{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,30]],"date-time":"2025-12-30T05:20:57Z","timestamp":1767072057391,"version":"3.48.0"},"reference-count":41,"publisher":"MDPI AG","issue":"1","license":[{"start":{"date-parts":[[2025,12,27]],"date-time":"2025-12-27T00:00:00Z","timestamp":1766793600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Diagnostics"],"abstract":"<jats:p>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\u2019s 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 (APIL\u03c4), which is the proportion of attributions (SHAP contributions) falling inside the lesion; Attribution Recall in Lesion (ARIL\u03c4), which is the proportion of the true lesion covered by the attributions; and leakage (Leak\u03c4 = 1 \u2212 APIL\u03c4), 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 \u03c4% 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 (\u03c4 = 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 characterization of SHAP validated the model results. In the clinic, peri-lesion and narrow-coverage conditions are the key situations that require careful interpretation.<\/jats:p>","DOI":"10.3390\/diagnostics16010097","type":"journal-article","created":{"date-parts":[[2025,12,29]],"date-time":"2025-12-29T08:47:27Z","timestamp":1766998047000},"page":"97","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Quantifying Explainability in OCT Segmentation of Macular Holes and Cysts: A SHAP-Based Coverage and Factor Contribution Analysis"],"prefix":"10.3390","volume":"16","author":[{"given":"\u0130lknur Tuncer","family":"F\u0131rat","sequence":"first","affiliation":[{"name":"Faculty of Medicine, Inonu University, Malatya 44050, Turkey"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Murat","family":"F\u0131rat","sequence":"additional","affiliation":[{"name":"Faculty of Medicine, Malatya Turgut Ozal University, Malatya 44900, Turkey"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0526-4526","authenticated-orcid":false,"given":"Taner","family":"Tuncer","sequence":"additional","affiliation":[{"name":"Department of Computer Engineering, Firat University, Elaz\u0131\u011f 23200, Turkey"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2025,12,27]]},"reference":[{"key":"ref_1","first-page":"7294952","article-title":"Idiopathic Macular Hole: A Comprehensive Review of Its Pathogenesis and of Advanced Studies on Metamorphopsia","volume":"2019","author":"Chen","year":"2019","journal-title":"J. Ophthalmol."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"1069","DOI":"10.1016\/j.jfo.2023.02.019","article-title":"Prognostic factors for visual recovery after successful large macular hole surgery using the inverted flap technique","volume":"46","author":"Zgolli","year":"2023","journal-title":"J. Fr. Ophtalmol."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"2611","DOI":"10.1016\/j.ophtha.2013.07.042","article-title":"The International Vitreomacular Traction Study Group classification of vitreomacular adhesion, traction, and macular hole","volume":"120","author":"Duker","year":"2013","journal-title":"Ophthalmology"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1016\/j.oret.2016.09.006","article-title":"The Charles Schepens Lecture: Management Options for Vitreomacular Traction: Use an Individualized Approach","volume":"1","author":"Flynn","year":"2017","journal-title":"Ophthalmol. Retin."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"F\u0131rat, M., F\u0131rat, \u0130.T., \u00dcst\u00fcnda\u011f, Z.F., \u00d6zt\u00fcrk, E., and Tuncer, T. (2025). AI-Based Response Classification After Anti-VEGF Loading in Neovascular Age-Related Macular Degeneration. Diagnostics, 15.","DOI":"10.3390\/diagnostics15172253"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"102539","DOI":"10.1016\/j.compmedimag.2025.102539","article-title":"Retinal OCT image segmentation with deep learning: A review of advances, datasets, and evaluation metrics","volume":"123","author":"Zhang","year":"2025","journal-title":"Comput. Med. Imaging Graph."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Quintana-Quintana, O.J., Aceves-Fern\u00e1ndez, M.A., Pedraza-Ortega, J.C., Alfonso-Francia, G., and Tovar-Arriaga, S. (2025). Deep Learning Techniques for Retinal Layer Segmentation to Aid Ocular Disease Diagnosis: A Review. Computers, 14.","DOI":"10.3390\/computers14080298"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"365","DOI":"10.1038\/s41597-024-03182-7","article-title":"OCTDL: Optical Coherence Tomography Dataset for Image-Based Deep Learning Methods","volume":"11","author":"Kulyabin","year":"2024","journal-title":"Sci. Data"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Herath, H.M.S.S., Yasakethu, S.L.P., Madusanka, N., Yi, M., and Lee, B.