Automatic classification and prediction models for early Parkinson’s disease diagnosis from SPECT imaging

• Published in: Expert systems with applications Vol. 41; no. 7; pp. 3333 - 3342
• Main Authors: Prashanth, R., Dutta Roy, Sumantra, Mandal, Pravat K., Ghosh, Shantanu
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Ltd 01.06.2014; Elsevier
• Subjects: Applied sciences; Biological and medical sciences; Classification; Computer science; control theory; systems; Computer-aided early diagnosis; Data processing. List processing. Character string processing; Diagnosis; Exact sciences and technology; Imaging; Investigative techniques, diagnostic techniques (general aspects); Linear inference, regression; Logistic regression; Logistics; Mathematical models; Mathematics; Medical sciences; Memory organisation. Data processing; Nervous system; Parkinson’s disease; Pattern analysis; Probability and statistics; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Radionuclide investigations; Risk; Risk prediction; Sciences and techniques of general use; Software; Statistics; Support vector machine; Support vector machines; Tomography; Logistic regression; SBR; PD; PPMI; Computer-aided early diagnosis; Risk prediction; SVM; Support vector machine; Pattern analysis; SPECT; Parkinson’s disease; Therapy; Automatic classification; Parkinson's disease; Parkinson disease; Modeling; Encephalon; Imaging; Database; Vector support machine; Diagnosis; Data analysis; Nervous system diseases; Statistical analysis; Probabilistic approach; Decision support system; Regression analysis; Marker; Single photon emission tomography; Forecasting; Single photon emission computed tomography; Computerized tomography; Kernel method; Radial basis function; Regression model; Medical imagery; Risk management
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2013.11.031

•Propose methods for very accurate classification of early PD using only 4 features.•Used public database which is large and diverse making the developed models robust.•First study to develop accurate prognostic model based on SBR features for early PD. Early and accurate diagnosis of Parkinson’s disease (PD) is important for early management, proper prognostication and for initiating neuroprotective therapies once they become available. Recent neuroimaging techniques such as dopaminergic imaging using single photon emission computed tomography (SPECT) with 123I-Ioflupane (DaTSCAN) have shown to detect even early stages of the disease. In this paper, we use the striatal binding ratio (SBR) values that are calculated from the 123I-Ioflupane SPECT scans (as obtained from the Parkinson’s progression markers initiative (PPMI) database) for developing automatic classification and prediction/prognostic models for early PD. We used support vector machine (SVM) and logistic regression in the model building process. We observe that the SVM classifier with RBF kernel produced a high accuracy of more than 96% in classifying subjects into early PD and healthy normal; and the logistic model for estimating the risk of PD also produced high degree of fitting with statistical significance indicating its usefulness in PD risk estimation. Hence, we infer that such models have the potential to aid the clinicians in the PD diagnostic process.