Local Clustering for Lung Cancer Image Classification via Sparse Solution Technique

TL;DR

This work reframes lung cancer CT image classification as seed-based local clustering on a graph of images, enabling small, label-consistent clusters to be identified around query images. It combines box-spline tight-wavelet based image simplification with a Modified Exponential Distance adjacency and two sparse-solution clustering algorithms (LCE and LSC) to perform classification without training. Empirical results on IQ-OTH/NCCD and Pneumonia datasets show competitive accuracy relative to benchmarks while offering faster inference, highlighting the method as a practical complementary tool for computer-aided diagnosis. The paper also discusses data augmentation via optimal transport and GBC-based deformations to mitigate limited labeled data and improve robustness.

Abstract

In this work, we propose to use a local clustering approach based on the sparse solution technique to study the medical image, especially the lung cancer image classification task. We view images as the vertices in a weighted graph and the similarity between a pair of images as the edges in the graph. The vertices within the same cluster can be assumed to share similar features and properties, thus making the applications of graph clustering techniques very useful for image classification. Recently, the approach based on the sparse solutions of linear systems for graph clustering has been found to identify clusters more efficiently than traditional clustering methods such as spectral clustering. We propose to use the two newly developed local clustering methods based on sparse solution of linear system for image classification. In addition, we employ a box spline-based tight-wavelet-framelet method to clean these images and help build a better adjacency matrix before clustering. The performance of our methods is shown to be very effective in classifying images. Our approach is significantly more efficient and either favorable or equally effective compared with other state-of-the-art approaches. Finally, we shall make a remark by pointing out two image deformation methods to build up more artificial image data to increase the number of labeled images.

Figures (12)

• Figure 1: An example of image simplification by using box spline-based tight-wavelet frames courtesy of Ming-Jun Lai, an author of GL13.
• Figure 2: Two examples of lung image simplification based on the LN method.
• Figure 3: Adjacency matrices built without (left) and with the LN method (right) based on the Modified Exponential Distance.
• Figure 4: Given a set of faces (left), we sorted these faces (right) which were done by using LCE or LSC.
• Figure 5: Given a set of faces (left), we sorted these faces (right) which were done by using LCE or LSC.

Theorems & Definitions (2)

• Example 1
• Example 2