Detection of subtle white matter lesions in MRI through texture feature extraction and boundary delineation using an embedded clustering strategy

Kokhaur Ong,Siti Mariyam Shamsuddin,Stephan J Sanders,Seepheng Hang,Weimiao Yu,David M Young,Sarina Sulaiman,Norzaini Rose Mohd Zain,Kahhay Yuen,Hilwati Hashim

doi:10.1038/s41598-022-07843-8

Abstract

White matter lesions (WML) underlie multiple brain disorders, and automatic WML segmentation is crucial to evaluate the natural disease course and effectiveness of clinical interventions, including drug discovery. Although recent research has achieved tremendous progress in WML segmentation, accurate detection of subtle WML present early in the disease course remains particularly challenging. Here we propose an approach to automatic WML segmentation of mild WML loads using an intensity standardisation technique, gray level co-occurrence matrix (GLCM) embedded clustering technique, and random forest (RF) classifier to extract texture features and identify morphology specific to true WML. We precisely define their boundaries through a local outlier factor (LOF) algorithm that identifies edge pixels by local density deviation relative to its neighbors. The automated approach was validated on 32 human subjects, demonstrating strong agreement and correlation (excluding one outlier) with manual delineation by a neuroradiologist through Intra-Class Correlation (ICC = 0.881, 95% CI 0.769, 0.941) and Pearson correlation (r = 0.895, p-value < 0.001), respectively, and outperforming three leading algorithms (Trimmed Mean Outlier Detection, Lesion Prediction Algorithm, and SALEM-LS) in five of the six established key metrics defined in the MICCAI Grand Challenge. By facilitating more accurate segmentation of subtle WML, this approach may enable earlier diagnosis and intervention.

Highlights

White matter lesions (WML) underlie multiple brain disorders, and automatic WML segmentation is crucial to evaluate the natural disease course and effectiveness of clinical interventions, including drug discovery
On Fluid-Attenuated Inversion Recovery (FLAIR) Magnetic Resonance Imaging (MRI), WMLs seen in multiple sclerosis (MS) exhibit similar characteristics to WMLs seen in ischaemia, allowing similar analytic methods to be used for both
To provide a consistent performance comparison across all state-of-the-art methods for 32 subjects, we found that the proposed method showed the optimal performance for each metric except for sensitivity, where it took second to our prior work, trading some sensitivity for increased precision (PPV)

Summary

Introduction

White matter lesions (WML) underlie multiple brain disorders, and automatic WML segmentation is crucial to evaluate the natural disease course and effectiveness of clinical interventions, including drug discovery. Distinguishing different types of outliers, such as iron deposition, from WML helps improve diagnosis and management of age-related cognitive decline These fully automated approaches have the advantage of reducing inter- and intra-rater variability through an objective, consistent analysis after training on a large body of images, at least in theory. With these advancements, fully automated methods have become a preferred scheme for neuroradiologists to analyze and quantify WML in large MRI datasets, especially for longitudinal progression of WML in developmental studies, where large numbers of subjects with subtle changes across time require sensitive, high-throughput techniques

Methods

Results

Discussion

Conclusion