Guwahati: In a great relief for knee osteoarthritis patients, researchers at the Indian Institute of Technology, Guwahati, have developed a Deep Learning (DL)-based framework, namely Osteo HRNet, which automatically assesses the severity of knee osteoarthritis. knee (OA) from X-rays. images
This AI-based model can be used to detect the severity level of the disease and remotely assist doctors for more accurate diagnosis.
Knee osteoarthritis has a prevalence of 28 percent in the country and there is no possible cure for the condition except total joint replacement at an advanced stage, so early diagnosis is essential for pain management and behavior corrections.
MRI and CT scans provide a 3D image of the knee joints for effective diagnosis of knee osteoarthritis, but their availability is limited and expensive.
For routine diagnosis, X-ray imaging is highly effective and more economically feasible.
Researchers have been working to improve automatic detection of knee osteoarthritis from X-ray images or radiographs to aid in clinical evaluation. In this direction, the IIT Guwahati The team has developed an AI-based model to automatically assess the severity of knee OA.
Speaking about the knee osteoarthritis prediction model, Palash Ghosh, Assistant Professor, Department of Mathematics at IIT Guwahati, said: “Compared to others, our model can identify the most medically important area to decide the severity level of osteoarthritis. of the knee, which helps doctors detect the disease accurately at an early stage.
The proposed approach is not a direct plug-and-play of popular deep models. The AI-based model uses an efficient deep convolutional neural network (CNN), i.e., an image recognition algorithm.
This model predicts the severity of knee osteoarthritis according to the World Health Organization (WHO) approved Kellgren and Lawrence (KL) grading scale ranging from grade 0 (low severity) to 4 (high severity). ).
It relies on one of the latest deep models, called High Resolution Network (HRNet), to capture the multi-scale features of knee radiographs.
Speaking about the further application of this work, Professor Arijit Sur, Department of Informatics and Engineering, IIT Guwahati, said: “Although simple, the proposed model can be a good starting point for analyzing inexpensive radiographic modalities such as X-rays. Our group is currently focusing on how efficient models based on deep learning can be designed so that we can work in cheap and easy-to-achieve modalities, such as very low-resolution x-ray images or even photos taken of x-ray plates with a smartphone. “.
The team is working further to reconfigure these models in such a way that they can be implemented in devices with limited resources so that medical professionals can easily obtain an initial but accurate guess for diagnosis.
This work has the potential to alleviate the severe shortage of skilled personnel in this field, especially in rural India.
The research has been accepted for publication in the journal Multimedia Tools and Applications. It was conducted by Rohit Kumar Jain, an MTech Data Science student (now graduate) under the joint supervision of Professor Sur and Dr. Palash Ghosh.
The research team also includes former PhD students of Prof. South at IIT Guwahati, Dr. Prasen Kumar Sharma and Dr. Sibaji Gaj (now a Research Fellow at Cleveland Clinic, Ohio, USA). (ME TOO)
