ehsan yasrebi; iman zabbah; morteza ramezani; hamid valeh; setare salarniya
Abstract
Abstract
Background: Knee osteoarthritis is one of the common diseases in humans and due to its increasing spread, early diagnosis of this disease is very important. Consideration of cartilage volume in knee osteoarthritis studies from radiological images is very necessary. The aim of this study is ...
Read More
Abstract
Background: Knee osteoarthritis is one of the common diseases in humans and due to its increasing spread, early diagnosis of this disease is very important. Consideration of cartilage volume in knee osteoarthritis studies from radiological images is very necessary. The aim of this study is to help improve the diagnosis of knee osteoarthritis with the help of artificial intelligence and image processing techniques.
Methods: This is a diagnostic study that has been evaluated on 957 MRI images. Images were collected from Tehran Hospital database, such that 111 samples were related to healthy individuals and 48 samples to people with knee osteoarthritis. In this study, in order to diagnose osteoarthritis automatically, a new method called “image distinguishing and teaching it to artificial neural network”,using MATLAB software was used. MRI images were received and after pre-processing they were processed to diagnose osteoarthritis conditions with the help of artificial neural networks.
Findings: Experiments show acceptable performance of the proposed method, such that using this technique the diagnose of knee osteoarthritis was possible, with 93% accuracy.
Results: the proposed model can be used in screening plans in order to identify people in danger of developing osteoarthritis and can serve as doctor assistants.
