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

Key points

• Thyroid ultrasound supports the assessment of nodule malignancy, but classification predictions often lack evidence linked to familiar clinical descriptions.1
• Radiomics provides explicitly defined measurements that can support such explanations through a clinical dictionary.1
• Within this model, Family-Aware Radiomic Learning (FARL) estimates measurements with balanced supervision across feature families.1
• Experiments on ThyroidXL and TN3K show competitive classification performance, while case studies illustrate feature contributions for benign and malignant nodules.1

Sources (1)

• [1]Learning to Measure and Diagnose: A Clinical-Dictionary Radiomic Bottleneck Model for Thyroid Nodule Classification
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 21, 01:38 PM

  We propose a clinical-dictionary radiomic bottleneck model for thyroid nodule classification.

  Thyroid ultrasound supports the assessment of nodule malignancy, but classification predictions often lack evidence linked to familiar clinical descriptions.

Extractive summary: sentences quoted from the sources.