Multiple RF classifier for the hippocampus segmentation: Method and validation on EADC-ADNI Harmonized Hippocampal Protocol

Autor: Martina Bocchetta, Stefania Bruno, Paolo Inglese, Giovanni B. Frisoni, Rosalia Maglietta, R. Errico, Nicola Amoroso, Marina Boccardi, Alberto Redolfi, Francesco Sensi, Roberto Bellotti, Sabina Tangaro, Andrea Tateo, Andrea Chincarini
Rok vydání: 2015
Předmět:
Computer science
Biophysics
General Physics and Astronomy
Physics and Astronomy(all)
Hippocampal formation
computer.software_genre
Hippocampus
Imaging
ddc:616.89
Imaging
Three-Dimensional
Voxel
Minimum bounding box
Three-Dimensional/methods
Random forest classifier
medicine
Radiology
Nuclear Medicine and imaging
Segmentation
Science & Technology
02 Physical Sciences
medicine.diagnostic_test
business.industry
Radiology
Nuclear Medicine & Medical Imaging
Hippocampus segmentation
Alzheimer's Disease Neuroimaging Initiative
Magnetic resonance imaging
Pattern recognition
11 Medical And Health Sciences
General Medicine
Alzheimer's disease
06 Biological Sciences
Magnetic Resonance Imaging
Random forest
Nuclear Medicine & Medical Imaging
Radiology Nuclear Medicine and imaging
Affine transformation
Data mining
Artificial intelligence
business
Life Sciences & Biomedicine
Classifier (UML)
computer
Algorithms
Zdroj: Physica Medica, Vol. 31, No 8 (2015) pp. 1085-1091
ISSN: 1120-1797
Popis: The hippocampus has a key role in a number of neurodegenerative diseases, such as Alzheimer's Disease. Here we present a novel method for the automated segmentation of the hippocampus from structural magnetic resonance images (MRI), based on a combination of multiple classifiers. The method is validated on a cohort of 50 T1 MRI scans, comprehending healthy control, mild cognitive impairment, and Alzheimer's Disease subjects. The preliminary release of the EADC-ADNI Harmonized Protocol training labels is used as gold standard. The fully automated pipeline consists of a registration using an affine transformation, the extraction of a local bounding box, and the classification of each voxel in two classes (background and hippocampus). The classification is performed slice-by-slice along each of the three orthogonal directions of the 3D-MRI using a Random Forest (RF) classifier, followed by a fusion of the three full segmentations. Dice coefficients obtained by multiple RF (0.87 ± 0.03) are larger than those obtained by a single monolithic RF applied to the entire bounding box, and are comparable to state-of-the-art. A test on an external cohort of 50 T1 MRI scans shows that the presented method is robust and reliable. Additionally, a comparison of local changes in the morphology of the hippocampi between the three subject groups is performed. Our work showed that a multiple classification approach can be implemented for the segmentation for the measurement of volume and shape changes of the hippocampus with diagnostic purposes.
Databáze: OpenAIRE
Externí odkaz: