Frontiers in Medicine (Oct 2022)
Deep learning to infer visual acuity from optical coherence tomography in diabetic macular edema
- Ting-Yi Lin,
- Hung-Ruei Chen,
- Hsin-Yi Huang,
- Hsin-Yi Huang,
- Yu-Ier Hsiao,
- Zih-Kai Kao,
- Kao-Jung Chang,
- Kao-Jung Chang,
- Tai-Chi Lin,
- Tai-Chi Lin,
- Chang-Hao Yang,
- Chung-Lan Kao,
- Chung-Lan Kao,
- Chung-Lan Kao,
- Chung-Lan Kao,
- Po-Yin Chen,
- Po-Yin Chen,
- Po-Yin Chen,
- Po-Yin Chen,
- Po-Yin Chen,
- Shih-En Huang,
- Shih-En Huang,
- Chih-Chien Hsu,
- Chih-Chien Hsu,
- Chih-Chien Hsu,
- Yu-Bai Chou,
- Yu-Bai Chou,
- Yu-Bai Chou,
- Ying-Chun Jheng,
- Ying-Chun Jheng,
- Ying-Chun Jheng,
- Ying-Chun Jheng,
- Shih-Jen Chen,
- Shih-Jen Chen,
- Shih-Jen Chen,
- Shih-Hwa Chiou,
- Shih-Hwa Chiou,
- Shih-Hwa Chiou,
- Shih-Hwa Chiou,
- De-Kuang Hwang,
- De-Kuang Hwang,
- De-Kuang Hwang
Affiliations
- Ting-Yi Lin
- Doctoral Degree Program of Translational Medicine, National Yang Ming Chiao Tung University and Academia Sinica, Taipei, Taiwan
- Hung-Ruei Chen
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Hsin-Yi Huang
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Hsin-Yi Huang
- Taipei Veterans General Hospital Biostatistics Task Force, Taipei, Taiwan
- Yu-Ier Hsiao
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Zih-Kai Kao
- Department of Physical Medicine and Rehabilitation, Taipei Veterans General Hospital, Taipei, Taiwan
- Kao-Jung Chang
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Kao-Jung Chang
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Tai-Chi Lin
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Tai-Chi Lin
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- Chang-Hao Yang
- Department of Ophthalmology, National Taiwan University, Taipei, Taiwan
- Chung-Lan Kao
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Chung-Lan Kao
- Department of Physical Medicine and Rehabilitation, Taipei Veterans General Hospital, Taipei, Taiwan
- Chung-Lan Kao
- Department of Physical Therapy and Assistive Technology, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Chung-Lan Kao
- Center for Intelligent Drug Systems and Smart Bio-devices (IDS2B), National Yang Ming Chiao Tung University, Hsinchu, Taiwan
- Po-Yin Chen
- Department of Physical Medicine and Rehabilitation, Taipei Veterans General Hospital, Taipei, Taiwan
- Po-Yin Chen
- Department of Physical Therapy and Assistive Technology, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Po-Yin Chen
- 0School of Gerontology and Long-Term Care, College of Nursing, Taipei Medical University, Taipei, Taiwan
- Po-Yin Chen
- 1Master Program in Long-Term Care, College of Nursing, Taipei Medical University, Taipei, Taiwan
- Po-Yin Chen
- 2International Ph.D. Program in Gerontology and Long-Term Care, College of Nursing, Taipei Medical University, Taipei, Taiwan
- Shih-En Huang
- Department of Physical Medicine and Rehabilitation, Taipei Veterans General Hospital, Taipei, Taiwan
- Shih-En Huang
- Department of Physical Therapy and Assistive Technology, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Chih-Chien Hsu
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Chih-Chien Hsu
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Chih-Chien Hsu
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- Yu-Bai Chou
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Yu-Bai Chou
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Yu-Bai Chou
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- Ying-Chun Jheng
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Ying-Chun Jheng
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- Ying-Chun Jheng
- 3Big Data Center, Department of Medical Research, Taipei Veterans General Hospital, Taipei, Taiwan
- Ying-Chun Jheng
- 4Center for Quality Management, Taipei Veterans General Hospital, Taipei, Taiwan
- Shih-Jen Chen
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Shih-Jen Chen
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Shih-Jen Chen
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- Shih-Hwa Chiou
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Shih-Hwa Chiou
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- Shih-Hwa Chiou
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- Shih-Hwa Chiou
- 3Big Data Center, Department of Medical Research, Taipei Veterans General Hospital, Taipei, Taiwan
- De-Kuang Hwang
- School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- De-Kuang Hwang
- Institute of Clinical Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
- De-Kuang Hwang
- Department of Ophthalmology, Taipei Veterans General Hospital, Taipei, Taiwan
- DOI
- https://doi.org/10.3389/fmed.2022.1008950
- Journal volume & issue
-
Vol. 9
Abstract
PurposeDiabetic macular edema (DME) is one of the leading causes of visual impairment in diabetic retinopathy (DR). Physicians rely on optical coherence tomography (OCT) and baseline visual acuity (VA) to tailor therapeutic regimen. However, best-corrected visual acuity (BCVA) from chart-based examinations may not wholly reflect DME status. Chart-based examinations are subjected findings dependent on the patient’s recognition functions and are often confounded by concurrent corneal, lens, retinal, optic nerve, or extraocular disorders. The ability to infer VA from objective optical coherence tomography (OCT) images provides the predicted VA from objective macular structures directly and a better understanding of diabetic macular health. Deviations from chart-based and artificial intelligence (AI) image-based VA will prompt physicians to assess other ocular abnormalities affecting the patients VA and whether pursuing anti-VEGF treatment will likely yield increment in VA.Materials and methodsWe enrolled a retrospective cohort of 251 DME patients from Big Data Center (BDC) of Taipei Veteran General Hospital (TVGH) from February 2011 and August 2019. A total of 3,920 OCT images, labeled as “visually impaired” or “adequate” according to baseline VA, were grouped into training (2,826), validation (779), and testing cohort (315). We applied confusion matrix and receiver operating characteristic (ROC) curve to evaluate the performance.ResultsWe developed an OCT-based convolutional neuronal network (CNN) model that could classify two VA classes by the threshold of 0.50 (decimal notation) with an accuracy of 75.9%, a sensitivity of 78.9%, and an area under the ROC curve of 80.1% on the testing cohort.ConclusionThis study demonstrated the feasibility of inferring VA from routine objective retinal images.Translational relevanceServes as a pilot study to encourage further use of deep learning in deriving functional outcomes and secondary surrogate endpoints for retinal diseases.
Keywords
