Application of information theoretic unsupervised learning to medical image analysis

Journal Title
Journal ISSN
Volume Title

Automated segmentation of medical images is a challenging problem. The number of segments in a medical image may be unknown a priori, due to the presence or absence of pathological anomalies. Some unsupervised learning techniques that take advantage of information theory concepts may provide a solid approach to the solution of this problem. To this end, there has been the recent development of the Improved “Jump” Method (IJM), a technique that efficiently finds a suitable number of clusters representing different tissue characteristics in a medical image. The IJM works by optimizing an objective function, the margin, that quantifies the quality of particular cluster configurations. Recent developments involving interesting relationships between Spectral Clustering (SC) and kernel Principal Component Analysis (kPCA) are used by the implementation of the IJM to cover the non-linear domain. In this novel SC approach the data is mapped to a new space where the points belonging to the same cluster are collinear if the parameters of a Radial Basis Function (RBF) kernel are adequately selected. After projecting these points onto the unit sphere, IJM measures the quality of different cluster configurations, yielding an algorithm that simultaneously selects the number of clusters, and the RBF kernel parameter. Validation of this method is sought via segmentation of MR brain images in a combination of all major modalities. Such labeled MRI datasets serve as benchmarks for any segmentation algorithm. The effectiveness of the nonlinear IJM is demonstrated in the segmentation of uterine cervix color images for early identification of cervical neoplasia, as an aid to cervical cancer diagnosis. Limitations of the current implementation of IJM are encountered when attempting to segment and MR brain images with multiple sclerosis (MS) lesions. These limitations and a strategy to overcome them are discussed. Finally, an outlook to applying this method to the segmentation of cells in Pap smear test micrographs is laid out.

Unsupervised learning, Medical images, Spectral clustering.