Automated Identification of Incomplete and Complete Retinal Epithelial Pigment and Outer Retinal Atrophy Using Machine Learning.

OBJECTIVE: To assess and validate a deep learning algorithm to automatically detect incomplete retinal pigment epithelial and outer retinal atrophy (iRORA) and complete retinal pigment epithelial and outer retinal atrophy (cRORA) in eyes with age-related macular degeneration. DESIGN: In a retrospective machine learning analysis, a deep learning model was trained to jointly classify the presence of iRORA and cRORA within a given B-scan. The algorithm was evaluated using 2 separate and independent datasets. PARTICIPANTS: OCT B-scan volumes from 71 patients with nonneovascular age-related macular degeneration captured at the Doheny-University of California Los Angeles Eye Centers and the following 2 external OCT B-scans testing datasets: (1) University of Pennsylvania, University of Miami, and Case Western Reserve University and (2) Doheny Image Reading Research Laboratory. METHODS: The images were annotated by an experienced grader for the presence of iRORA and cRORA. A Resnet18 model was trained to classify these annotations for each B-scan using OCT volumes collected at the Doheny-University of California Los Angeles Eye Centers. The model was applied to 2 testing datasets to assess out-of-sample model performance. MAIN OUTCOMES MEASURES: Model performance was quantified in terms of area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC). Sensitivity, specificity, and positive predictive value were also compared against additional clinician annotators. RESULTS: On an independently collected test set, consisting of 1117 volumes from the general population, the model predicted iRORA and cRORA presence within the entire volume with nearly perfect AUROC performance and AUPRC scores (iRORA, 0.61; 95% confidence interval [CI] [0.45, 0.82]: cRORA, 0.83; 95% CI [0.68, 0.95]). On another independently collected set, consisting of 60 OCT B-scans enriched for iRORA and cRORA lesions, the model performed with AUROC (iRORA: 0.68, 95% CI [0.54, 0.81]; cRORA: 0.84, 95% CI [0.75, 0.94]) and AUPRC (iRORA: 0.70, 95% CI [0.55, 0.86]; cRORA: 0.82, 95% CI [0.70, 0.93]). CONCLUSIONS: A deep learning model can accurately and precisely identify both iRORA and cRORA lesions within the OCT B-scan volume. The model can achieve similar sensitivity compared with human graders, which potentially obviates a laborious and time-consuming annotation process and could be developed into a diagnostic screening tool.

Comparison between B-Scan and En Face Images for Incomplete and Complete Retinal Pigment Epithelium and Outer Retinal Atrophy.

PURPOSE: To evaluate and compare the detection of incomplete retinal pigment epithelium and outer retinal atrophy (iRORA) and complete retinal pigment epithelium and outer retinal atrophy (cRORA) assessed on OCT B-scans versus persistent choroidal hypertransmission defects (hyperTDs) assessed by en face choroidal OCT images. DESIGN: Retrospective, cross-sectional study. PARTICIPANTS: Patients with late atrophic age-related macular degeneration imaged on the same day using both Spectralis OCT and Cirrus OCT. MAIN OUTCOME MEASURE: Agreement between the B-scan and en face OCT for the detection of hyperTDs, cRORA, and iRORA. METHODS: Two independent graders examined en face OCT and structural OCT to determine the presence and location of hyperTDs, iRORA, and cRORA. RESULTS: A total of 239 iRORA and cRORA lesions were detected on the B-scans, and 249 hyperTD lesions were identified on the en face OCT images. There was no significant difference (P = 0.88) in the number of lesions. There was no significant difference in the 134 cRORA lesions identified on B-scans and the 131 hyperTDs detected on en face OCT images (P = 0.13). A total of 105 iRORA lesions were identified by B-scan assessment; however, 50 of these iRORA lesions met the criteria for persistent hyperTDs on en face OCT images (P < 0.001). When considering the topographic correspondence between B-scan and en face OCT detected lesions, the mean percentage of agreement between B-scan detection of cRORA lesions with en face OCT detection was 97.6 % (P = 0.13). CONCLUSIONS: We observed high overall agreement between cRORA lesions identified on B-scans and persistent hyperTDs identified on en face OCT. However, en face imaging was able to detect iRORA lesions that had a greatest linear dimension ≥ 250 µm in a nonhorizontal en face dimension. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.