{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T08:46:55Z","timestamp":1785833215621,"version":"3.56.0"},"posted":{"date-parts":[[2026,7,31]]},"group-title":"Cardiovascular Medicine","reference-count":26,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T00:00:00Z","timestamp":1785456000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/publicdomain\/zero\/1.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,7,31]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>Purpose<\/jats:title>\n                  <jats:p>Quantifying aortic morphology is central to surgical planning for thoracic endovascular aortic repair (TEVAR), yet no consensus exists on how best to represent three-dimensional aortic shape for outcome prediction. Two broad strategies have emerged: statistical shape analysis (SSA), which relies on statistical methods and dimensionality reduction to capture the most significant shape modes, and geometrically-informed approaches that extract descriptors grounded in differential geometry. Here, we directly compare these paradigms on a cohort of 290 CTA scans classified by surgical outcome (non-pathological, successful TEVAR, failed TEVAR).<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>\n                    For the geometrically-informed approach, we use a two-dimensional feature space using normalized fluctuation in integrated Gaussian curvature\n                    <jats:inline-formula>\n                      <jats:inline-graphic xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"26359299v1_inline1.gif\"\/>\n                    <\/jats:inline-formula>\n                    and mean aortic radius\n                    <jats:bold>\n                      <jats:italic>R<\/jats:italic>\n                    <\/jats:bold>\n                    . For SSA, we construct a point-cloud shape model with dimensionality reduction using Principal Component Analysis (PCA) and evaluate classification performance as a function of the number of retained principal components.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>\n                    SSA\u2019s leading principal components encode variations in global aortic size and are statistically redundant with (\n                    <jats:bold>\n                      <jats:italic>R<\/jats:italic>\n                    <\/jats:bold>\n                    ,\n                    <jats:inline-formula>\n                      <jats:inline-graphic xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"26359299v1_inline2.gif\"\/>\n                    <\/jats:inline-formula>\n                    ), yet they lack a one-to-one correspondence with interpretable anatomical quantities. Testing on an unseen, independent dataset reveals that the geometrically-informed approach provided better generalizability than SSA. Using Gaussian process classification with 10-fold cross-validation, we find that the geometrically-informed approach achieves a higher weighted\n                    <jats:bold>\n                      <jats:italic>F<\/jats:italic>\n                      <jats:sub>1<\/jats:sub>\n                    <\/jats:bold>\n                    score than SSA achieves with up to 20 principal components. While SSA\u2019s full-dataset accuracy rises above 90% with increasing dimensionality, this gain is driven by overfitting rather than genuine discriminative power.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusion<\/jats:title>\n                  <jats:p>These results demonstrate that geometrically-informed descriptors offer a more interpretable, robust, and clinically translatable framework for aortic disease classification than data-driven statistical shape representations.