F J Martinez-Murcia1, J M Górriz1, J Ramírez1, M Moreno-Caballero2, M Gómez-Río2. 1. Signal Processing and Biomedical Applications Research Group, Department of Signal Theory, Networking and Communications, University of Granada, Granada 18071, Spain. 2. Department of Nuclear Medicine, Virgen de las Nieves Hospital, Granada 18071, Spain.
Abstract
PURPOSE: A novel approach to a computer aided diagnosis system for the Parkinson's disease is proposed. This tool is intended as a supporting tool for physicians, based on fully automated methods that lead to the classification of (123)I-ioflupane SPECT images. METHODS: (123)I-ioflupane images from three different databases are used to train the system. The images are intensity and spatially normalized, then subimages are extracted and a 3D gray-level co-occurrence matrix is computed over these subimages, allowing the characterization of the texture using Haralick texture features. Finally, different discrimination estimation methods are used to select a feature vector that can be used to train and test the classifier. RESULTS: Using the leave-one-out cross-validation technique over these three databases, the system achieves results up to a 97.4% of accuracy, and 99.1% of sensitivity, with positive likelihood ratios over 27. CONCLUSIONS: The system presents a robust feature extraction method that helps physicians in the diagnosis task by providing objective, operator-independent textural information about (123)I-ioflupane images, commonly used in the diagnosis of the Parkinson's disease. Textural features computation has been optimized by using a subimage selection algorithm, and the discrimination estimation methods used here makes the system feature-independent, allowing us to extend it to other databases and diseases.
PURPOSE: A novel approach to a computer aided diagnosis system for the Parkinson's disease is proposed. This tool is intended as a supporting tool for physicians, based on fully automated methods that lead to the classification of (123)I-ioflupane SPECT images. METHODS: (123)I-ioflupane images from three different databases are used to train the system. The images are intensity and spatially normalized, then subimages are extracted and a 3D gray-level co-occurrence matrix is computed over these subimages, allowing the characterization of the texture using Haralick texture features. Finally, different discrimination estimation methods are used to select a feature vector that can be used to train and test the classifier. RESULTS: Using the leave-one-out cross-validation technique over these three databases, the system achieves results up to a 97.4% of accuracy, and 99.1% of sensitivity, with positive likelihood ratios over 27. CONCLUSIONS: The system presents a robust feature extraction method that helps physicians in the diagnosis task by providing objective, operator-independent textural information about (123)I-ioflupane images, commonly used in the diagnosis of the Parkinson's disease. Textural features computation has been optimized by using a subimage selection algorithm, and the discrimination estimation methods used here makes the system feature-independent, allowing us to extend it to other databases and diseases.
Authors: Antonio Augimeri; Andrea Cherubini; Giuseppe Lucio Cascini; Domenico Galea; Maria Eugenia Caligiuri; Gaetano Barbagallo; Gennarina Arabia; Aldo Quattrone Journal: EJNMMI Phys Date: 2016-02-16
Authors: Arman Rahmim; Yousef Salimpour; Saurabh Jain; Stephan A L Blinder; Ivan S Klyuzhin; Gwenn S Smith; Zoltan Mari; Vesna Sossi Journal: Neuroimage Clin Date: 2016-02-23 Impact factor: 4.881
Authors: Francisco J Martinez-Murcia; Juan M Górriz; Javier Ramírez; Ignacio A Illán; Fermín Segovia; Diego Castillo-Barnes; Diego Salas-Gonzalez Journal: Front Neuroinform Date: 2017-11-14 Impact factor: 4.081
Authors: Ivan S Klyuzhin; Jessie F Fu; Andy Hong; Matthew Sacheli; Nikolay Shenkov; Michele Matarazzo; Arman Rahmim; A Jon Stoessl; Vesna Sossi Journal: PLoS One Date: 2018-11-05 Impact factor: 3.240
Authors: Victor Comte; Hugo Schmutz; David Chardin; Fanny Orlhac; Jacques Darcourt; Olivier Humbert Journal: Eur J Nucl Med Mol Imaging Date: 2022-05-14 Impact factor: 10.057