-I. (2025). Comparative Analysis of Deep Learning Architectures for Macular Hole Segmentation in OCT Images: A Performance Evaluation of U-Net Variants. J. Imaging, 11.","DOI":"10.3390\/jimaging11020053"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"51","DOI":"10.21037\/atm-20-1789","article-title":"A multi-center study of prediction of macular hole status after vitrectomy and internal limiting membrane peeling by a deep learning model","volume":"9","author":"Hu","year":"2021","journal-title":"Ann. Transl. Med."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"6","DOI":"10.1167\/tvst.11.4.6","article-title":"Predicting Visual Improvement After Macular Hole Surgery: A Combined Model Using Deep Learning and Clinical Features","volume":"11","author":"Lachance","year":"2022","journal-title":"Transl. Vis. Sci. Technol."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"12","DOI":"10.1016\/j.survophthal.2024.09.003","article-title":"The role of artificial intelligence in macular hole management: A scoping review","volume":"70","author":"Mikhail","year":"2025","journal-title":"Surv. Ophthalmol."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"6913","DOI":"10.1038\/s41598-024-57562-5","article-title":"Accuracy of generative deep learning model for macular anatomy prediction from optical coherence tomography images in macular hole surgery","volume":"14","author":"Kwon","year":"2024","journal-title":"Sci. Rep."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"1184","DOI":"10.1097\/IAE.0000000000004409","article-title":"Generative Deep Learning Approach to Predict Posttreatment Optical Coherence Tomography Images of Age-Related Macular Degeneration After 12 Months","volume":"45","author":"Lee","year":"2025","journal-title":"Retina"},{"key":"ref_15","unstructured":"OIMHS Dataset (2025, September 08). Springer Nature Figshare, Available online: https:\/\/springernature.figshare.com\/articles\/dataset\/OIMHS_dataset\/23508453?file=42522673."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"769","DOI":"10.1038\/s41597-023-02675-1","article-title":"OIMHS: An Optical Coherence Tomography Image Dataset Based on Macular Hole Manual Segmentation","volume":"10","author":"Ye","year":"2023","journal-title":"Sci. Data"},{"key":"ref_17","first-page":"4765","article-title":"A unified approach to interpreting model predictions","volume":"30","author":"Lundberg","year":"2017","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Kulyabin, M., Zhdanov, A., Pershin, A., Sokolov, G., Nikiforova, A., Ronkin, M., Borisov, V., and Maier, A. (2024). Segment Anything in Optical Coherence Tomography: SAM 2 for Volumetric Segmentation of Retinal Biomarkers. Bioengineering, 11.","DOI":"10.20944\/preprints202409.0703.v1"},{"key":"ref_19","unstructured":"Frawley, J., Willcocks, C.G., Habib, M., Geenen, C., Steel, D.H., and Obara, B. (2021). Robust 3D U-Net segmentation of macular holes. arXiv."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"e000404","DOI":"10.1136\/bmjophth-2019-000404","article-title":"Macular hole morphology and measurement using an automated three-dimensional image segmentation algorithm","volume":"5","author":"Chen","year":"2020","journal-title":"BMJ Open Ophthalmol."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"208","DOI":"10.3928\/23258160-20220315-02","article-title":"Proof-of-Concept Analysis of a Deep Learning Model to Conduct Automated Segmentation of OCT Images for Macular Hole Volume","volume":"53","author":"Pereira","year":"2022","journal-title":"Ophthalmic Surg. Lasers Imaging Retin."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"296","DOI":"10.1109\/JBHI.2018.2810379","article-title":"Segmentation of Intra-Retinal Cysts From Optical Coherence Tomography Images Using a Fully Convolutional Neural Network Model","volume":"23","author":"Girish","year":"2019","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"1647","DOI":"10.1002\/ima.22893","article-title":"A generalizable approach based on the U-Net model for automatic intraretinal cyst segmentation in SD-OCT images","volume":"33","author":"Ganjee","year":"2023","journal-title":"Int. J. Imaging Syst. Technol."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"273","DOI":"10.1109\/JBHI.2018.2793534","article-title":"Segmentation of Retinal Cysts From Optical Coherence Tomography Volumes via Selective Enhancement","volume":"23","author":"Gopinath","year":"2019","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"\u00c7i\u00e7ek, O., Abdulkadir, A., Lienkamp, S.S., Brox, T., and Ronneberger, O. (2016, January 17\u201321). 