Cone Identification in Choroideremia: Repeatability, Reliability, and Automation Through Use of a Convolutional Neural Network

Abstract

Adaptive optics imaging has enabled the visualization of photoreceptors both in health and disease. However, there remains a need for automated accurate cone photoreceptor identification in images of disease. Here, we apply an open-source convolutional neural network (CNN) to automatically identify cones in images of choroideremia (CHM). We further compare the results to the repeatability and reliability of manual cone identifications in CHM. Methods: We used split-detection adaptive optics scanning laser ophthalmoscopy to image the inner segment cone mosaic of 17 patients with CHM. Cones were manually identified twice by one experienced grader and once by two additional experienced graders in 204 regions of interest (ROIs). An open-source CNN either pre-trained on normal images or trained on CHM images automatically identified cones in the ROIs. True and false positive rates and Dice's coefficient were used to determine the agreement in cone locations between data sets. Interclass correlation coefficient was used to assess agreement in bound cone density. Results: Intra-and intergrader agreement for cone density is high in CHM. CNN performance increased when it was trained on CHM images in comparison to normal, but had lower agreement than manual grading. Conclusions: Manual cone identifications and cone density measurements are repeatable and reliable for images of CHM. CNNs show promise for automated cone selections, although additional improvements are needed to equal the accuracy of manual measurements. Translational Relevance: These results are important for designing and interpreting longitudinal studies of cone mosaic metrics in disease progression or treatment intervention in CHM.

Cone Photoreceptor Morphology in Choroideremia Assessed Using Non-Confocal Split-Detection Adaptive Optics Scanning Light Ophthalmoscopy

“…In the present study, our results from the centrally retained retinal area showed significantly lower cone density in CHM subjects compared to normal-sighted controls at 1°, 2°, and 4° temporal from the fovea, as well as larger intercone spaces. These findings are consistent with other recent publications on parafoveal cone density in CHM which have found decreased cone density and heterogeneous cell morphology 8 , 12 , 21 , 22 , 36 Abnormal cone coverage and increased spacing in CHM compared to controls have also been reported in previous studies. 21 , 50 , 51…”

“…Intergrader agreement for cone identifications in CHM is high; however, differences do occur. 22 , 34 When reviewer PX disagreed on a cone selection made by JIWM (either a false positive or a false negative), the segmentation was performed according to PX's discretion, with the location of the cone recorded. Investigators JIWM and PX then reviewed jointly all segmentations to come to agreement.…”

“… 8 Within the retained central islands, studies have found local regions of normal and reduced cone photoreceptor density and abnormal cone reflectivity. 8 , 12 , 21 , 22 Recent studies using AOSLO demonstrated the RPE cells are enlarged in CHM in comparison to normal. 23 …”

Short-term Assessment of Subfoveal Injection of Adeno-Associated Virus-Mediated hCHM Gene Augmentation in Choroideremia Using Adaptive Optics Ophthalmoscopy

“…One grader (J.I.W.M.) manually identified cones in all ROIs using custom software . The grader was masked to injected vs control eye and time point for each study participant and was able to adjust the brightness and contrast of the image while selecting cones.…”

“…manually identified cones in all ROIs using custom software. 47 The grader was masked to injected vs control eye and time point for each study participant and was able to adjust the brightness and contrast of the image while selecting cones. Cone centers were used to determine Voronoi boundaries and bound cone density was calculated for each ROI.…”