3-T magnetic resonance imaging simultaneous automated multimodal approach improves detection of ambiguous visual hippocampal sclerosis.

BACKGROUND AND PURPOSE: To evaluate if an automatic magnetic resonance imaging (MRI) processing system may improve detection of hippocampal sclerosis (Hs) in patients with mesial temporal lobe epilepsy (MTLE). METHODS: Eighty consecutive patients with a diagnosis of MTLE and 20 age- and sex-matched controls were prospectively recruited and included in our study. The entire group had 3-T MRI visual assessment of Hs analysed by two blinded imaging epilepsy experts. Logistic regression was used to evaluate the performances of neuroradiologists and multimodal analysis. RESULTS: The multimodal automated tool gave no evidence of Hs in all 20 controls and classified the 80 MTLE patients as follows: normal MRI (54/80), left Hs (14/80), right Hs (11/80) and bilateral Hs (1/80). Of note, this multimodal automated tool was always concordant with the side of MTLE, as determined by a comprehensive electroclinical evaluation. In comparison with standard visual assessment, the multimodal automated tool resolved five ambiguous cases, being able to lateralize Hs in four patients and detecting one case of bilateral Hs. Moreover, comparing the performances of the three logistic regression models, the multimodal approach overcame performances obtained with a single image modality for both the hemispheres, reaching a global accuracy value of 0.97 for the right and 0.98 for the left hemisphere. CONCLUSIONS: Multimodal quantitative automated MRI is a reliable and useful tool to depict and lateralize Hs in patients with MTLE, and may help to lateralize the side of MTLE especially in subtle and uncertain cases.

Machine learning on brain MRI data for differential diagnosis of Parkinson's disease and Progressive Supranuclear Palsy.

BACKGROUND: Supervised machine learning has been proposed as a revolutionary approach for identifying sensitive medical image biomarkers (or combination of them) allowing for automatic diagnosis of individual subjects. The aim of this work was to assess the feasibility of a supervised machine learning algorithm for the assisted diagnosis of patients with clinically diagnosed Parkinson's disease (PD) and Progressive Supranuclear Palsy (PSP). METHOD: Morphological T1-weighted Magnetic Resonance Images (MRIs) of PD patients (28), PSP patients (28) and healthy control subjects (28) were used by a supervised machine learning algorithm based on the combination of Principal Components Analysis as feature extraction technique and on Support Vector Machines as classification algorithm. The algorithm was able to obtain voxel-based morphological biomarkers of PD and PSP. RESULTS: The algorithm allowed individual diagnosis of PD versus controls, PSP versus controls and PSP versus PD with an Accuracy, Specificity and Sensitivity>90%. Voxels influencing classification between PD and PSP patients involved midbrain, pons, corpus callosum and thalamus, four critical regions known to be strongly involved in the pathophysiological mechanisms of PSP. COMPARISON WITH EXISTING METHODS: Classification accuracy of individual PSP patients was consistent with previous manual morphological metrics and with other supervised machine learning application to MRI data, whereas accuracy in the detection of individual PD patients was significantly higher with our classification method. CONCLUSIONS: The algorithm provides excellent discrimination of PD patients from PSP patients at an individual level, thus encouraging the application of computer-based diagnosis in clinical practice.