{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,5,22]],"date-time":"2026-05-22T18:04:45Z","timestamp":1779473085577,"version":"3.53.1"},"posted":{"date-parts":[[2026,5,20]]},"group-title":"Ophthalmology","reference-count":33,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,5,20]],"date-time":"2026-05-20T00:00:00Z","timestamp":1779235200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,5,20]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>\n                  Manufacturer-defined signal-strength indices are frequently employed as quality benchmarks for automated optical coherence tomography analysis, yet their empirical relationship with deep learning segmentation accuracy remains unclear. Because these metrics were originally developed for conventional image-processing pipelines, their ability to predict modern model-based segmentation accuracy has not been empirically validated. To address this gap, we evaluated the Heidelberg Spectralis Q-score against U-Net segmentation performance across 5,047 B-scans from 103 eyes for three anatomical boundaries of the posterior segment of the eye: the Ellipsoid Zone (EZ), Bruch\u2019s Membrane (BM), and Choroid Outer Boundary (COB). Alongside standard boundary agreement metrics (MAE, MSE, Dice Similarity Coefficient), we adapted the Earth Mover\u2019s Distance (EMD) from optimal transport theory as a boundary evaluation metric. Unlike column-wise averages, EMD quantifies boundary agreement as a 2-D geometric displacement, directly measuring residual spatial displacement between the model segmented boundary and the ground-truth boundary. Our results demonstrate that the Q-score \u2014 originally designed to gate image-processing-based automated analysis \u2014 is a poor predictor of deep learning boundary segmentation accuracy, with explained variance (\n                  <jats:italic>R<\/jats:italic>\n                  <jats:sup>2<\/jats:sup>\n                  ) failing to exceed 1.4% across all three boundaries. We further observed a monotonically increasing error hierarchy with anatomical depth (EZ &lt; BM &lt; COB), consistent across metrics, which is unexplained by the signal strength. At the COB, correlations were paradoxically positive, explained by a B-scan-level mediation chain: higher Q-scores correspond to greater choroidal thickness (\n                  <jats:italic>r<\/jats:italic>\n                  = 0.113,\n                  <jats:italic>\u03c1<\/jats:italic>\n                  = 0.158), which in turn predicts higher COB segmentation error (\n                  <jats:italic>r<\/jats:italic>\n                  = 0.165,\n                  <jats:italic>\u03c1<\/jats:italic>\n                  = 0.191) \u2014 a localization difficulty that global signal strength cannot capture. Collectively, these findings challenge the implicit assumption that signal-strength-based quality thresholds are a reliable proxy for deep learning model performance, and motivate a shift toward task-specific acquisition quality criteria calibrated to model performance rather than signal interpretability.\n                <\/jats:p>","DOI":"10.64898\/2026.05.17.26353399","type":"posted-content","created":{"date-parts":[[2026,5,20]],"date-time":"2026-05-20T20:00:19Z","timestamp":1779307219000},"source":"Crossref","is-referenced-by-count":0,"title":["Evaluating OCT Device-Reported Image Quality Score: Towards a Task-Specific Quality Gate for Deep Learning-based Outer-Retina and Choroid Boundary Segmentation"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-8238-1957","authenticated-orcid":false,"given":"Adarsh","family":"Gadari","sequence":"first","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Atharva Ajay","family":"Vichare","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Francesca","family":"Corona","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Universit\u00e0 degli Studi di Cagliari, Cagliari, Sardinia, Italy"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sharat Chandra","family":"Vupparaboina","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shreyaa Rohindra","family":"Lall","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Giulia","family":"Gregori","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Polytechnic University of Marche, Ancona, Italy"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nasiq","family":"Hasan","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jos\u00e9-Alain","family":"Sahel","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jay","family":"Chhablani","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sandeep Chandra","family":"Bollepalli","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kiran Kumar","family":"Vupparaboina","sequence":"additional","affiliation":[{"name":"University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA"},{"name":"Netramind Innovations Inc, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026052210151256000_2026.05.17.26353399v1.1","doi-asserted-by":"crossref","first-page":"2114","DOI":"10.3390\/medicina59122114","article-title":"Optical coherence tomography (oct): a brief look at the uses and technological evolution of ophthalmology","volume":"59","year":"2023","journal-title":"Medicina"},{"key":"2026052210151256000_2026.05.17.26353399v1.2","unstructured":"S. R. Singh , Optical Coherence Tomography Imaging: Advances in Ophthalmology (MDPI-Multidisciplinary Digital Publishing Institute, 2023)."