Knee Joint Osteoarthritis Data of ''Exploring the Efficacy of Base Data Augmentation Methods in Deep Learning-Based Radiograph Classification of Knee Joint Osteoarthritis''

Published: 16 January 2024| Version 1 | DOI: 10.17632/btkf32tb8n.1
Contributors:
Fabi Prezja,
,
,

Description

This repository contains the datasets generated and utilized to achieve the results presented in our study, available at https://www.mdpi.com/1999-4893/17/1/8. The augmentation and processing name of each dataset match those in the study. Additionally, the Top 20 Grad-CAM images from the adversarial outlier analysis are provided. The initial images originated from https://data.mendeley.com/datasets/56rmx5bjcr/1, which obtained raw data organized by OAI (https://nda.nih.gov/oai). We further processed and modified this data in several ways, including correction of all negative radiographs, contrast adjustment and adversarial/supportive additive augmentations (see the paper for more approaches and details). Any potential citations to our modified datasets must include all prior works mentioned to acknowledge the foundational public datasets.

Files

Steps to reproduce

Please refer to the provided description and the accompanying article for more information.

Categories

Osteoarthritis, Computer Imaging, Deep Learning

Licence