Using transfer learning for automated microbleed segmentation

Mahsa Dadar; Maryna Zhernovaia; Sawsan Mahmoud; Richard Camicioli; Josefina Maranzano; Simon Duchesne

doi:10.3389/fnimg.2022.940849

Using transfer learning for automated microbleed segmentation

Front Neuroimaging. 2022 Aug 26:1:940849. doi: 10.3389/fnimg.2022.940849. eCollection 2022.

Authors

Mahsa Dadar¹, Maryna Zhernovaia², Sawsan Mahmoud², Richard Camicioli³, Josefina Maranzano², Simon Duchesne^{4

5}

Affiliations

¹ Department of Psychiatry, Faculty of Medicine, McGill University, Montreal, QC, Canada.
² Department of Anatomy, Université du Québec à Trois-Rivières, Trois-Rivières, QC, Canada.
³ Department of Medicine, Division of Neurology, University of Alberta, Edmonton, AB, Canada.
⁴ CERVO Brain Research Center, Quebec City, QC, Canada.
⁵ Department of Radiology and Nuclear Medicine, Faculty of Medicine, Université Laval, Quebec City, QC, Canada.

Abstract

Introduction: Cerebral microbleeds are small perivascular hemorrhages that can occur in both gray and white matter brain regions. Microbleeds are a marker of cerebrovascular pathology and are associated with an increased risk of cognitive decline and dementia. Microbleeds can be identified and manually segmented by expert radiologists and neurologists, usually from susceptibility-contrast MRI. The latter is hard to harmonize across scanners, while manual segmentation is laborious, time-consuming, and subject to interrater and intrarater variability. Automated techniques so far have shown high accuracy at a neighborhood ("patch") level at the expense of a high number of false positive voxel-wise lesions. We aimed to develop an automated, more precise microbleed segmentation tool that can use standardizable MRI contrasts.

Methods: We first trained a ResNet50 network on another MRI segmentation task (cerebrospinal fluid vs. background segmentation) using T1-weighted, T2-weighted, and T2^* MRIs. We then used transfer learning to train the network for the detection of microbleeds with the same contrasts. As a final step, we employed a combination of morphological operators and rules at the local lesion level to remove false positives. Manual segmentation of microbleeds from 78 participants was used to train and validate the system. We assessed the impact of patch size, freezing weights of the initial layers, mini-batch size, learning rate, and data augmentation on the performance of the Microbleed ResNet50 network.

Results: The proposed method achieved high performance, with a patch-level sensitivity, specificity, and accuracy of 99.57, 99.16, and 99.93%, respectively. At a per lesion level, sensitivity, precision, and Dice similarity index values were 89.1, 20.1, and 0.28% for cortical GM; 100, 100, and 1.0% for deep GM; and 91.1, 44.3, and 0.58% for WM, respectively.

Discussion: The proposed microbleed segmentation method is more suitable for the automated detection of microbleeds with high sensitivity.

Keywords: cerebrovascular disease; deep neural networks; magnetic resonance imaging; microbleeds; transfer learning.