Automatic classification of 123I-FP-CIT (DaTSCAN) SPECT images

Abstract

The combination of data reduction by SVD with automatic classifiers such as NB can provide good diagnostic accuracy and may be a useful adjunct to clinical reporting.

High-Accuracy Detection of Early Parkinson's Disease through Multimodal Features and Machine Learning

“…There are studies which use these features to classify subjects as early PD and healthy normal [11,[14][15][16][17][18][19][20][21][22]. [14] performed smell identification tests using culturally adapted translations of the University of Pennsylvania Smell Identification Test (UPSIT) and Sniffin' Sticks (SS) test in 106 PD and 118 normal subjects.…”

“…[11] used Cerebrospinal Fluid (CSF) measurements from 63 early PD and 39 healthy normal and observed that these measures were statistically significant but showed a low diagnostic utility as the area under the ROC (AUC) was less than 0.8 for PD diagnosis. [17] used 123I-Ioflupane (DaTSCAN TM , GE Healthcare; also known as [ 123 I]FP-CIT) SPECT scan data from 79 patients with parkinsonism (PS) and 37 non-PS subjects (which is not the healthy normal group and instead it represent subjects with non-PS conditions like essential tremor that shows normal SPECT scans), and obtained a classification accuracy of 94.8% with 93.7% sensitivity and 97.3% specificity using Naïve Bayes classifier. [18] also used SPECT scan data from 95 PD and 94 normal subjects and obtained a maximum accuracy of 94.7%, sensitivity 93.7% and specificity 95.7% using an approach based on partial least squares and support vector machine (SVM).…”

Texture Features Based Detection of Parkinson’s Disease on DaTSCAN Images

“…A wide range of supervised learning techniques have been combined to generate Computer Aided Diagnosis (CAD) systems that allow to detect neurodegenerative diseases, such as Alzheimer's Disease [9] or Parkinson [16]. Techniques range from the use of selection of Regions of Interest (ROIs) [5], or Single Value Decomposition strategies (SVD) [14] to more complex approaches such as Empirical Mode Decomposition (EMD) combined with Principal Component Analysis (PCA) combined method in [13].…”

Clinical utility of DaTscan™ (123I-Ioflupane Injection) in the diagnosis of Parkinsonian Syndromes

“…44 From a methodology point of view, images of SPECT with 123I-Ioflupane can be quantified through several methods such as a region of interest (ROI), one-shape analysis of the uptake, Statistical Parametric Mapping (SPM) analysis based on ROI/background ratios, and principal component analysis. 59 Others are based on the machine learning paradigm that can be applied to the SPECT images through computeraided diagnosis systems that can assist the diagnosis of PSs. 60 An interesting method for quantitative measure of uptake has been proposed to assist the diagnosis of uncertain PSs and possibly measure disease progression 59 .…”

“…59 Others are based on the machine learning paradigm that can be applied to the SPECT images through computeraided diagnosis systems that can assist the diagnosis of PSs. 60 An interesting method for quantitative measure of uptake has been proposed to assist the diagnosis of uncertain PSs and possibly measure disease progression 59 . 54.191.190.102 on 12-May-2018 For personal use only.…”