Quantifying Explainability in OCT Segmentation of Macular Holes and Cysts - A SHAP-Based Coverage and Factor Contribution Analysis

Sep 30, 2026

Uploaded by Mia

Optical coherence tomography (OCT) enables measurement of macular hole morphology and size for diagnosis and treatment planning. This study develops an automatic deep learning segmentation model for macular holes and cysts from OCT macular volumes and assesses decision reliability through GradientSHAP-based explainability. Segmentation performance is quantified with Dice/IoU, boundary accuracy with HD95, and calibration with ECE. Explainability is evaluated using lesion coverage, spatial focus, and leakage metrics derived from SHAP attributions clustered by PCA and k-means.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages24

Reading Time1 h 0 min

AccessFree to download

Abstract

Background and objective

Methods

Results

Conclusions

Introduction

Article Quantifying Explainability in OCT Segmentation of Macular Holes and Cysts: A SHAP-Based Coverage and Factor Contribution Analysis ˙Ilknur Tuncer Fırat 1, Murat Fırat 2 and Taner Tuncer 3, * Academic Editors: Michał Strzelecki, Adam Piórkowski, Karolina Nurzynska and Rafał Obuchowicz Received: 11 October 2025 Revised: 17 December 2025 Accepted: 24 December 2025 Published: 27 December 2025 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license. 1 Faculty of Medicine, Inonu University, Malatya 44050, Turkey; [ilknur.tuncer@inonu.edu.tr](ilknur.tuncer@inonu.edu.tr) 2 Faculty of Medicine, Malatya Turgut Ozal University, Malatya 44900, Turkey; [murat.firat@ozal.edu.tr](murat.firat@ozal.edu.tr) 3 Department of Computer Engineering, Firat University, Elazı ˘g 23200, Turkey * [Correspondence: ttuncer@firat.edu.tr](Correspondence: ttuncer@firat.edu.tr) Abstract Background: Optical coherence tomography (OCT) can quantify the morphology and dimensions of a macular hole for diagnosis and treatment planning. Objective: The aim of this study was to perform automatic segmentation of macular holes (MHs) and cysts from OCT macular volumes using a deep learning-based model and to quantitatively evaluate decision reliability using the model’s focus regions and GradientSHAP-based explainability. Methods: In this study, we automatically segmented MHs and cysts in OCT images from the open-access OIMHS dataset. The dataset comprises 125 eyes from 119 patients and 3859 OCT B-scans. OCT B-scan slices were input to a UNet-48-based model with a 2.5D stacking strategy. Performance was evaluated using Dice and intersection-over-union (IoU), boundary accuracy was evaluated using the 95th-percentile Hausdorff distance (HD95), and calibration was evaluated using the expected calibration error (ECE). Explainability was quantified from GradientSHAP maps using lesion coverage and spatial focus metrics: Attribution Precision in Lesion (APILτ), which is the proportion of attributions (SHAP contributions) falling inside the lesion; Attribution Recall in Lesion (ARILτ), which is the proportion of the true lesion covered by the attributions; and leakage (Leakτ = 1 − APILτ), which is the proportion of attributions falling outside the lesion. Spatial focus was monitored using the center-of-mass distance (COM-dist), which is the Euclidean distance between the attribution center and the segmentation center. All metrics were calculated using the top τ% of the pixels with the highest SHAP values. SHAP features were clustered using PCA and k-means. Explanations were calculated using the clinical mask in ground truth (GT) mode and the model segmentation in prediction (Pred) mode. Results: The Dice/IoU values for holes and cysts were 0.94/0.91 and 0.87/0.81, respectively. Across lesion classes, HD95 = 6 px and ECE = 0.008, indicating good boundary accuracy and calibration. In GT mode (τ = 20), three regimes were observed: (i) retina-dominant: high ARIL (hole: 0.659; cyst: 0.654), high Leak (hole: 0.983; cyst: 0.988), and low COM-dist (hole: 7.84 px; cyst: 6.91 px), with the focus lying within the retina and largely confined to the retinal tissue; (ii) peri-lesional: highest ARIL (hole: 0.684; cyst: 0.719), relatively lower Leak (hole: 0.917; cyst: 0.940), and medium/high COM-dist (hole: 16.22 px; cyst: 10.17 px), with the focus located around the lesion; (iii) narrow-coverage: primarily seen for cysts in GT mode (ARIL: 0.494; Leak: 1.000; COM-dist: 52.02 px), with markedly reduced coverage. In Pred mode, the ARIL20 for holes increased in the retina-dominant cluster (0.758) and COM-dist decreased (6.24 px), indicating better agreement with the model segmentation. Conclusions: The model exhibited high accuracy and good calibration for MH and cyst segmentation in OCT images. Quantitative characte

FAQs

What task does the study focus on for OCT images?

The study automatically segments macular holes and cysts from OCT macular volumes using a deep learning model, then evaluates the reliability of decisions with explainability methods.

How is segmentation performance and boundary accuracy measured?

Performance is evaluated with Dice and intersection-over-union (IoU), while boundary accuracy is assessed using the 95th-percentile Hausdorff distance (HD95).

Which explainability metrics are used and what do they represent?

The study quantifies lesion coverage with Attribution Precision in Lesion (APILτ) and Attribution Recall in Lesion (ARILτ), and leakage (Leakτ = 1 − APILτ) for attributions outside the lesion; spatial focus is measured by the center-of-mass distance (COM-dist).

Sep 30, 2026

Uploaded by Mia