Local ischaemia morphology in diabetics differs from healthy controls

Abstract

Aims/Purpose: To evaluate local macular ischaemia morphology additional to global metrics in diabetic macular ischaemia (DMI) patients and healthy controls.Methods: The study was registered: DRKS‐ID DRKS00024399. Informed written consent was obtained from all participants. Seven DMI patients and seven patients after cataract surgery were included as healthy controls. 10 × 10° Heidelberg Spectralis OCT Angiography was recorded and divided to superficial/deep vascular complexes (SVC/DVC) following the inbuilt segmentation. Further processing was done in MATLAB R2022a. After binarization following Otsu's method for threshold determination, minimum distances to blood vessels were calculated for each pixel applying a sliding window. More than 30 μm distance to the next vessel was considered ischaemic. Morphology analysis of the resulting ischaemic lesions was conducted for SVC and DVC separately. The total ischaemic area; mean ischaemic area per lesion; area and circularity of the second largest ischaemic lesion—avoiding the foveal avascular zone—was extracted. Two‐sample Student t‐tests were performed testing differences of healthy controls and DMI patients in SVC and DVC for each of the four named metrics. Statistical significance was defined when p < 0.05.Results: Analysing total ischaemic area, statistically significant differences were found for SVC and DVC data. Mean ischaemic area per lesion was significantly larger in DMI patients in SVC but not in DVC. The area of the largest ischaemic lesion avoiding FAZ was significantly larger in DMI patients in both SVC and DVC. Lesion circularity did not reveal significant differences.Conclusions: Supplementing reported global nonperfusion differences, local nonperfusion morphology differs significantly between DMI patients and healthy controls. Findings need to be replicated in a larger trial in the future. Research on structure function correlation in areas of nonperfusion will shed light on clinical implications.