Clinical Characteristics of Patients with Recurrent Cytomegalovirus Retinitis after Hematopoietic Stem Cell Transplantation
by Xiaona Wang, Yao Lu, Haiping Li, Zhizhong Ma, Jing Hong and Changguan Wang
J. Pers. Med. 2023, 13(4), 639; https://doi.org/10.3390/jpm13040639 - 7 Apr 2023
Viewed by 1270
Abstract
Objective: To analyze and summarize the clinical and imaging characteristics of patients with cytomegalovirus retinitis (CMVR) relapse after hematopoietic stem cell transplantation (HSCT). Methods: This retrospective case series study recruited patients with CMVR after HSCT. The study compared the patients with stable lesions [...] Read more.
Objective: To analyze and summarize the clinical and imaging characteristics of patients with cytomegalovirus retinitis (CMVR) relapse after hematopoietic stem cell transplantation (HSCT). Methods: This retrospective case series study recruited patients with CMVR after HSCT. The study compared the patients with stable lesions and CMV-negative aqueous humor after treatment with those with relapse lesions and a CMV DNA load in aqueous humor which had increased again after treatment. The observation indexes were basic clinical information, best-corrected visual acuity, wide-angle fundus photography, optical coherence tomography (OCT), blood CD4+ T lymphocyte count, and aqueous humor CMV load of the patients. We summarized the data and statistically analyzed the differences between the relapse and non-relapse groups, as well as the correlations of the observed indicators. Results: The study recruited 52 patients with CMVR (82 eyes) after HSCT, of whom 11 patients (15 eyes) had recurrence after treatment (21.2%). The recurrence interval was 6.4 ± 4.9 months. The final best-corrected visual acuity of recurrent patients was 0.3 ± 0.3. The number of CD4+ T lymphocytes in recurrence patients at the time of onset was 126.7 ± 80.2/mm3. The median CMV DNA load detected in aqueous humor at the time of recurrence was 8.63 × 103 copies/mL. There was a significant difference in the CD4+ T lymphocyte count between the recurrence and the non-recurrence groups at onset. The onset of visual acuity in recurrence patients was significantly correlated with final visual acuity and recurrence lesion area. The fundus of recurred CMVR showed increased marginal activity of the original stable lesion. Concurrently, yellow-white new lesions appeared around the stable, atrophic, and necrotic lesions. OCT showed new diffuse hyperreflexic lesions in the retinal neuroepithelial layer near the old lesions. Inflammatory punctate hyperreflexes were observed in the vitreous, with vitreous liquefaction and contraction. Conclusion: This study suggests that the clinical features, fundus manifestations, and imaging features of CMVR recurrence after HSCT are different from those at the initial onset. Patients should be closely followed up after their condition is stable to be alert for CMVR recurrence. Full article
Show Figures

Figure 1

11 pages, 3192 KiB  
Article
Automatic Diagnosis of Infectious Keratitis Based on Slit Lamp Images Analysis
by Shaodan Hu, Yiming Sun, Jinhao Li, Peifang Xu, Mingyu Xu, Yifan Zhou, Yaqi Wang, Shuai Wang and Juan Ye
J. Pers. Med. 2023, 13(3), 519; https://doi.org/10.3390/jpm13030519 - 13 Mar 2023
Cited by 2 | Viewed by 3373
Abstract
Infectious keratitis (IK) is a common ophthalmic emergency that requires prompt and accurate treatment. This study aimed to propose a deep learning (DL) system based on slit lamp images to automatically screen and diagnose infectious keratitis. This study established a dataset of 2757 [...] Read more.
Infectious keratitis (IK) is a common ophthalmic emergency that requires prompt and accurate treatment. This study aimed to propose a deep learning (DL) system based on slit lamp images to automatically screen and diagnose infectious keratitis. This study established a dataset of 2757 slit lamp images from 744 patients, including normal cornea, viral keratitis (VK), fungal keratitis (FK), and bacterial keratitis (BK). Six different DL algorithms were developed and evaluated for the classification of infectious keratitis. Among all the models, the EffecientNetV2-M showed the best classification ability, with an accuracy of 0.735, a recall of 0.680, and a specificity of 0.904, which was also superior to two ophthalmologists. The area under the receiver operating characteristics curve (AUC) of the EffecientNetV2-M was 0.85; correspondingly, 1.00 for normal cornea, 0.87 for VK, 0.87 for FK, and 0.64 for BK. The findings suggested that the proposed DL system could perform well in the classification of normal corneas and different types of infectious keratitis, based on slit lamp images. This study proves the potential of the DL model to help ophthalmologists to identify infectious keratitis and improve the accuracy and efficiency of diagnosis. Full article
Show Figures

