AUTOMATED SEGMENTATION OF BRAIN LESIONS BY COMBINING INTENSITY AND SPATIAL INFORMATION Bilwaj Gaonkar, Guray Erus, Nick Bryan, Christos Davatzikos Section of Biomedical Image Analysis, Department of Radiology, University of Pennsylvania, Philadelphia, PA, 19104, USA ABSTRACT Quantitative analysis of brain lesions in large clinical tri- als is becoming more and more important. We present a new automated method, that combines intensity based lesion seg- mentation with a false positive elimination method based on the spatial distribution of lesions. A Support Vector Regres- sor (SVR) is trained on expert-defined lesion masks using im- age histograms as features, in order to obtain an initial lesion segmentation. A lesion probability map that represents the spatial distribution of true and false positives on the inten- sity based segmentation is constructed using the segmented lesions and manual masks. A k-Nearest Neighbor (kNN) clas- sifier based on the lesion probability map is applied to refine the segmentation. Index Terms— Lesion Segmentation, Machine Learning, kNN, SVR, Spatial Learning 1. INTRODUCTION Brain lesions, especially White Matter Lesions (WMLs), are associated with cerebrovascular disease, and also with aging. WMLs are common abnormalities of the brain, which may be the result of different brain diseases, such as multiple scle- rosis and vascular dementia, or may appear in normal elderly subjects. MR imaging is widely used for diagnosing such dis- eases clinically. Manual lesion segmentation by trained ex- perts, a commonly used method, is extremely time consum- ing, and suffers from high intra-observer and inter-observer variability. This raises the demand for automated lesion seg- mentation methods which can reduce both the cost of analysis and the intra-observer and inter-observer variability. Several methods have been developed to automate the process of lesion detection [1, 2, 3, 4]. In [1] the authors use Support Vector Machines with Adaboost to learn a classifier from multi-modality MRIs. While this method can learn from expert delineated datasets, it completely ignores spatial distribution of lesions. On FLAIR images WMLs show up as hyperintensities with respect to surrounding healthy white matter (WM) tissues. However, their intensity range also overlaps with normal gray matter (GM) tissues. Furthermore, artefacts from skull stripping such as regions near the eyes, or artefacts inside the ventricles may fall in the same intensity range as WMLs. This intensity overlap causes the failure of segmentation methods based solely on image intensity and is the major difficulty in accurately segmenting WMLs. In [2] the intensity and the spatial information are com- bined together in a voxel based feature vector and a kNN classifier is used to segment lesions. We followed a similar approach in using the spatial information; however we pro- pose a two level method, where the spatial information is used to refine an intensity based segmentation. One major contribution of our method is that it does not require training on individual voxel intensities, unlike [1, 2], but learns to predict an adaptive threshold from the image his- tograms using a Support Vector Regressor (SVR), which is a robust machine learning method known to generalize well in high dimensional spaces. It effectively examines the pat- tern presented by the histogram, and determines a ‘dynamic’ threshold, i.e. a unique threshold for each individual. Based on the correspondences between manually ex- tracted lesions and the segmented hyperintensities, we con- struct spatial probability maps of true and false positives resulting from the intensity based segmentation. We propose a compact representation of the spatial information, by using centroids and volumes of connected components. The spatial maps are used by a kNN classifier in order to eliminate the false positives and to obtain the final segmentation. We evaluated our method on a WML segmentation task on 38 MR scans for which expert defined truth was available. We obtained very promising results, which outperformed the ones reported in [1]. 2. METHOD We present a detailed description of the two main modules of the proposed method in the following subsections. The intensity based segmentation method is presented in section 2.1. The construction of the spatial probability maps and the kNN classifier is explained in section 2.2