Validation and Clinical Applicability of Whole-Volume Automated Segmentation of Optical Coherence Tomography in Retinal Disease Using Deep Learning

Abstract

IMPORTANCE Quantitative volumetric measures of retinal disease in optical coherence tomography (OCT) scans are infeasible to perform owing to the time required for manual grading. Expert-level deep learning systems for automatic OCT segmentation have recently been developed. However, the potential clinical applicability of these systems is largely unknown.OBJECTIVE To evaluate a deep learning model for whole-volume segmentation of 4 clinically important pathological features and assess clinical applicability. DESIGN, SETTING, PARTICIPANTSThis diagnostic study used OCT data from 173 patients with a total of 15 558 B-scans, treated at Moorfields Eye Hospital. The data set included 2 common OCT devices and 2 macular conditions: wet age-related macular degeneration (107 scans) and diabetic macular edema (66 scans), covering the full range of severity, and from 3 points during treatment. Two expert graders performed pixel-level segmentations of intraretinal fluid, subretinal fluid, subretinal hyperreflective material, and pigment epithelial detachment, including all B-scans in each OCT volume, taking as long as 50 hours per scan. Quantitative evaluation of whole-volume model segmentations was performed. Qualitative evaluation of clinical applicability by 3 retinal experts was also conducted.

Validation of an Automated Fluid Algorithm on Real-World Data of Neovascular Age-Related Macular Degeneration Over Five Years

“…Comparison between different reported fluid segmentation algorithms is difficult because they are not publicly available and have not been applied on the same imaging data sets. In particular, the algorithms developed by DeepMind 9,25 and the Notal OCT Analyzer 26 are the most closely related to our Vienna Fluid Monitor. For reported volumetric measurements, a comparison of our algorithm with the DeepMind algorithm is reported above and with Notal OCT Analyzer was part of the analysis reported by Keenan et al 27 Ideally, the algorithms should be compared directly on the same open data set, like the one from the RETOUCH challenge 28 on fluid detection and quantification.…”

Quantitative assessment of retinal fluid in neovascular age-related macular degeneration under anti-VEGF therapy

“…The discrepancies between readers and experts might be tremendously higher when total macular fluid is assessed and expert opinion might differ when annotating total volume scans. 57 Especially, IRF when present in only small amounts might be missed by the human investigator. 7 Nonetheless, IRF is associated with worse visual outcomes, whereas the decision to treat SRF in the long term is still not fully solved.…”

“…An overview of the presence of SRF, IRF, or PED can be quickly performed, whereas manual quantification is unfeasible in clinical routine. 8 , 57 However, patient satisfaction and functional outcomes are not only based on the presence of fluid. The amount of fluid in the central millimeter is highly correlated with visual function, 28 , 40 but subclinical markers might be as important as exudative fluid.…”

Future of artificial intelligence for the diagnosis and treatment of retinal diseases

“…Возможности структурной ОКТ в диагностике возрастной макулярной дегенерации (ВМД) включают визуализацию и оценку размера и типа макулярных друз, выявление зон атрофии, гиперплазии и миграции клеток ретинального пигментного эпителия (РПЭ), отслойки РПЭ и нейроэпителия, интра-и субретинальной жидкости, субретинального гиперрефлективного материала, оценку типа хориоидальной неоваскуляризации и признаков ее активности [20]. В последние годы были опубликованы результаты исследований, посвященных разработке программ ИИ для автоматического распознавания макулярных друз [21], географической атрофии [22], морфологических признаков хориоидальной неоваскуляризации (ХНВ) [23].…”