Figure 1

16 pages, 1954 KiB  
Article
Using the C-Read as a Portable Device to Evaluate Reading Ability in Young Chinese Adults: An Observational Study
by Tian Cheng, Taikang Yao, Boxuan Xu, Wanwei Dai, Xuejiao Qin, Juan Ye, Lingge Suo and Chun Zhang
J. Pers. Med. 2023, 13(3), 463; https://doi.org/10.3390/jpm13030463 - 1 Mar 2023
Viewed by 1434
Abstract
We evaluated the reading characteristics of normal-sighted young adults using C-Read to provide baseline healthy population values. We also investigated the relationship between the National Eye Institute’s Visual Functioning Questionnaire (VFQ-25) score and reading ability, myopia, and hours of screen use, focusing on [...] Read more.
We evaluated the reading characteristics of normal-sighted young adults using C-Read to provide baseline healthy population values. We also investigated the relationship between the National Eye Institute’s Visual Functioning Questionnaire (VFQ-25) score and reading ability, myopia, and hours of screen use, focusing on the extent to which these factors affect participants’ visual function and, ultimately, their vision-related quality of life (QoL). Overall, 207 young, healthy participants (414 eyes) aged 18–35 years were tested for reading speed using C-Read connected to a smartphone-based application between December 2022 and January 2023. Each participant received a VFQ-25 questionnaire to evaluate vision-related QoL. Data on daily e-screen usage hours were collected. Among the participants, 91 (44.0%) were women; their mean (SD) age was 22.45 (4.01) years. The mean (SD) reading acuity (RA) was 0.242 (0.124), 0.249 (0.120), and 0.193 (0.104) logarithmic minimum angle of resolution (logMAR) for the right, left, and both eyes, respectively. The mean (SD) maximum reading speed (MRS) was 171.65 (46.27), 168.59 (45.68), and 185.16 (44.93) words per minute (wpm) with the right, left, and both eyes, respectively. The mean (SD) critical print size (CPS) was 0.412 (0.647), 0.371 (0.229), and 0.419 (1.05) logMAR per the right, left, and both eyes, respectively. The RA and CPS were significantly different between sexes (p = 0.002 and p = 0.001). MRS was significantly different between the education level (p = 0.005) and myopia level groups (p = 0.010); however, it was not clear whether this difference was confounded by age. The myopic power in diopters significantly affected RA (coefficient, −0.012; 95% CI, −0.018 to −0.006; p = 0.001); screen time significantly affected MRS (coefficient, 0.019; 95% CI, 0.57 to 6.33; p = 0.019). RA (coefficient, −21.41; 95% CI, −33.74 to −9.08; p = 0.001) and duration of screen use (coefficient, -0.86; 95% CI, −1.29 to −0.43; p < 0.001) independently had a significantly negative correlation with VFQ-25 scores. Our findings provide a baseline value for C-Read in normal-sighted young adults. Refractive status significantly affected RA, while screen time significantly affected MRS. Interventions aimed at enhancing RA may have the potential to maximize vision-related QoL and enable older adults with impaired vision to achieve greater outcomes. Future, larger-scale, C-Read experiments will help provide newer, more optimal methods for the early diagnosis of visual impairment. Full article
Show Figures

