Newest papers

Plausible but Unimplantable: A Trustworthiness Audit of Shape Completion Foundation Models for Cranial Defect Reconstruction

Marek Wodziński, Kamil Kwarciak, Daria Hemmerling, Mateusz Danioł · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.30035 · cited by 0 works (OpenAlex) · open-access PDF

French

Un patient qui a perdu une partie de la boîte à crânes a besoin d'un implant qui correspond exactement au défaut, et de tels implants sont toujours dessinés à la main dans le CAD par des spécialistes, ce qui rend chaque cas lent et coûteux. La tâche peut être considérée comme un problème de finition de la forme, de sorte que les modèles de finition prétrainés, et maintenant les premiers modèles de base de finition de la forme, ressemblent à un moyen de supprimer cette étape manuelle. Nous avons audité trois modèles de finition de la forme prétrainés publiquement publiés, deux d'entre eux sont des modèles de base, appliqués à zéro tir à 239 défauts crâniens, certains ont été sculptés à partir de n'importe quel

Machine translation (Meta M2M100); the English original follows.

Abstract

A patient who has lost part of the skull vault needs an implant that fits the defect exactly, and such implants are still drawn by hand in CAD by specialists, which makes each case slow and expensive. The task can be considered as a shape completion problem, so pretrained completion models, and now the first shape completion foundation models, look like a way to remove that manual step. We audited three publicly released pretrained shape completion models, two of them foundation models, applied zero-shot to 239 cranial defects, some carved from public skull datasets and some real post-craniectomy cases. Each output was compared with simply returning the patient's own remaining bone, and tested for whether the predicted material could be placed without intersecting the bone that is still there. On the 210 carved defects none of the three improved on doing nothing, and on the 29 real cases only one did so beyond chance. Yet not one output, from any model in any configuration, was implantable: none managed both to stay clear of the bone the patient still has and to place more material than doing nothing, where a task-specific volumetric model, trained for this task and selected on these cases, was implantable on all but one. The output that most convincingly resembles a skull is no more placeable than the rest, and at these point densities implant thickness cannot be read off the emitted points at all, which is a property of the output representation rather than of any one model. On this evidence the shape completion foundation models we could obtain cannot be trusted and, used as released, are not usable in clinical practice. Substantial research towards trustworthy foundation models for shape completion is still needed.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Audit, Foundation (evidence), Task (project management), Property (philosophy), Skull, Trustworthiness

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange