Extraction Of Pharmaceutical Data From Medical Prescriptions Using Optical Character Recognition (OCR) Model

This study focuses on effectively adding a layer of abstraction while searching for generic medicines or supplements from prescriptions. We leverage the use of AI which demonstrates effective extraction of all the necessary medicinal information from the intricate prescriptions that most users find hard to read or correlate. By utilizing an Optical Character Recognition (OCR) model we solve this problem. The model is designed to be able to extract all the text information provided in the prescription first and apply some parameters that can filter personal information. It also filters harmful or severe medication materials if mentioned; and warns the user. To provide relevant information about the medications or drugs, the system relies on a trusted API endpoint to display precautions, dosage, sources and other important parameters to better assist users regarding their prescription. This reduces the effort of manually trying to understand details behind a medicine and searching countless sites to look for exactly what the user wants. We developed an automated pipeline using OCR and Large Language Models (LLM). The primary goal of this system is to identify the medicines prescribed to a patient accurately, reducing manual data entry and enhancing the efficiency of medical records processing in healthcare settings.