},{"key":"2026052210151256000_2026.05.17.26353399v1.3","first-page":"6655021","article-title":"Optical coherence tomography (angiography) biomarkers in the assessment and monitoring of diabetic macular edema","volume":"2020","year":"2020","journal-title":"J. Diabetes Res"},{"key":"2026052210151256000_2026.05.17.26353399v1.4","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2022.02.001"},{"key":"2026052210151256000_2026.05.17.26353399v1.5","doi-asserted-by":"crossref","unstructured":"A. Gadari , S. Chandra Bollepalli , M. N. Ibrahim , et al., \u201cRobust retinal layer segmentation using OCT B-scans: A novel approach based on Pix2Pix generative adversarial network,\u201d in Proceedings of the 14th ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics, (ACM, 2023), BCB\u201923, pp. 1\u20136.","DOI":"10.1145\/3584371.3612979"},{"key":"2026052210151256000_2026.05.17.26353399v1.6","doi-asserted-by":"crossref","first-page":"104822","DOI":"10.1016\/j.compbiomed.2021.104822","article-title":"Modular deep neural networks for automatic quality control of retinal optical coherence tomography scans","volume":"141","year":"2022","journal-title":"Comput. Biol. Med"},{"key":"2026052210151256000_2026.05.17.26353399v1.7","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0220063"},{"key":"2026052210151256000_2026.05.17.26353399v1.8","doi-asserted-by":"crossref","first-page":"2133","DOI":"10.1167\/iovs.11-8755","article-title":"Signal quality assessment of retinal optical coherence tomography images","volume":"53","year":"2012","journal-title":"Investig. ophthalmology & visual science"},{"key":"2026052210151256000_2026.05.17.26353399v1.9","unstructured":"\u201cOptical Coherence Tomography (OCT) Normative Data Collection Study (S-2013-3),\u201d https:\/\/clinicaltrials.gov\/study\/NCT02014688 (2013). ClinicalTrials.gov Identifier: NCT02014688. Last update posted August 20, 2015. Accessed May 2024."},{"key":"2026052210151256000_2026.05.17.26353399v1.10","doi-asserted-by":"crossref","unstructured":"M. Alhelaly , C. Soylu , D. Oncel , et al., \u201cOct-based evaluation of vitreous haze in the first-line antimetabolites as steroid-sparing treatment (fast) uveitis trial,\u201d Ophthalmol. Sci. p. 101106 (2026).","DOI":"10.1016\/j.xops.2026.101106"},{"key":"2026052210151256000_2026.05.17.26353399v1.11","doi-asserted-by":"crossref","first-page":"34","DOI":"10.1167\/iovs.67.2.34","article-title":"Inner retinal reflectivity loss and concurrent thickening predict short-term conversion to neovascular amd","volume":"67","year":"2026","journal-title":"Investig. Ophthalmol. & Vis. Sci"},{"key":"2026052210151256000_2026.05.17.26353399v1.12","unstructured":"T. E. Clemons et al., \u201cCell-based ciliary neurotrophic factor therapy for macular telangiectasia type 2,\u201d New Engl. J. Med. (2024). ClinicalTrials.gov: NCT03316300, NCT03319849."},{"key":"2026052210151256000_2026.05.17.26353399v1.13","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0034823"},{"key":"2026052210151256000_2026.05.17.26353399v1.14","doi-asserted-by":"crossref","first-page":"1528","DOI":"10.1002\/acn3.51320","article-title":"Artificial intelligence extension of the oscar-ib criteria","volume":"8","year":"2021","journal-title":"Ann. clinical translational neurology"},{"key":"2026052210151256000_2026.05.17.26353399v1.15","doi-asserted-by":"crossref","first-page":"2972","DOI":"10.1364\/BOE.564188","article-title":"Roqus: A retinal oct quality and usability score","volume":"16","year":"2025","journal-title":"Biomed. Opt. Express"},{"key":"2026052210151256000_2026.05.17.26353399v1.16","doi-asserted-by":"crossref","first-page":"496","DOI":"10.1016\/j.ajo.2008.05.032","article-title":"Enhanced depth imaging spectral-domain optical coherence tomography","volume":"146","year":"2008","journal-title":"Am. journal ophthalmology"},{"key":"2026052210151256000_2026.05.17.26353399v1.17","unstructured":"Heidelberg Engineering GmbH, HEYEX 2 (Heidelberg Eye Explorer) PACS and Image Management Software, Heidelberg Engineering GmbH, Heidelberg, Germany (2023). Version 2.6.6 or later."