<\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.07.29.26359299","type":"posted-content","created":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:40:14Z","timestamp":1785537614000},"source":"Crossref","is-referenced-by-count":0,"title":["The Case for Interpretable Geometry: Statistical Shape Models vs. Curvature-Based Descriptors in Aortic Disease Classification"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-5898-9424","authenticated-orcid":false,"given":"Duc","family":"Nguyen","sequence":"first","affiliation":[{"name":"Department of Physics, University of Chicago, Chicago, 60637, Illinois, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joseph A.","family":"Pugar","sequence":"additional","affiliation":[{"name":"Department of Surgery, University of Chicago, Chicago, 60637, Illinois, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Luka","family":"Pocivavsek","sequence":"additional","affiliation":[{"name":"Department of Surgery, University of Chicago, Chicago, 60637, Illinois, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026080401250457000_2026.07.29.26359299v1.1","doi-asserted-by":"publisher","DOI":"10.1161\/CIR.0000000000001106"},{"key":"2026080401250457000_2026.07.29.26359299v1.2","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2023.107052"},{"key":"2026080401250457000_2026.07.29.26359299v1.3","doi-asserted-by":"publisher","DOI":"10.1016\/j.jtcvs.2016.09.018"},{"key":"2026080401250457000_2026.07.29.26359299v1.4","doi-asserted-by":"publisher","DOI":"10.1007\/s10439-012-0691-4"},{"key":"2026080401250457000_2026.07.29.26359299v1.5","unstructured":"Carmo, M.P. : Differential Geometry of Curves and Surfaces. Prentice\u2013Hall, Englewood Cliffs, NJ (1976)"},{"key":"2026080401250457000_2026.07.29.26359299v1.6","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1011815"},{"key":"2026080401250457000_2026.07.29.26359299v1.7","doi-asserted-by":"publisher","DOI":"10.1007\/s13239-026-00827-z"},{"key":"2026080401250457000_2026.07.29.26359299v1.8","doi-asserted-by":"publisher","DOI":"10.3390\/jpm10020028"},{"key":"2026080401250457000_2026.07.29.26359299v1.9","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2509.09718"},{"key":"2026080401250457000_2026.07.29.26359299v1.10","doi-asserted-by":"publisher","DOI":"10.3390\/jcm11061687"},{"key":"2026080401250457000_2026.07.29.26359299v1.11","doi-asserted-by":"publisher","DOI":"10.3390\/jcm13154577"},{"key":"2026080401250457000_2026.07.29.26359299v1.12","doi-asserted-by":"publisher","DOI":"10.1007\/s12265-022-10335-9"},{"key":"2026080401250457000_2026.07.29.26359299v1.13","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-025-94310-9"},{"key":"2026080401250457000_2026.07.29.26359299v1.14","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2025.108924"},{"key":"2026080401250457000_2026.07.29.26359299v1.15","doi-asserted-by":"publisher","DOI":"10.1186\/s12880-016-0142-z"},{"key":"2026080401250457000_2026.07.29.26359299v1.16","doi-asserted-by":"publisher","DOI":"10.1093\/ejcts\/ezy339"},{"key":"2026080401250457000_2026.07.29.26359299v1.17","doi-asserted-by":"publisher","DOI":"10.1080\/14786440109462720"},{"key":"2026080401250457000_2026.07.29.26359299v1.18","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2016.06.025"},{"key":"2026080401250457000_2026.07.29.26359299v1.19","doi-asserted-by":"publisher","DOI":"10.1002\/hbm.22525"},{"key":"2026080401250457000_2026.07.29.26359299v1.20","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2008.2006604"},{"key":"2026080401250457000_2026.07.29.26359299v1.21","unstructured":"NUREA: PraevAorta: Advanced Aortic Image Analysis Software. https:\/\/www.nurea.com\/praevaorta (2025)"},{"key":"2026080401250457000_2026.07.29.26359299v1.22","doi-asserted-by":"publisher","DOI":"10.1109\/TDPVT.2004.1335277"},{"key":"2026080401250457000_2026.07.29.26359299v1.23","doi-asserted-by":"publisher","DOI":"10.1080\/10255842.2022.2128672"},{"key":"2026080401250457000_2026.07.29.26359299v1.24","first-page":"2825","article-title":"Scikit-learn: Machine Learning in Python","volume":"12","year":"2011","journal-title":"Journal of Machine Learning Research"},{"key":"2026080401250457000_2026.07.29.26359299v1.25","doi-asserted-by":"publisher","DOI":"10.2196\/26151"},{"key":"2026080401250457000_2026.07.29.26359299v1.26","doi-asserted-by":"publisher","DOI":"10.1016\/j.avsg.2024.12.008"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.07.29.26359299","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T08:25:13Z","timestamp":1785831913000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.07.29.26359299"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,31]]},"references-count":26,"URL":"https:\/\/doi.org\/10.64898\/2026.07.29.26359299","relation":{},"subject":[],"published":{"date-parts":[[2026,7,31]]},"subtype":"preprint"}}