Graphical abstract

12 pages, 2307 KiB  
Article
A Novel PAX6 Frameshift Mutation Identified in a Large Chinese Family with Congenital Aniridia
by Chenghu Wang, Weihua Yang, Xiumiao Li, Chenchen Zhou, Jinghua Liu, Ling Jin, Qin Jiang and Yun Wang
J. Pers. Med. 2023, 13(3), 442; https://doi.org/10.3390/jpm13030442 - 28 Feb 2023
Viewed by 1284
Abstract
Congenital aniridia is a rare autosomal dominant congenital ocular disorder. Genetic studies suggest that heterozygous mutations in the developmental regulator PAX6 gene or the related regulatory regions leading to haploinsufficiency are the main cause of congenital aniridia. In this study, the clinical characteristics [...] Read more.
Congenital aniridia is a rare autosomal dominant congenital ocular disorder. Genetic studies suggest that heterozygous mutations in the developmental regulator PAX6 gene or the related regulatory regions leading to haploinsufficiency are the main cause of congenital aniridia. In this study, the clinical characteristics and pathogenic mutation of a four-generation Chinese family with congenital aniridia were investigated. All members recruited in this study underwent comprehensive ophthalmic examinations. Targeted gene capture sequencing and Sanger sequencing were performed to screen and confirm the candidate pathogenicity gene and its mutation. A multiple alignment of homologous sequences covering the identified mutation from different species was investigated, and the mutant protein structure was predicted using Swiss-Model. Additionally, the prediction of pathogenicity was analyzed using the ACMG Guidelines. Thirteen patients in this pedigree were diagnosed with congenital aniridia. A novel heterozygous frameshift mutation (c.391_398dupATACCAAG, p.Ser133Argfs*8) in exon 7 of the PAX6 gene was identified in all affected individuals in the family. This study demonstrates that this frameshift mutation of the PAX6 gene might be the causative genetic defect of congenital aniridia in this family. This mutation is predicted to cause the premature truncation of the PAX6 protein, leading to the functional haploinsufficiency of PAX6, which may be the major molecular mechanism underlying the aniridia phenotype. To the best of our knowledge, this is the first report of a novel pathogenic PAX6 gene variant c.391_398dupATACCAAG(p.Ser133Argfs*8) identified in a Chinese family with congenital aniridia. Full article
Show Figures

Figure 1

10 pages, 1520 KiB  
Article
A Modified Technique for Preventing Lens–Iris Diaphragm Retropulsion Syndrome in Vitrectomized Eyes during Phacoemulsification
by Zhiyi Wu, Tian He, Zhitao Su, Ye Liu, Jingliang He and Yanan Huo
J. Pers. Med. 2023, 13(3), 418; https://doi.org/10.3390/jpm13030418 - 26 Feb 2023
Viewed by 1887
Abstract
Background: Lens–iris diaphragm retropulsion syndrome (LIDRS) is common in vitrectomized or high myopic eyes during phacoemulsification. We evaluated the results of a modified technique for cataract treatment using phacoemulsification in vitrectomized eyes. Methods: In this retrospective study, we enrolled thirty-four vitrectomized eyes treated [...] Read more.
Background: Lens–iris diaphragm retropulsion syndrome (LIDRS) is common in vitrectomized or high myopic eyes during phacoemulsification. We evaluated the results of a modified technique for cataract treatment using phacoemulsification in vitrectomized eyes. Methods: In this retrospective study, we enrolled thirty-four vitrectomized eyes treated with modified phacoemulsification (Modified Group) and nineteen vitrectomized eyes treated with routine phacoemulsification (Control Group). The modified technique comprised irrigation with a balanced salt solution underneath the pupil before phacoemulsification instrument entry, lens implantation and stromal hydration to stabilize the anterior chamber and equilibrate the pressure between the anterior chamber and posterior cavity. Results: We compared the incidences of intra and postoperative complications and visual outcomes between modified and routine phacoemulsification. Pain, LIDRS and difficulty in stromal hydration were significantly more common in the Control Group than in the Modified Group (p < 0.05). There were no significant differences in the rates of posterior capsular rupture, iris trauma, lens dislocation, or posterior capsular opacification between the Modified and Control Groups (p > 0.05). However, there was no significant difference in visual acuity between the groups (p > 0.05). Complications such as loss of nuclear fragments into the vitreous cavity, cystoid macular edema, retina redetachment, suprachoroidal hemorrhage and vitreous hemorrhage did not occur either intra or postoperatively in any of our patients. Conclusions: Our modified technique prevents LIDRS and complications arising during cataract surgery in vitrectomized eyes. Aside from this, the results of modified and routine phacoemulsification are similar in vitrectomized eyes. Full article
Show Figures