},{"key":"2026052210151256000_2026.05.17.26353399v1.18","doi-asserted-by":"crossref","unstructured":"N. Hasan , A. Gadari , S. C. Vupparaboina , et al., \u201cDeep learning\u2013based automated segmentation and quantification of ellipsoid zone and rpe\u2013bruch\u2019s membrane complex in healthy subjects and geographic atrophy,\u201d Preprints (2026).","DOI":"10.20944\/preprints202604.1517.v1"},{"key":"2026052210151256000_2026.05.17.26353399v1.19","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"2026052210151256000_2026.05.17.26353399v1.20","doi-asserted-by":"crossref","first-page":"737","DOI":"10.3390\/diagnostics16050737","article-title":"Deep learning\u2013based choroidal boundary detection in geographic atrophy using spectral-domain optical coherence tomography","volume":"16","year":"2026","journal-title":"Diagnostics"},{"key":"2026052210151256000_2026.05.17.26353399v1.21","doi-asserted-by":"publisher","DOI":"10.2307\/2286407"},{"key":"2026052210151256000_2026.05.17.26353399v1.22","doi-asserted-by":"crossref","unstructured":"B. S. Everitt and A. Skrondal , The Cambridge Dictionary of Statistics (Cambridge University Press, 2010), 4th ed.","DOI":"10.1017\/CBO9780511779633"},{"key":"2026052210151256000_2026.05.17.26353399v1.23","doi-asserted-by":"crossref","first-page":"315","DOI":"10.1016\/j.compmedimag.2015.09.008","article-title":"Automated estimation of choroidal thickness distribution and volume based on oct images of posterior visual section","volume":"46","year":"2015","journal-title":"Comput. Med. Imaging Graph"},{"key":"2026052210151256000_2026.05.17.26353399v1.24","doi-asserted-by":"crossref","first-page":"104667","DOI":"10.1016\/j.compbiomed.2021.104667","article-title":"Automatic extraction and stenosis evaluation of coronary arteries in invasive coronary angiograms","volume":"136","year":"2021","journal-title":"Comput. biology medicine"},{"key":"2026052210151256000_2026.05.17.26353399v1.25","doi-asserted-by":"crossref","first-page":"99","DOI":"10.1023\/A:1026543900054","article-title":"The earth mover\u2019s distance as a metric for image retrieval","volume":"40","year":"2000","journal-title":"Int. journal computer vision"},{"key":"2026052210151256000_2026.05.17.26353399v1.26","doi-asserted-by":"crossref","unstructured":"C. Villani et al., Optimal transport: old and new, vol. 338 (Springer, 2009).","DOI":"10.1007\/978-3-540-71050-9"},{"key":"2026052210151256000_2026.05.17.26353399v1.27","first-page":"I","article-title":"Fast contour matching using approximate earth mover\u2019s distance,\u201d in Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004","volume":"1","year":"2004","journal-title":"CVPR 2004"},{"key":"2026052210151256000_2026.05.17.26353399v1.28","doi-asserted-by":"crossref","first-page":"509","DOI":"10.1109\/34.993558","article-title":"Shape matching and object recognition using shape contexts","volume":"24","year":"2002","journal-title":"IEEE transactions on pattern analysis machine intelligence"},{"key":"2026052210151256000_2026.05.17.26353399v1.29","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1016\/j.tjem.2018.08.001","article-title":"User\u2019s guide to correlation coefficients","volume":"18","year":"2018","journal-title":"Turkish journal emergency medicine"},{"key":"2026052210151256000_2026.05.17.26353399v1.30","doi-asserted-by":"publisher","DOI":"10.1213\/ANE.0000000000002864"},{"key":"2026052210151256000_2026.05.17.26353399v1.31","doi-asserted-by":"crossref","first-page":"4898","DOI":"10.1002\/mp.17059","article-title":"Evaluating the relationship between magnetic resonance image quality metrics and deep learning-based segmentation accuracy of brain tumors","volume":"51","year":"2024","journal-title":"Med. physics"},{"key":"2026052210151256000_2026.05.17.26353399v1.32","doi-asserted-by":"crossref","unstructured":"S. Dodge and L. Karam , \u201cUnderstanding how image quality affects deep neural networks,\u201d in 2016 eighth international conference on quality of multimedia experience (QoMEX), (IEEE, 2016), pp. 1\u20136.","DOI":"10.1109\/QoMEX.2016.7498955"},{"key":"2026052210151256000_2026.05.17.26353399v1.33","article-title":"Diagnosing model performance under distribution shift","year":"2023","journal-title":"arXiv preprint"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.05.17.26353399","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,22]],"date-time":"2026-05-22T17:15:22Z","timestamp":1779470122000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.05.17.26353399"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,20]]},"references-count":33,"URL":"https:\/\/doi.org\/10.64898\/2026.05.17.26353399","relation":{},"subject":[],"published":{"date-parts":[[2026,5,20]]},"subtype":"preprint"}}