AI-Based Intelligent Medical Report Understanding and Personal Health Management System

Medical reports contain important health information, but their technical terminology and numerical results may be difficult for patients to understand. Recent studies have explored the use of large language models to simplify radiology reports, while also identifying concerns related to accuracy, omitted information, and potentially misleading explanations [1–3]. This paper proposes an AI-Based Intelligent Medical Report Understanding and Personal Health Management System designed to make medical information easier to understand and organize. The proposed framework will accept medical reports in PDF or image format, extract relevant text, and generate patient-friendly explanations while aiming to preserve important values, units, and reference ranges. A retrieval-based component is proposed to support explanations using reliable medical sources. Additional planned features include date-wise health-record organization, comparison of comparable report values, and reminders based on user-entered or clinician-provided instructions. The framework emphasizes privacy, factual consistency, and the distinction between educational information and professional medical advice. Its evaluation will examine text-extraction accuracy, preservation of medical information, factual consistency, readability, and usability after implementation. As the system has not yet been experimentally evaluated, this paper presents the proposed design and evaluation plan rather than measured performance results.