Medical Report Simplifier

Abstract

Medical reports from an MRI, CT, or X-ray imaging are usually filled with complicated terms that a patient fails to understand. Such a failure to comprehend may result in a heightened state of anxiety, lower treatment adherence, and limited patient participation in their care. The paper, therefore, introduces a Hybrid Natural Language Processing (NLP) framework that comprises T5 and LLaMA-based models for converting complex medical text into simple language without losing diagnostic accuracy. Besides, the proposed system has Optical Character Recognition (OCR) for the scanned reports, entity recognition for the medical terms, and deep learning-based simplification with fine-tuned transformer models. The backend is in FastAPI, and the frontend in Flask, thus, combining the whole process from report uploading, text extraction to patient-friendly simplification. Assessment outcomes indicate that the text becomes more readable, and essential medical entities are preserved. This is an indication of the role of AI powered tools in lessening the communication barrier between doctors and patients.