Fine-grained Abnormality Detection and Natural Language Description of Medical CT Images Using Large Language Models

Abstract

Medical report generation demands accurate abnormality detection and precise description generation from CT images. While large language models have shown promising results in natural language processing tasks, their application in medical imaging analysis faces challenges due to the complexity of fine-grained feature detection and the requirement for domain-specific knowledge. This paper presents a novel framework integrating large language models with specialized medical image processing techniques for fine-grained abnormality detection and natural language description generation. Our approach incorporates a multi-modal knowledge enhancement module and a hierarchical attention mechanism to bridge the gap between visual understanding and textual description. The framework employs an adapter-based architecture for efficient domain adaptation and introduces a medical knowledge-enhanced loss function to improve description accuracy. Experimental results on three public datasets demonstrate the effectiveness of our approach, achieving 94.6% detection accuracy and a BLEU-4 score of 0.421 for description generation, surpassing current state-of-the-art methods. The system shows particular strength in handling subtle abnormalities, with a 91.2% average precision in fine-grained detection tasks. Comprehensive ablation studies validate the contribution of each component, while qualitative analysis demonstrates the clinical relevance of generated descriptions. The proposed framework represents a significant advancement in automated medical image analysis, offering potential benefits for clinical workflow optimization and diagnostic support.

Authors and Affiliations

Zhongwen Zhou , Siwei Xia , Mengying Shu , Hong Zhou

Keywords

Related Articles

Disease Prediction System using Support Vector Machine and Multilinear Regression

Evolution of modern technologies like data science and machine learning has opened the path for healthcare communities and medical institutions, to detect the diseases earliest as possible and it helps to provide better...

Quality Evaluation of Drinking Water in Softening Plant at Libyan Iron and Steel Company

The evaluation of quality for drinking water produced in softening plant of Libyan Iron and Steel Company in Misurata were studied at four sites in distribution network form January to June 2013. Some physicochemical par...

A Compact Circularly Polarized MIMO Array Diversified Antenna for 5G Mobile Applications

One of the most active areas of research in the field of communication systems today is wireless technology, and a study of communication systems would be lacking without an understanding of how antennas work and are con...

Ethylene Glycol: Kinetics of the Formation from Methanol–Formaldehyde Solutions

The mechanism and kinetics are developed for the initiated nonbranched-chain formation of ethylene glycol in methanol–formaldehyde solutions at formaldehyde concentrations of 0.1–3.1 mol dm–3 and temperatures of 373–473...

Properties of Concrete on Adding Polypropylene Fibre and Polyvinyl Chloride Fibre

Concrete is a structure material that can't be ignored indeed if it's weak in tension and lead to environmental problems. Properties of concrete can be modified using fibre in concrete. Using fibre from waste of plastic...

Download PDF file
  • EP ID EP753591
  • DOI 10.55524/ijircst.2024.12.6.8
  • Views 44
  • Downloads 0

How To Cite

Zhongwen Zhou, Siwei Xia, Mengying Shu, Hong Zhou (2025). Fine-grained Abnormality Detection and Natural Language Description of Medical CT Images Using Large Language Models. International Journal of Innovative Research in Computer Science and Technology, 13(1), -. https://europub.co.uk/articles/-A-753591