Radius-Optimized Efficient Template Matching for Lesion Detection from Brain Images

Abstract

Computer-aided detection of brain lesions from volumetric magnetic resonance imaging (MRI) is in demand for fast and automatic diagnosis of neural diseases. The templatematching technique can provide satisfactory outcome for automatic localization of brain lesions; however, finding the optimal template size that maximizes similarity of the template and the lesion remains challenging. This increases the complexity of the algorithm and the requirement for computational resources, while processing large MRI volumes with three-dimensional (3D) templates. Hence, reducing the computational complexity of template matching is needed. In this paper, we first propose a mathematical framework for computing the normalized crosscorrelation coefficient (NCCC) as the similarity measure between the MRI volume and approximated 3D Gaussian template with linear time complexity, , as opposed to the conventional fast Fourier transform (FFT) based approach with the complexity , where is the number of voxels in the image and is the number of tried template radii. We then propose a mathematical formulation to analytically estimate the optimal template radius for each voxel in the image and compute the NCCC with the location-dependent optimal radius, reducing the complexity to . We test our methods on one synthetic and two real multiple-sclerosis databases, and compare its performance in lesion detection with FFT and a state-of-the-art lesion prediction algorithm. We demonstrate through our experiments the efficiency of the proposed method for brain lesion detection and its comparable performance with existing techniques.Index Terms-Brain lesion detection, computational complexity, FFT, MRI, NCCC, template matching.

Feasibility of automatic detection of small hepatocellular carcinoma (≤2 cm) in cirrhotic liver based on pattern matching and deep learning

“…This process can then be repeated for a set of templates with various lesion sizes. Pattern matching algorithm has been used in other applications for lesion identifications; for example, small brain metastases (Farjam et al 2012), and brain lesions (Koley et al 2016(Koley et al , 2020. In our study, given overlapping imaging characteristics between sHCC and benign cirrhosis-associated nodules, applying deep learning algorithms directly on the multiparametric images for sHCC detection is challenging.…”