Article
Pharmacology & Pharmacy
Thibaud Mathis, Batoul El Ameen, Cristina Vartin, Yasmine Serrar, Frederic Matonti, Aditya Sudhalkar, Alper Bilgic, Amina Rezkallah, Laurent Kodjikian
Summary: This study examined the validity of using anatomical criteria alone to make retreatment decisions in patients with DME, compared to the gold standard that incorporates visual acuity. The OCT-guided strategy yielded equivalent results to the gold standard in 82.7% of cases, with variations in sensitivity and specificity based on the patient's treatment regimen.
Article
Medicine, General & Internal
Yun Bai, Jing Li, Lianjun Shi, Qin Jiang, Biao Yan, Zhenhua Wang
Summary: The study proposes a lightweight model (DME-DeepLabV3+) for extracting diabetic macular edema in OCT images, utilizing DeepLabV3+ architecture with MobileNetV2 backbone, improved ASPP for high-level feature extraction, and a decoder for feature fusion and refinement. The DME-DeepLabV3+ model shows better extraction performance compared to other models, with higher accuracy and F1-score.
FRONTIERS IN MEDICINE
(2023)
Article
Biochemistry & Molecular Biology
Huan-Yu Hsu, Yu-Bai Chou, Ying-Chun Jheng, Zih-Kai Kao, Hsin-Yi Huang, Hung-Ruei Chen, De-Kuang Hwang, Shih-Jen Chen, Shih-Hwa Chiou, Yu-Te Wu
Summary: This study proposes a modified U-net deep learning algorithm to segment fluid and the ellipsoid zone (EZ) from OCT images of patients with diabetic macular edema (DME). The model achieves high performance in segmenting these features and correlates EZ disruption with best corrected visual acuity (BCVA).
Article
Engineering, Electrical & Electronic
Jun Wu, Yaxin Zhang, Zhitao Xiao, Fang Zhang, Lei Geng
Summary: In this paper, an attention mechanism based on residual convolution module U-Net (RCU-Net) is proposed for the automatic segmentation of the retinal layer and cystoid edema lesions in DME. By fusing the residual structure and CBAM for feature extraction, the network can effectively learn different levels of information. Experimental results show that the proposed method achieves high accuracy and MIoU in DME segmentation.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Medicine, General & Internal
Fabao Xu, Shaopeng Liu, Yifan Xiang, Jiaming Hong, Jiawei Wang, Zheyi Shao, Rui Zhang, Wenjuan Zhao, Xuechen Yu, Zhiwen Li, Xueying Yang, Yanshuang Geng, Chunyan Xiao, Min Wei, Weibin Zhai, Ying Zhang, Shaopeng Wang, Jianqiao Li
Summary: This study utilizes generative adversarial network (GAN) to generate and evaluate individualized post-therapeutic optical coherence tomography (OCT) images based on pre-therapeutic images, aiming to predict the short-term response of anti-vascular endothelial growth factor (VEGF) therapy for diabetic macular edema (DME). Results show that the synthetic OCT images generated by the GAN model are comparable to the actual images and can accurately predict edema resorption.
JOURNAL OF CLINICAL MEDICINE
(2022)
Article
Medicine, General & Internal
Xingguo Wang, Yanyan Zhang, Yuhui Ma, William Robert Kwapong, Jianing Ying, Jiayi Lu, Shaodong Ma, Qifeng Yan, Quanyong Yi, Yitian Zhao
Summary: In this study, a new automated framework driven by deep learning was introduced to extract three-dimensional parameters from retinal OCT images for quantification of retinal HRF in DME patients. The results revealed significant differences in the amount and volume of HRF subtypes and showed that the distance between HRF and retinal pigmented epithelium decreased after treatment. The improvement in central macular thickness was positively correlated with the distance from HRF subtypes to the fovea.
FRONTIERS IN MEDICINE
(2023)
Article
Medicine, General & Internal
Haifan Huang, Liangjiu Zhu, Weifang Zhu, Tian Lin, Leonoor Inge Los, Chenpu Yao, Xinjian Chen, Haoyu Chen
Summary: The study developed an algorithm using deep learning technology to detect and quantify HRDs on OCT for DME patients. The algorithm showed stronger correlation and higher ICC with rater 1 compared to inter-rater agreement, providing an objective and repeatable tool for OCT analysis in clinical practice and research.
