The diagnostic process of many neurodegenerative diseases, such as Parkinson, Progressive Supranuclear Palsy, etc., involves the study of brain MRI scans in order to individuate morphological markers that can highlight… Click to show full abstract
The diagnostic process of many neurodegenerative diseases, such as Parkinson, Progressive Supranuclear Palsy, etc., involves the study of brain MRI scans in order to individuate morphological markers that can highlight on the healthy status of the subject. A fundamental step in the pre-processing and analysis of MRI is the identification of the Mid-Sagittal Plane, which corresponds to the mid-brain and allows a coordinate reference system for the whole MRI scans set. To improve the identification of the Mid-Sagittal, we have developed an algorithm in MatlabĀ®, based on the k-means clustering function. The results have been compared with the evaluation of four experts that manually identified the mid-sagittal and whose performances have been crossed with a cognitive decisional algorithm in order to define a gold standard. The comparison provided a mean percentage error of 0.96%. To further refine the automatic procedure, we trained a machine learning considering the results coming from the proposed algorithm and the gold standard. Therefore, we tested the machine learning and obtained results comparable to medical raters with a mean percentage error of 0.65%. Even if the sample of data analyzed needs to be increased, the system is promising and it could be directly incorporated into broader diagnostic support systems.
               
Click one of the above tabs to view related content.