Automatic symmetry based cluster approach for anomalous brain identification in PET scan image : An Analysis
This work addresses brain anomaly detection for medical diagnosis using PET scans, but it appears incremental as it focuses on analyzing existing symmetry approaches without introducing a new method.
The paper tackled the problem of segmenting PET scan images for brain anomaly identification by analyzing symmetry-based distances in clustering algorithms, but it did not report concrete numerical results.
Medical image segmentation is referred to the segmentation of known anatomic structures from different medical images. Normally, the medical data researches are more complicated and an exclusive structures. This computer aided diagnosis is used for assisting doctors in evaluating medical imagery or in recognizing abnormal findings in a medical image. To integrate the specialized knowledge for medical data processing is helpful to form a real useful healthcare decision making system. This paper studies the different symmetry based distances applied in clustering algorithms and analyzes symmetry approach for Positron Emission Tomography (PET) scan image segmentation. Unlike CT and MRI, the PET scan identifies the structure of blood flow to and from organs. PET scan also helps in early diagnosis of cancer and heart, brain and gastro intestinal ailments and to detect the progress of treatment. In this paper, the scope diagnostic task expands for PET image in various brain functions.