Improving Medical Diagnosis with LLM Reprompting

Abstract

In this paper, an innovative approach to enhancing medical diagnosis through the use of prompt engineering in large language models by introducing a systematic, multi-step algorithm that utilizes prompt engineering to construct refined prompts for a language model, optimizing its accuracy and relevance in generating diagnostic insights. The algorithm includes components for input processing, trigger token matching, template selection, and response generation, effectively structuring the interaction between user input and model output to ensure high-quality responses. The system’s efficacy was evaluated using clinical prompts, demonstrating its potential to improve diagnostic accuracy and support clinicians in decision-making. This paper contributes to the growing field of AI in healthcare by providing a scalable, adaptable framework for automatic prompt generation, offering valuable insights for improving clinical support.