Related Experiment Video
Updated: Nov 29, 2025
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
Published on: September 25, 2019
Accuracy and reproducibility of automated white matter hyperintensities segmentation with lesion segmentation tool: A
Federica Ribaldi1, Daniele Altomare2, Jorge Jovicich3
1Laboratory of Alzheimer's Neuroimaging and Alzheimer's Epidemiology, IRCCS Istituto Centro San Giovanni di Dio Fatebenefratelli, Brescia, Italy; Department of Molecular and Translational Medicine, University of Brescia, Brescia, Italy; Laboratory of Neuroimaging of Aging (LANVIE), University of Geneva, Geneva, Switzerland; Memory Clinic, Geneva University Hospitals, Geneva, Switzerland.
Abstract:
Brain vascular damage accumulate in aging and often manifest as white matter hyperintensities (WMHs) on MRI. Despite increased interest in automated methods to segment WMHs, a gold standard has not been achieved and their longitudinal reproducibility has been poorly investigated. The aim of present work is to evaluate accuracy and reproducibility of two freely available segmentation algorithms. A harmonized MRI protocol was implemented in 3T-scanners across 13 European sites, each scanning five volunteers twice (test-retest) using 2D-FLAIR. Automated segmentation was performed using Lesion segmentation tool algorithms (LST): the Lesion growth algorithm (LGA) in SPM8 and 12 and the Lesion prediction algorithm (LPA). To assess reproducibility, we applied the LST longitudinal pipeline to the LGA and LPA outputs for both the test and retest scans. We evaluated volumetric and spatial accuracy comparing LGA and LPA with manual tracing, and for reproducibility the test versus retest. Median volume difference between automated WMH and manual segmentations (mL) was -0.22[IQR = 0.50] for LGA-SPM8, -0.12[0.57] for LGA-SPM12, -0.09[0.53] for LPA, while the spatial accuracy (Dice Coefficient) was 0.29[0.31], 0.33[0.26] and 0.41[0.23], respectively. The reproducibility analysis showed a median reproducibility error of 20%[IQR = 41] for LGA-SPM8, 14% [31] for LGA-SPM12 and 10% [27] with the LPA cross-sectional pipeline. Applying the LST longitudinal pipeline, the reproducibility errors were considerably reduced (LGA: 0%[IQR = 0], p < 0.001; LPA: 0% [3], p < 0.001) compared to those derived using the cross-sectional algorithms. The DC using the longitudinal pipeline was excellent (median = 1) for LGA [IQR = 0] and LPA [0.02]. LST algorithms showed moderate accuracy and good reproducibility. Therefore, it can be used as a reliable cross-sectional and longitudinal tool in multi-site studies.
Insights
This study evaluated automated brain lesion segmentation tools for aging-related white matter hyperintensities (WMHs). The Lesion Segmentation Tool (LST) longitudinal pipeline significantly improved reproducibility, making it a reliable tool for multi-site research.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Gerontology
Background:
- Aging is associated with accumulating brain vascular damage, often visualized as white matter hyperintensities (WMHs) on MRI.
- Automated WMH segmentation methods are of increasing interest, but their accuracy and longitudinal reproducibility require further investigation.
Purpose of the Study:
- To evaluate the accuracy and reproducibility of two freely available WMH segmentation algorithms: Lesion Segmentation Tool (LST) algorithms (Lesion Growth Algorithm - LGA, and Lesion Prediction Algorithm - LPA).
- To assess the performance of LST algorithms in a multi-site setting using a harmonized MRI protocol.
Main Methods:
- A harmonized 2D-FLAIR MRI protocol was used across 13 European sites with 3T scanners.
- Automated segmentation of WMHs was performed using LGA (SPM8/12) and LPA.
- Reproducibility was assessed using both cross-sectional and a dedicated LST longitudinal pipeline for test-retest scans.
- Accuracy was evaluated against manual tracings for volume and spatial overlap (Dice Coefficient).
Main Results:
- LPA demonstrated the best cross-sectional spatial accuracy (Dice Coefficient: 0.41).
- Cross-sectional reproducibility errors were 20% (LGA-SPM8), 14% (LGA-SPM12), and 10% (LPA).
- The LST longitudinal pipeline significantly reduced reproducibility errors to 0% for LGA and 0% for LPA, with excellent Dice Coefficients (median=1).
Conclusions:
- LST algorithms offer moderate accuracy but good reproducibility for WMH segmentation.
- The LST longitudinal pipeline provides excellent reproducibility, establishing it as a reliable tool for cross-sectional and longitudinal multi-site studies.