FRONTIERS IN MEDICINE
(2021)
Article
Medicine, General & Internal
Carolina Arruabarrena, Antonio Rodriguez-Miguel, Fernando de Aragon-Gomez, Purificacion Escamez, Ingrid Rosado, Miguel A. Teus
Summary: The purpose of this study was to establish normative data for macular thickness and volume in a diabetic population. The researchers used spectral-domain optical coherence tomography (SD-OCT) to observe and compare macular thickness and volume between diabetic patients and a reference population. The results showed that diabetic patients had thinner macular thickness and volume compared to the reference population.
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Ophthalmology
Dawei Yang, Zihan Sun, Jian Shi, Anran Ran, Fangyao Tang, Ziqi Tang, Jerry Lok, Simon Szeto, Jason Chan, Fanny Yip, Liang Zhang, Qianli Meng, Martin Rasmussen, Jakob Grauslund, Carol Y. Cheung
Summary: The purpose of this study was to develop and test a deep-learning system for image quality and diabetic macular ischemia (DMI) assessment on optical coherence tomography angiography (OCTA) images. The system achieved high accuracy in both image quality and DMI assessment, indicating its potential for simplified evaluation of DMI on OCTA images.
RETINA-THE JOURNAL OF RETINAL AND VITREOUS DISEASES
(2022)
Article
Medicine, General & Internal
Xue Bai, Rui Hua
Summary: The study found that fluorescein angiography had a higher accuracy in detecting the severity of diabetic retinopathy and macular leakages compared to optical coherence tomography. Central foveal thickness is not a sensitive parameter for detecting latent DMEs.
FRONTIERS IN MEDICINE
(2021)
Review
Engineering, Biomedical
K. C. Pavithra, Preetham Kumar, M. Geetha, Sulatha V. Bhandary
Summary: Diabetic Macular Edema (DME) is a serious complication of Diabetic Retinopathy (DR) and it is the leading cause of vision loss in diabetics. DME is characterized by the accumulation of fluid in the macula due to leaky blood vessels. Advanced imaging techniques such as Color Fundus Photography (CFP) and Optical Coherence Tomography (OCT) can detect the presence of DME at different stages of DR. This review article discusses the latest automated DME detection methods using traditional Machine Learning (ML) and Deep Learning (DL) techniques with retinal fundus or OCT images.
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING
(2023)
Article
Medicine, General & Internal
Huiping Yao, Zijian Yang, Yu Cheng, Xi Shen
Summary: This study investigated the changes in macular status and choroidal thickness following phacoemulsification in patients with mild to moderate nonproliferative diabetic retinopathy. The results showed that the increase in SCP-VD, MT, and CT after surgery was significantly greater in diabetic retinopathy patients compared to the control group.
FRONTIERS IN MEDICINE
(2023)
Article
Medicine, General & Internal
Kai Jin, Yan Yan, Shuai Wang, Ce Yang, Menglu Chen, Xindi Liu, Hiroto Terasaki, Tun-Hang Yeo, Neha Gulab Singh, Yao Wang, Juan Ye
Summary: This study developed a two-stage deep learning system named iERM, which improved the severity grading assessment of epiretinal membranes (ERM) based on optical coherence tomography (OCT) images. The iERM achieved high accuracy scores in internal and external test datasets, comparable to retinal specialists. It has the potential to provide precise guidance for ERM diagnosis and treatment.
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Medicine, General & Internal
Sunjin Hwang, Mincheol Seong, Min Ho Kang, Zheng Xian Thng, Heeyoon Cho, Yong Un Shin
Summary: This study examined the association between bioimpedance profiles and optical coherence tomography features in patients with diabetic macular edema. The results suggest a connection between systemic fluid status, including extracellular water-to-total body water ratio and phase angle, and DME development.
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Endocrinology & Metabolism
Tsung-Cheng Hsieh, Guang-Hong Deng, Yung-Ching Chang, Fang-Ling Chang, Ming-Shan He
Summary: This study aimed to investigate optical coherence tomography (OCT) biomarkers of diabetic macular edema (DME) refractory to anti-vascular endothelial growth factor (anti-VEGF) therapy. The results showed that partially continuous inner segment-outer segment (IS-OS) layers were predictive of better response, while epiretinal membrane (ERM) was a significant predictor of poor response.
