Can Multimodal Decision Models Classify Breast Ultrasound Images? Benchmarking CLEF and CLEF-Flash Across Prompts and Datasets

Key points

• Multimodal decision models return probabilities over predefined alternatives, offering a structured interface for image classification.1
• Dataset B and BUS-BRA use binary benign-malignant classification, whereas BUSI includes normal, benign, and malignant categories.1
• Clinical instructions improve CLEF-Flash accuracy relative to the minimal prompt but reduce CLEF accuracy on all datasets.1
• These findings demonstrate dataset-and prompt-dependent performance and motivate class-specific evaluation before applying multimodal decision models to breast ultrasound analysis.1

Sources (1)

• [1]Can Multimodal Decision Models Classify Breast Ultrasound Images? Benchmarking CLEF and CLEF-Flash Across Prompts and Datasets
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 9, 02:41 PM

  This study benchmarks inference with CLEF and CLEF-Flash on 2818 breast ultrasound images from Dataset B, BUSI, and BUS-BRA.

  Multimodal decision models return probabilities over predefined alternatives, offering a structured interface for image classification.

Extractive summary: sentences quoted from the sources.