The problem KOA predictor solves

Knee osteoarthritis is currently diagnosed by assessing symptoms and evaluating the radiograph manually by the practitioner. We propose a computer-aided diagnosis method to automatically evaluate score knee Osteoarthritis severity according to the KL scale. This method allows transparent and fair diagnosis that is not subjective and based on previous data and medical parameters. Data set of radio-graphs classified with KL levels related to knee osteoarthritis will be used for training and the patient may know the severity of osteoarthritis in future time periods so that he will be able to take necessary precautions

Challenges I ran into

Dataset gathering, space requirements, CPU and GPU requirements.