Towards Shape-based Knee Osteoarthritis Classification using Graph Convolutional Networks

TL;DR

This paper introduces a graph convolutional network approach for classifying knee osteoarthritis severity based on shape analysis, combining intrinsic dimension reduction with semi-supervised learning.

Contribution

It presents a novel semi-supervised graph-based method for osteophyte grading that effectively handles high-dimensional, non-Euclidean shape data.

Findings

Outperforms extrinsic approaches in osteophyte classification

Effective handling of high-dimensional shape data

Demonstrates potential for improved osteoarthritis diagnosis

Abstract

We present a transductive learning approach for morphometric osteophyte grading based on geometric deep learning. We formulate the grading task as semi-supervised node classification problem on a graph embedded in shape space. To account for the high-dimensionality and non-Euclidean structure of shape space we employ a combination of an intrinsic dimension reduction together with a graph convolutional neural network. We demonstrate the performance of our derived classifier in comparisons to an alternative extrinsic approach.