Newest papers

Learning to Measure and Diagnose: A Clinical-Dictionary Radiomic Bottleneck Model for Thyroid Nodule Classification

Guibin Zhang, Weier Luo, Lei Cao · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.29738 · cited by 0 works (OpenAlex) · open-access PDF

French

L'échographie de la thyroïde soutient l'évaluation de la malignité des nodules, mais les prédictions de classification manquent souvent de preuves liées à des descriptions cliniques familières. La radiomics fournit des mesures explicitement définies qui peuvent soutenir de telles explications à travers un dictionnaire clinique. Cependant, en utilisant ces mesures en tant que preuves interprétables, une estimation fiable des images à ultrasons et la préservation de la fidélité de la mesure lors de l'optimisation des performances de la classification sont nécessaires. Nous proposons un modèle de bouchon radiomique clinique-dictatoire pour la classification des nodules de la thyroïde. Dans ce modèle, l'apprentissage radiomique famili

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

Abstract

Thyroid ultrasound supports the assessment of nodule malignancy, but classification predictions often lack evidence linked to familiar clinical descriptions. Radiomics provides explicitly defined measurements that can support such explanations through a clinical dictionary. However, using these measurements as interpretable evidence requires reliable estimation from ultrasound images and preservation of measurement fidelity when optimizing classification performance. We propose a clinical-dictionary radiomic bottleneck model for thyroid nodule classification. Within this model, Family-Aware Radiomic Learning (FARL) estimates measurements with balanced supervision across feature families. Fidelity-Guided Bottleneck Modeling (FGBM) constructs a fixed bottleneck, followed by joint optimization of measurement estimation and classification under fidelity guidance. Decision evidence decomposition then traces feature contributions through a linear classifier and organizes them by clinical dictionary categories. Experiments on ThyroidXL and TN3K show competitive classification performance, while case studies illustrate feature contributions for benign and malignant nodules. These findings indicate interpretable classification with inspectable quantitative evidence informed by clinical knowledge. Code is available online.

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

Topics: Artificial intelligence, Classifier (UML), Bottleneck, Computer science, Feature (linguistics), Pattern recognition (psychology)

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