Figure 1

9 pages, 546 KiB  
Article
Pneumatic Retinopexy Involving the Use of Intravitreal Air Injection and Laser Photocoagulation for Rhegmatogenous Retinal Detachment in Phakic Eyes
by Tiepei Zhu, Zhenyang Xiang, Qinzhu Huang, Gaochun Li, Shenchao Guo and Enhui Li
J. Pers. Med. 2023, 13(2), 328; https://doi.org/10.3390/jpm13020328 - 14 Feb 2023
Viewed by 2014
Abstract
The clinical efficacy of pneumatic retinopexy (PR) using intravitreal pure air injection and laser photocoagulation for rhegmatogenous retinal detachment (RRD) remains unknown. Thirty-nine consecutive patients with RRD (39 eyes) were included in this prospective case series. All patients underwent two-step PR surgery containing [...] Read more.
The clinical efficacy of pneumatic retinopexy (PR) using intravitreal pure air injection and laser photocoagulation for rhegmatogenous retinal detachment (RRD) remains unknown. Thirty-nine consecutive patients with RRD (39 eyes) were included in this prospective case series. All patients underwent two-step PR surgery containing pure air intravitreal injection and laser photocoagulation retinopexy during hospitalization. The main outcomes were best-corrected visual acuity (BCVA) and primary anatomic success rates after PR treatment. The mean follow-up was 18.3 ± 9.7 months, ranging from 6 to 37 months. The primary anatomic success rate was 89.7% (35/39) after PR treatment. Final reattachment of the retina was achieved in 100% of cases. Macular epiretinal membrane was developed in two patients (5.7%) among successful PR cases during the follow-up. The mean logMAR BCVA value was significantly improved from 0.94 ± 0.69 before surgery to 0.39 ± 0.41 after surgery. The average central retinal thickness was significantly thinner in the RRD eyes of macula-off patients (206.8 ± 56.13 μm) when compared with the fellow eyes (234.6 ± 48.4 μm) at the last follow-up (p = 0.005). This study concluded that an inpatient PR procedure with pure air injection and laser photocoagulation is a safe and effective approach to treating patients with RRD, who may achieve a high single-operation success rate and good visual acuity recovery. Full article
Show Figures

Figure 1

11 pages, 2025 KiB  
Article
Analysis of Retinal Microstructure in Eyes with Dissociated Optic Nerve Fiber Layer (DONFL) Appearance following Idiopathic Macular Hole Surgery: An Optical Coherence Tomography Study
by Shucheng He, Xin Ye, Wangli Qiu, Shangchao Yang, Xiaxing Zhong, Yiqi Chen, Rui He and Lijun Shen
J. Pers. Med. 2023, 13(2), 255; https://doi.org/10.3390/jpm13020255 - 30 Jan 2023
Cited by 1 | Viewed by 1426
Abstract
(1) Purpose: This study aimed to evaluate morphological changes of the retina in eyes with dissociated optic nerve fiber layer (DONFL) appearance following internal limiting membrane (ILM) peeling for full-thickness idiopathic macular hole (IMH) on spectral-domain optical coherence tomography (SD-OCT). (2) Methods: We [...] Read more.
(1) Purpose: This study aimed to evaluate morphological changes of the retina in eyes with dissociated optic nerve fiber layer (DONFL) appearance following internal limiting membrane (ILM) peeling for full-thickness idiopathic macular hole (IMH) on spectral-domain optical coherence tomography (SD-OCT). (2) Methods: We retrospectively analyzed 39 eyes of 39 patients with type 1 macular hole closure after a vitrectomy with ILM peeling procedure at a six-month minimum postoperative follow-up. The retinal thickness maps and cross-sectional OCT images were obtained from a clinical OCT device. The cross-sectional area of the retinal nerve fiber layer (RNFL) on cross-sectional OCT images was manually measured by ImageJ software. (3) Results: The inner retinal layers (IRLs) thickness thinned down much more in the temporal quadrant than in nasal quadrants at 2 and 6 months postoperatively (p < 0.001). However, the cross-sectional area of the RNFL did not change significantly at 2 and 6 months postoperatively (p > 0.05) when compared to preoperative data. In addition, the thinning of the IRL did not correlate with the best-corrected visual acuity (BCVA) at 6 months postoperatively. (4) Conclusions: The thickness of the IRL decreased in eyes with a DONFL appearance after ILM peeling for IMH. The thickness of the IRL decreased more in the temporal retina than in the nasal retina, but the change did not affect BCVA during the 6 months after surgery. Full article
Show Figures

Figure 1

12 pages, 1492 KiB  
Article
Could Mid- to Late-Onset Glaucoma Be Associated with an Increased Risk of Incident Dementia? A Nationwide Retrospective Cohort Study
by Dong-Kyu Kim and So Yeon Lee
J. Pers. Med. 2023, 13(2), 214; https://doi.org/10.3390/jpm13020214 - 26 Jan 2023
Cited by 3 | Viewed by 1311
Abstract
This study aimed to investigate the possible link between glaucoma and dementia using a nationwide cohort sample of data. The glaucoma group (n = 875) included patients diagnosed between 2003 and 2005, aged over 55 years; the comparison group was selected using [...] Read more.
This study aimed to investigate the possible link between glaucoma and dementia using a nationwide cohort sample of data. The glaucoma group (n = 875) included patients diagnosed between 2003 and 2005, aged over 55 years; the comparison group was selected using propensity score matching (n = 3500). The incidence of all-cause dementia was 18.67 (7014.7 person-years) among those with glaucoma aged over 55 years. The glaucoma group developed all-cause dementia more frequently than those in the comparison group (adjusted hazard ratio (HR) = 1.43, 95% confidence interval (CI), 1.17–1.74). In a subgroup analysis, primary open-angle glaucoma (POAG) showed a significantly increased adjusted HR for all-cause dementia events (1.52, 95% CI: 1.23–1.89), whereas we could not find any significant association in patients with primary angle-closure glaucoma (PACG). Additionally, POAG patients showed an increased risk of the development of Alzheimer’s disease (adjusted HR = 1.57, 95% CI, 1.21–2.04) and Parkinson’s disease (adjusted HR = 2.29, 95% CI, 1.46–3.61), but there was no significant difference in PACG patients. Moreover, the risk of Alzheimer’s disease and Parkinson’s disease was higher within 2 years of POAG diagnosis. Although our findings have some limitations, such as confounding factor bias, we suggest that clinicians should pay attention to the early detection of dementia in patients with POAG. Full article
Show Figures

