Handwritten Prescription Analysis

Abstract

"Doctors are issuing a prescription to patients, which have prescribed medicines for relevant illness/disease. Usually, these prescriptions are handwritten documents. In general, every person has a unique handwriting style. When it comes to doctors’ handwritings, the written content is mostly unidentifiable. Majority of them use cursive handwriting style to write. Therefore, the written content of the prescription is mostly incomprehensive. It requires a lot of training to understand what is written in cursive handwriting. Since it contains medical information, it needs to interpret properly. As a solution for above mentioned problem, from this research the author has proposed the solution to develop a system where the user can upload the image of the prescription and make it interpret. Therefore, the pharmacists can issue medicines according to the interpreted prescription. The system will be used a dataset extracted from prescriptions to train the machine learning model to convert the handwritten characters to digital text.

"