AI in Healthcare Documentation: A Tool for Validating Specificity and Accuracy in Medical Notes

Abstract

Accurate clinical documentation is essential for effective care and diagnosis, as vague or incomplete notes can contribute to serious medical errors.This research proposes an AI-powered validation tool that uses ChatGPT and Google Gemini AI to evaluate the accuracy and specificity of SOAP-formatted clinical notes. The tool integrates a Forest-Based Validation Algorithm to classify each section as Valid, Vague, Not Defined, or Missing. Built using Agile methodology, the tool was tested through both quantitative metrics (accuracy, precision, error rate) and human review. Results show the tool enhances documentation quality while supporting clinician workflows and ethical AI use in healthcare.