Figure 1

13 pages, 2423 KiB  
Article
End-to-End Deep-Learning-Based Diagnosis of Benign and Malignant Orbital Tumors on Computed Tomography Images
by Ji Shao, Jiazhu Zhu, Kai Jin, Xiaojun Guan, Tianming Jian, Ying Xue, Changjun Wang, Xiaojun Xu, Fengyuan Sun, Ke Si, Wei Gong and Juan Ye
J. Pers. Med. 2023, 13(2), 204; https://doi.org/10.3390/jpm13020204 - 23 Jan 2023
Cited by 6 | Viewed by 2310
Abstract
Determining the nature of orbital tumors is challenging for current imaging interpretation methods, which hinders timely treatment. This study aimed to propose an end-to-end deep learning system to automatically diagnose orbital tumors. A multi-center dataset of 602 non-contrast-enhanced computed tomography (CT) images were [...] Read more.
Determining the nature of orbital tumors is challenging for current imaging interpretation methods, which hinders timely treatment. This study aimed to propose an end-to-end deep learning system to automatically diagnose orbital tumors. A multi-center dataset of 602 non-contrast-enhanced computed tomography (CT) images were prepared. After image annotation and preprocessing, the CT images were used to train and test the deep learning (DL) model for the following two stages: orbital tumor segmentation and classification. The performance on the testing set was compared with the assessment of three ophthalmologists. For tumor segmentation, the model achieved a satisfactory performance, with an average dice similarity coefficient of 0.89. The classification model had an accuracy of 86.96%, a sensitivity of 80.00%, and a specificity of 94.12%. The area under the receiver operating characteristics curve (AUC) of the 10-fold cross-validation ranged from 0.8439 to 0.9546. There was no significant difference on diagnostic performance of the DL-based system and three ophthalmologists (p > 0.05). The proposed end-to-end deep learning system could deliver accurate segmentation and diagnosis of orbital tumors based on noninvasive CT images. Its effectiveness and independence from human interaction allow the potential for tumor screening in the orbit and other parts of the body. Full article
Show Figures

Figure 1

16 pages, 1508 KiB  
Article
MAC-ResNet: Knowledge Distillation Based Lightweight Multiscale-Attention-Crop-ResNet for Eyelid Tumors Detection and Classification
by Xingru Huang, Chunlei Yao, Feng Xu, Lingxiao Chen, Huaqiong Wang, Xiaodiao Chen, Juan Ye and Yaqi Wang
J. Pers. Med. 2023, 13(1), 89; https://doi.org/10.3390/jpm13010089 - 29 Dec 2022
Cited by 2 | Viewed by 1831
Abstract
Eyelid tumors are tumors that occur in the eye and its appendages, affecting vision and appearance, causing blindness and disability, and some having a high lethality rate. Pathological images of eyelid tumors are characterized by large pixels, multiple scales, and similar features. Solving [...] Read more.
Eyelid tumors are tumors that occur in the eye and its appendages, affecting vision and appearance, causing blindness and disability, and some having a high lethality rate. Pathological images of eyelid tumors are characterized by large pixels, multiple scales, and similar features. Solving the problem of difficult and time-consuming fine-grained classification of pathological images is important to improve the efficiency and quality of pathological diagnosis. The morphology of Basal Cell Carcinoma (BCC), Meibomian Gland Carcinoma (MGC), and Cutaneous Melanoma (CM) in eyelid tumors are very similar, and it is easy to be misdiagnosed among each category. In addition, the diseased area, which is decisive for the diagnosis of the disease, usually occupies only a relatively minor portion of the entire pathology section, and screening the area of interest is a tedious and time-consuming task. In this paper, deep learning techniques to investigate the pathological images of eyelid tumors. Inspired by the knowledge distillation process, we propose the Multiscale-Attention-Crop-ResNet (MAC-ResNet) network model to achieve the automatic classification of three malignant tumors and the automatic localization of whole slide imaging (WSI) lesion regions using U-Net. The final accuracy rates of the three classification problems of eyelid tumors on MAC-ResNet were 96.8%, 94.6%, and 90.8%, respectively. Full article
Show Figures

Graphical abstract

Back to TopTop