FRONTIERS IN ENDOCRINOLOGY
(2023)
Article
Ophthalmology
Simon K. H. Szeto, Vivian W. K. Hui, Fang Yao Tang, Dawei Yang, Zi Han Sun, Shaheeda Mohamed, Carmen K. M. Chan, Timothy Y. Y. Lai, Carol Cheung
Summary: This study aims to investigate whether a combination of baseline and change in SD-OCT biomarkers can predict visual outcomes in patients with DMO treated with anti-VEGF injections. The results indicate that baseline DRIL, hyper-reflective foci in retina, and disruption of ELM/EZ and COST were associated with worse baseline and subsequent VA up to 24 months after treatment. Improvement in DRIL, ELM/EZ, and COST after treatment was associated with greater improvement in VA at 12 months.
BRITISH JOURNAL OF OPHTHALMOLOGY
(2023)
Article
Ophthalmology
Simon K. H. Szeto, Shaheeda Mohamed, Chi Wai Tsang, Carmen K. M. Chan
Summary: This article describes the clinical and optical coherence tomography (OCT) features of two cases with bilateral diffuse retinal infiltrates as the only presenting feature of chronic myeloid leukemia (CML) on initial diagnosis and upon relapse. The findings highlight the important role of ophthalmologists in promptly diagnosing leukemia in patients with visual complaints, and the potential for systemic investigation and hematologist referrals to improve outcomes and preserve vision.
EUROPEAN JOURNAL OF OPHTHALMOLOGY
(2023)
Article
Ophthalmology
Da Wei Yang, Zi Qi Tang, Fang Yao Tang, Simon K. H. Szeto, Jason Chan, Fanny Yip, Cherie Y. K. Wong, An Ran Ran, Timothy Y. Y. Lai, Carol Y. Cheung
Summary: This study investigated the factors associated with diabetic macular ischaemia (DMI) as assessed by optical coherence tomography angiography (OCTA). The results showed that age, visual acuity, ganglion cell-inner plexiform layer thickness, diabetic retinopathy severity, haemoglobin A1c level, estimated glomerular filtration rate, and low-density lipoprotein cholesterol level were associated with SCP-DMI. In addition, the presence of diabetic macular oedema and shorter axial length were associated with DCP-DMI.
BRITISH JOURNAL OF OPHTHALMOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Xi Wang, Fangyao Tang, Hao Chen, Carol Y. Cheung, Pheng-Ann Heng
Summary: Supervised deep learning has achieved success in DME recognition from OCT volumetric images. However, the shortage of labeled data and the expensive annotation make accurate analysis difficult. To tackle this problem, the authors propose a deep semi-supervised multiple instance learning framework that leverages a small amount of labeled data and a large amount of unlabeled data.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Ophthalmology
Shaheeda Mohamed, Carmen K. M. Chan, Chi Wai Tsang, Simon K. H. Szeto, Angie H. C. Fong, Jason C. K. Chan, Cherie Y. K. Wong
Summary: This study reports two cases of bilateral retinal vasculitis in adolescents following COVID-19 vaccination. It suggests a possible association between mRNA vaccinations and the development of autoimmune-related eye diseases. Clinicians should be aware of the potential risk of retinal vasculitis in adolescents after COVID-19 vaccination.
OCULAR IMMUNOLOGY AND INFLAMMATION
(2023)
Review
Medicine, General & Internal
Dawei Yang, An Ran Ran, Truong X. X. Nguyen, Timothy P. H. Lin, Hao Chen, Timothy Y. Y. Lai, Clement C. C. Tham, Carol Y. Y. Cheung
Summary: Optical coherence tomography angiography (OCT-A) provides non-invasive visualization of retinal microvasculature, and deep learning (DL) has been applied in OCT-A image analysis to enhance its clinical values. However, the deployment of this combination in real-world clinics is still in the proof-of-concept stage due to limitations in training sample size, data preprocessing, external dataset testing, and results interpretation standardization. In this review, we introduce the applications of DL in OCT-A, summarize the challenges of clinical deployment, and discuss future research directions.
