Malaria Detection from Blood Smears using Deep Learning

MedDES is a medical diagnostic system for medical diagnosis using deep learning which has 4 diagnostic tests for Malaria, COVID-19, Pneumonia and Brain Tumour. The system also generates a detailed patient report after testing and can be accessed here.

Built 4 light weight CNN models using Keras and Tensorflow backend, one for each test and were trained on respective official datasets. The highly accurate models have an avg inference time of 84 msecs. For enabling easy access, system was built as a web application using Streamlit and deployed using herokuapp and can be accessed here.