3D U-Net: Learning dense volumetric segmentation from sparse annotation. Proceedings of the Medical Image Computing and Computer-Assisted Intervention\u2013MICCAI 2016, Athens, Greece.","DOI":"10.1007\/978-3-319-46723-8_49"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Milletari, F., Navab, N., and Ahmadi, S.A. (2016, January 25\u201328). V-Net: Fully convolutional neural networks for volumetric medical image segmentation. Proceedings of the 2016 Fourth International Conference on 3D Vision (3DV), Stanford, CA, USA.","DOI":"10.1109\/3DV.2016.79"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"2663","DOI":"10.1109\/TMI.2018.2845918","article-title":"H-DenseUNet: Hybrid densely connected U-Net for liver and tumor segmentation from CT volumes","volume":"37","author":"Li","year":"2018","journal-title":"IEEE Trans. Med. Imaging"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"152452","DOI":"10.1109\/ACCESS.2020.3017449","article-title":"Retinal Fluid Segmentation Using Ensembled 2-Dimensionally and 2.5-Dimensionally Deep Learning Networks","volume":"8","author":"Alsaih","year":"2020","journal-title":"IEEE Access"},{"key":"ref_29","unstructured":"Zeyu, Z., Huiyu, D., Guangtao, Z., Bo, J., and Yuye, L. (2025, January 17\u201320). Video-inspired lesion segmentation algorithm for OCT retinal images. Proceedings of the SPIE 13406, Medical Imaging 2025: Image Processing, San Diego, CA, USA."},{"key":"ref_30","first-page":"113","article-title":"Cross-Slice Attention and Evidential Critical Loss for Uncertainty-Aware Prostate Cancer Detection","volume":"15008","author":"Hung","year":"2024","journal-title":"Med. Image Comput. Comput. Assist. Interv."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"646","DOI":"10.1016\/j.bbe.2022.05.005","article-title":"Use of dominant activations obtained by processing OCT images with the CNNs and slime mold method in retinal disease detection","volume":"42","author":"Ergen","year":"2022","journal-title":"Biocybern. Biomed. Eng."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"390","DOI":"10.1136\/bjo.86.4.390","article-title":"Macular hole size as a prognostic factor in macular hole surgery","volume":"86","author":"Ullrich","year":"2002","journal-title":"Br. J. Ophthalmol."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1136\/bjo.2009.159582","article-title":"Macular morphology and visual acuity after macular hole surgery with or without internal limiting membrane peeling","volume":"94","author":"Christensen","year":"2010","journal-title":"Br. J. Ophthalmol."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"549","DOI":"10.1016\/j.ophtha.2017.10.031","article-title":"Fully Automated Detection and Quantification of Macular Fluid in OCT Using Deep Learning","volume":"125","author":"Schlegl","year":"2018","journal-title":"Ophthalmology"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"3292","DOI":"10.1364\/BOE.8.003292","article-title":"Robust total retina thickness segmentation in optical coherence tomography images using convolutional neural networks","volume":"8","author":"Venhuizen","year":"2017","journal-title":"Biomed. Opt. Express"},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"102470","DOI":"10.1016\/j.media.2022.102470","article-title":"Explainable artificial intelligence (XAI) in deep learning-based medical image analysis","volume":"79","author":"Kuijf","year":"2022","journal-title":"Med. Image Anal."},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Hernandez, M., Ramon-Julvez, U., Vilades, E., Cordon, B., Mayordomo, E., and Garcia-Martin, E. (2023). Explainable artificial intelligence toward usable and trustworthy computer-aided diagnosis of multiple sclerosis from Optical Coherence Tomography. PLoS ONE, 18.","DOI":"10.1371\/journal.pone.0289495"},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Mariotti, C., Mangoni, L., Iorio, S., Lombardo, V., Fruttini, D., Rizzo, C., Chhablani, J., Midena, E., and Lupidi, M. (2024). Novel Artificial Intelligence-Based Assessment of Imaging Biomarkers in Full-Thickness Macular Holes: Preliminary Data from a Pivotal Trial. J. Clin. Med., 13.","DOI":"10.3390\/jcm13020628"},{"key":"ref_39","unstructured":"Adebayo, J., Gilmer, J., Muelly, M., Goodfellow, I., Hardt, M., and Kim, B. (2018, January 3\u20138). Sanity checks for saliency maps. Proceedings of the Advances in Neural Information Processing Systems 31 (NeurIPS 2018), Montr\u00e9al, QC, Canada."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"867","DOI":"10.1038\/s42256-022-00536-x","article-title":"Benchmarking saliency methods for chest X-ray interpretation","volume":"4","author":"Saporta","year":"2022","journal-title":"Nat. Mach. Intell."},{"key":"ref_41","unstructured":"Ren, T., Rivera, J.H., Oswal, H., Pan, Y., Chopra, A., Ruzevick, J., and Kurt, M. (2025). Here Comes the Explanation: A Shapley Perspective on Multi-contrast Medical Image Segmentation. arXiv."}],"container-title":["Diagnostics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2075-4418\/16\/1\/97\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,30]],"date-time":"2025-12-30T05:16:35Z","timestamp":1767071795000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2075-4418\/16\/1\/97"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,12,27]]},"references-count":41,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,1]]}},"alternative-id":["diagnostics16010097"],"URL":"https:\/\/doi.org\/10.3390\/diagnostics16010097","relation":{},"ISSN":["2075-4418"],"issn-type":[{"type":"electronic","value":"2075-4418"}],"subject":[],"published":{"date-parts":[[2025,12,27]]}}}