Article
Ophthalmology
Dawei Yang, Ziqi Tang, Anran Ran, Truong X. X. Nguyen, Simon Szeto, Jason Chan, Cherie Y. K. Wong, Vivian Hui, Ken Tsang, Carmen K. M. Chan, Clement C. C. Tham, Sobha Sivaprasad, Timothy Y. Y. Lai, Carol Y. Y. Cheung
Summary: The presence of diabetic macular ischemia (DMI) on optical coherence tomography angiography (OCTA) images is important for predicting the progression of diabetic retinal disease and visual acuity deterioration. An automated binary DMI algorithm using OCTA images was found to have prognostic value for DR progression, diabetic macular edema development, and visual acuity deterioration.
JAMA OPHTHALMOLOGY
(2023)
Review
Ophthalmology
Yu Huang, Carol Y. Cheung, Dawei Li, Yih Chung Tham, Bin Sheng, Ching Yu Cheng, Ya Xing Wang, Tien Yin Wong
Summary: Cardiovascular disease (CVD) is a major cause of death globally, and the use of artificial intelligence (AI) in analyzing ocular images has the potential to enhance CVD risk prediction. This review provides an overview of AI-based ocular image analysis for predicting CVD and discusses the limitations in AI research and the challenges in clinical practice.
Article
Health Care Sciences & Services
Kang-An Wong, Bryan Chin Hou Ang, Dinesh Visva Gunasekeran, Rahat Husain, Joewee Boon, Krishna Vikneson, Zyna Pei Qi Tan, Gavin Siew Wei Tan, Tien Yin Wong, Rupesh Agrawal
Summary: This study investigates the performance of a novel remote perimetry application designed in a virtual reality metaverse environment to enable functional testing in community-based and primary care settings. The results show that the remote perimetry application has good concordance with the gold standard perimetry and could potentially be used for functional eye screening in out-of-hospital settings.
JOURNAL OF MEDICAL INTERNET RESEARCH
(2023)
Article
Computer Science, Information Systems
Chan Joo Lee, Tyler Hyungtaek Rim, Hyun Goo Kang, Joseph Keunhong Yi, Geunyoung Lee, Marco Yu, Soo-Hyun Park, Jin-Taek Hwang, Yih-Chung Tham, Tien Yin Wong, Ching-Yu Cheng, Dong Wook Kim, Sung Soo Kim, Sungha Park
Summary: This study validated a personalized cardiovascular disease (CVD) risk scoring system, named Reti-CVD, based on retinal images. The risk groups defined by Reti-CVD were significantly associated with increased CVD risk and remained effective in a comprehensive model with other risk factors.
JOURNAL OF THE AMERICAN MEDICAL INFORMATICS ASSOCIATION
(2023)
Article
Oncology
Xi Wang, Yuming Jiang, Hao Chen, Taojun Zhang, Zhen Han, Chuanli Chen, Qingyu Yuan, Wenjun Xiong, Wei Wang, Guoxin Li, Pheng-Ann Heng, Ruijiang Li
Summary: This study presents a multi-modal deep learning radiomics approach to predict the immunotherapy response in gastric cancer patients using both clinical data and computed tomography images. The deep learning model achieved promising performance for predicting immunotherapy response and further improved the accuracy when combined with PD-L1 expression.
RADIOTHERAPY AND ONCOLOGY
(2023)
Article
Urology & Nephrology
Cynthia Ciwei Lim, Crystal Chong, Gavin Tan, Chieh Suai Tan, Carol Y. Cheung, Tien Y. Wong, Ching Yu Cheng, Charumathi Sabanayagam
Summary: This study found that retinal vessel calibre measurements obtained by a deep learning system were significantly associated with incident cardiovascular disease (CVD) in chronic kidney disease (CKD) patients. The addition of kidney function and retinal vessel calibre parameters improved the prediction of CVD risk among CKD patients.
CLINICAL KIDNEY JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Yang Liu, Xiaoqi Wang, Xi Wang, Zhen Wang, Jurgen Kurths
Summary: This article studies the diffusion-source-inference (DSI) problem and proposes a percolation-based evolutionary framework (PrEF) to optimize the observer set and minimize the candidate set. The effectiveness of the proposed method is validated on both synthetic and empirical networks in varied circumstances and shows better performance compared to the state of the art. This research provides a framework for the analysis of the DSI problem in large-scale networks.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)