Enhancing multiclass classification of knee osteoarthritis severity grades using oneDNN
by Akshay Bhuvaneswari Ramakrishnan; Shriram K. Vasudevan; T.S. Murugesh; Sini Raj Pulari
International Journal of Bioinformatics Research and Applications (IJBRA), Vol. 19, No. 3, 2023

Abstract: Osteoarthritis of the knee (OA) is a degenerative joint condition affecting around 23% of adult patients in the USA and globally. To diagnose OA early and assess severity grades, knee images were classified into five severity categories using the Knee Osteoarthritis Dataset with Severity Grading dataset from Kaggle. The highest severity grade was grade 4. Pre-processing processes, including data reduction and augmentation, were required to correct the uneven distribution of class information. As the dataset had an uneven distribution of class information, pre-processing processes including data reduction and augmentation were required to correct the problem In addition, the classification process was carried out with the assistance of three widely used convolutional neural network models. We proposed a new architecture and have used three of the following models for classification EfficientNetB5, DenseNet201, and Inception V3. Additionally, all these models are optimised with oneDNN library using oneAPI.

Online publication date: Fri, 29-Sep-2023

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Bioinformatics Research and Applications (IJBRA):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com