Design and Implementation of a Thyroid Disease Health Management Platform Based on Multimodal Models

Le résumé fourni par la source

Thyroid diseases are common chronic disorders of the endocrine system, with a steadily increasing incidence that poses a major threat to public health. Existing diagnostic and treatment systems still exhibit significant limitations in primary screening capability, individualized health management, and equitable allocation of medical resources. To address these issues, this paper develops a thyroid disease health management platform based on multimodal models, enabling closed-loop management that integrates disease risk prediction, health assessment, and coordinated diagnosis and treatment. The platform takes biochemical and demographic data as textual inputs and thyroid CT images as visual inputs, supporting both unimodal and multimodal risk prediction models. By combining expert knowledge with multimodal-based inference, the system facilitates collaborative decision-making, effectively enhancing primary screening and health management capacity, reducing hospital workload, and providing a novel paradigm for intelligent chronic disease management.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.