Effect of Manually Corrected Segmentation Errors on OCTA Measurements
Publish Year: 1397
نوع سند: مقاله کنفرانسی
زبان: English
View: 467
نسخه کامل این Paper ارائه نشده است و در دسترس نمی باشد
- Certificate
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
ACSOMED28_182
تاریخ نمایه سازی: 1 دی 1397
Abstract:
Purpose: To evaluate impact of segmentation error correction on optical coherence tomography (OCTA)measurements. Methods: OCTA scans of eyes from 45 subjects (25 patients with diabetic macular edema and 20 normal subjects) were obtained using an RTVue Avanti spectral-domain optical coherence tomography dvice. Two independent expert graders evaluated the images for segmentation error, and the errors were corrected manually. Frequency of segmentation error and the impact of error correction on retinal thickness and vessel density were assessed. Results: Intraclass correlation coefficients (ICC) demonstrated excellent inter-grader agreement (ICC ≥ 0.9) in all parameters of both groups. Frequency of segmentation error was 100% in diabetic eyes and 15% in healthy eyes need segmentation correction. Segmentation error in outer plexiform layer slabs was more common than inner plexiform layer slabs in both groups (P<0.05). After segmentation error correction, retinal layers thickness and vessel density in superficial capillary plexus were not statistically significantly different from baseline (P> 0.05).However, vessel density increased significantly in deep capillary plexus in diabetic patients after correction of segmentation error (P<0.05). Conclusion: Segmentation error is common in OCTA images of patients with diabetic macular edema. Proper attention is needed to the segmentation boundaries when interpreting the OCTA images in diabetic eyes with macular edema.
Authors
Khalil Ghasemi Falavarjani
Eye Research Center, Iran University of Medical Sciences
Abbas Habibi
Eye Research Center, Iran University of Medical Sciences
Maryam Ashraf Khorasani
Eye Research Center, Iran University of Medical Sciences
Shahriar Ghasemizadeh
Eye Research Center, Iran University of Medical Sciences