Investigating the impact of supervoxel segmentation for unsupervised abnormal brain asymmetry detection

Abstract

Several brain disorders are associated with abnormal brain asymmetries (asymmetric anomalies). Several computerbased methods aim to detect such anomalies automatically. Recent advances in this area use automatic unsupervised techniques that extract pairs of symmetric supervoxels in the hemispheres, model normal brain asymmetries for each pair from healthy subjects, and treat outliers as anomalies. Yet, there is no deep understanding of the impact of the supervoxel segmentation quality for abnormal asymmetry detection, especially for small anomalies, nor of the added value of using a specialized model for each supervoxel pair instead of a single global appearance model. We aim to answer these questions by a detailed evaluation of different scenarios for supervoxel segmentation and classification for detecting abnormal brain asymmetries. Experimental results on 3D MR-T1 brain images of stroke patients confirm the importance of high-quality supervoxels fit anomalies and the use of a specific classifier for each supervoxel. Next, we present a refinement of the detection method that reduces the number of false-positive supervoxels, thereby making the detection method easier to use for visual inspection and analysis of the found anomalies.

Unsupervised Brain Anomaly Detection in MR Images

“…tion (SAAD) [9], [10], [28], for detecting abnormal asymmetries associated with anomalies in 3D MR brain images. Fig.…”

“…We next summarize some experimental results from our method and the baselines extracted from [12], [28]. We refer to these works for a more deep evaluation for SAAD, including more baselines and metrics.…”

“…SymmISF is based on the recent Iterative Spanning Forest (ISF) framework [29] for superpixel segmentation and has three steps: (i) seed sampling followed by multiple iterations of (ii) connected supervoxel delineation, and (iii) seed recomputation to improve delineation. We refer to [12], [28] for more details of different strategies for each of these steps.…”

“…The thesis is the result of the following publications (in chronological order): [32], [33] (honorable mention), [9], [10], [17], [22], [28], [8] (best student paper), [34].…”