1,243 research outputs found
Magic-5: Medical Application on a Grid Infrastructure Connection
No full textSi tratta di un progetto finanziato (sin dal 2003) dall'Istituto Nazionale di Fisica Nucleare e concernente la messa a punto di sistemi di Computer Assisted Diagnosis (CAD) mediante lo sviluppo di algoritmi per l'elaborazione di immagini biomediche, anche attraverso l'uso di GRID.
Al progetto partecipano le sedi universitarie e INFN di Torino, Genova, Pisa, Palermo, Lecce, Sassari, Napoli, Bari.
Responsabile nazionale e' il dott. Cerello (INFN Torino).
Il gruppo leccese e' costutiuto da circa 10 persone tra docenti della Facolta' di ScienzeMM.FF.NN., dottorandi e borsisti. Il coordinatore del gruppo leccese e' il dott. Ivan De Mitri. Egli e' anche responsabile scientifico di una convenzione tra l'Università del Salento e la ASL-Lecce riguardante gli stessi temi
Prediction of the outcome of subjects with mild cognitive impairment: whole-brain analysis of structural MRI data
No full textPredictive models based on Support Vector Machines: whole-brain versus regional analysis of structural MRI in the Alzheimer’s disease
No full textDecision-making systems trained on structural magnetic resonance imaging data of subjects affected by the Alzheimer's disease (AD) and healthy controls (CTRL) are becoming widespread prognostic tools for subjects with mild cognitive impairment (MCI). This study compares the performances of three classification methods based on support vector machines (SVMs), using as initial sets of brain voxels (ie, features): (1) the segmented grey matter (GM); (2) regions of interest (ROIs) by voxel-wise t-test filtering; (3) parceled ROIs, according to prior knowledge. The recursive feature elimination (RFE) is applied in all cases to investigate whether feature reduction improves the classification accuracy. We analyzed more than 600 AD Neuroimaging Initiative (ADNI) subjects, training the SVMs on the AD/CTRL dataset, and evaluating them on a trial MCI dataset. The classification performance, evaluated as the area under the receiver operating characteristic curve (AUC), reaches AUC = (88.9 ± .5)% in 20-fold cross-validation on the AD/CTRL dataset, when the GM is classified as a whole. The highest discrimination accuracy between MCI converters and nonconverters is achieved when the SVM-RFE is applied to the whole GM: with AUC reaching (70.7 ± .9)%, it outperforms both ROI-based approaches in predicting the AD conversion
202 STRUCTURAL MRI ANALYSIS AND HIPPOCAMPAL SEGMENTATION IN THE ASSESSMENT OF ALZHEIMER'S DISEASE
No full textFully automated hippocampus segmentation with virtual ant colonies
No full textThe development of tools for a fully automatic segmenta- tion of the relevant brain structures, such as the hippocam- pus, is potentially very useful for pathologies detection.
In this paper, a method for the automated hippocampal seg- mentation, based on virtual ant colonies, is proposed.
The algorithm used, the Channeler Ant Model (CAM), rep- resents an effective way to segment 3D objects with a com- plex shape in a noisy background. The CAM was modified by inserting a shape knowledge that is crucial to face the hippocampus segmentation.
The algorithm was trained and tested using a database of 56 T1 weighted MRI images with a known manual segmen- tation of the hippocampus volume.
The results are comparable to other methods: an average Dice Index of 0.74 and 0.72 is obtained over the left and right hippocampi, respectively. The lack of a heavy training procedure, because all the model parameters are fixed, and the speed make this approach very effective
- …
