Quantitative prediction of acute ischemic tissue fate using support vector machine

Research Imaging Institute, University of Texas Health Science Center, San Antonio, TX 78229, USA.
Brain research (Impact Factor: 2.84). 08/2011; 1405:77-84. DOI: 10.1016/j.brainres.2011.05.066
Source: PubMed


Accurate and quantitative prediction of ischemic tissue fate could improve decision-making in the clinical treatment of acute stroke. The goal of the present study is to explore the novel use of support vector machine (SVM) to predict infarct on a pixel-by-pixel basis using only acute cerebral blood flow (CBF), apparent diffusion coefficient (ADC) MRI data. The efficacy of SVM prediction model was tested on three stroke groups: 30-min, 60-min, and permanent middle cerebral-artery occlusion (n=12 rats for each group). CBF, ADC and relaxation time constant (T2) were acquired during the acute phase up to 3h and again at 24h. Infarct was predicted using only acute (30-min) stroke data. Receiver-operating characteristic (ROC) analysis was used to quantify prediction accuracy. The areas under the receiver-operating curves were 86±2.7%, 89±1.4%, and 93±0.8% using ADC+CBF data for the 30-min, 60-min and permanent middle cerebral artery occlusion (MCAO) group, respectively. Adding neighboring pixel information and spatial infarction incidence improved performance to 88±2.8%, 94±0.8%, and 97±0.9%, respectively. SVM prediction compares favorably to a previously published artificial neural network (ANN) prediction algorithm operated on the same data sets. SVM prediction model has the potential to provide quantitative frameworks to aid clinical decision-making in the treatment of acute stroke.

Download full-text


Available from: Qiang Shen, Mar 17, 2014
20 Reads
  • Source
    • "A hyper-plane can be written as the set of points X satisfying WTX + b = 0, where WT is a normal vector perpendicular to the hyper-plane, X is the vector of electrocardiographic features, and b is bias or offset of the hyper-plane from the origin. Inputs of SVM are mapped onto a high dimensional feature space via kernel functions, and the optimal hyper-planes are constructed to separate samples into two classes [16]. SVMs were trained and tested using the leave-one-out and cross-validation methods, which were applied to evaluate the accuracy of classification. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Previous studies indicate that decreased heart-rate variability (HRV) is related to the risk of death in patients after acute myocardial infarction (AMI). However, the conventional indices of HRV have poor predictive value for mortality. Our aim was to develop novel predictive models based on support vector machine (SVM) to study the integrated features of HRV for improving risk stratification after AMI. Methods A series of heart-rate dynamic parameters from 208 patients were analyzed after a mean follow-up time of 28 months. Patient electrocardiographic data were classified as either survivals or cardiac deaths. SVM models were established based on different combinations of heart-rate dynamic variables and compared to left ventricular ejection fraction (LVEF), standard deviation of normal-to-normal intervals (SDNN) and deceleration capacity (DC) of heart rate. We tested the accuracy of predictors by assessing the area under the receiver-operator characteristics curve (AUC). Results We evaluated a SVM algorithm that integrated various electrocardiographic features based on three models: (A) HRV complex; (B) 6 dimension vector; and (C) 8 dimension vector. Mean AUC of HRV complex was 0.8902, 0.8880 for 6 dimension vector and 0.8579 for 8 dimension vector, compared with 0.7424 for LVEF, 0.7932 for SDNN and 0.7399 for DC. Conclusions HRV complex yielded the largest AUC and is the best classifier for predicting cardiac death after AMI.
    BMC Cardiovascular Disorders 05/2014; 14(1):59. DOI:10.1186/1471-2261-14-59 · 1.88 Impact Factor
  • Source
    • "Incorporating additional imaging data (such as CBF) could further improve prediction accuracy. Finally, although performance was evaluated by sensitivity and specificity calculations, future studies will utilize more sophisticated algorithms, such as support vector machine with separate training and experimental groups, to quantitatively predict tissue fate (Huang et al., 2011). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Algorithms to predict ischemic tissue fate based on acute stroke MRI typically utilized data at a single time point. The goal of this study was to investigate the potential improvement in prediction accuracy when incorporating MRI diffusion data from multiple time points during acute phase to improve prediction accuracy. This study was carried out using MRI data from rats subjected to permanent, 60-min and 30-min of middle cerebral artery occlusion (MCAO). The sensitivity and specificity of prediction accuracy were calculated. In the permanent MCAO group, prediction with multiple time-point diffusion data improved sensitivity and specificity compared with prediction using a single time point. In the 60-min MCAO group, multiple time-point analysis improved specificity but decreased sensitivity compared to the single time-point analysis. In the 30-min MCAO group, multiple time-point analysis showed no statistically significant improvement in specificity and sensitivity compared with the single time point analysis. This is because reperfusion transiently or permanently reversed the decline in ADC values, resulting in increased uncertainty and thus decreased prediction accuracy. Incorporating this a priori information could further improve prediction accuracy in the reperfusion group. These findings suggest that incorporating MRI data from multiple time points could improve prediction accuracy under certain ischemic conditions.
    Brain research 04/2012; 1458:86-92. DOI:10.1016/j.brainres.2012.04.004 · 2.84 Impact Factor
  • Source
    • "While the OC percent changes may be biased by the low basal T 2 * -weighted signal, it may be more sensitive. Additional studies are needed to determine which analysis method will provide more accurate prediction of final infarct volume (Huang et al., 2011). With improved MRI sensitivity, the ability of basal T 2 * or T 2 weighted MRI to identify salvageable and non-salvageable tissue may be warrant. "
    [Show abstract] [Hide abstract]
    ABSTRACT: It has been recently shown that at-risk tissue exhibits exaggerated T(2)⁎-weighted MRI signal increases during transient oxygen challenge (OC), suggesting that the tissue is still metabolically active. This study further characterized the effects of transient OC on T(2)⁎-weighted MRI in permanent focal stroke rats (N=8) using additional quantitative measures. The major findings were: i) the ischemic core cluster showed no significant response, whereas the mismatch cluster showed markedly higher percent changes relative to normal tissue in the acute phase. ii) Many of the mismatch pixels showed exaggerated OC responses which became hyperintense on T(2)-weighted MRI at 24h. The area with exaggerated OC responses was larger than the mismatch, suggesting that some tissue with reduced diffusion were potentially at risk. iii) Basal T(2)⁎-weighted intensities on the perfusion-diffusion contourplot were high in normal tissue and low in the core, with a sharp transition in the mismatch. iv) OC-induced changes on the perfusion-diffusion contourplot dropped as perfusion and diffusion values fell below their respective viability thresholds. v) Basal T(1) increased slightly in the ischemic core (P<0.05). OC decreased T(1) in normal (P<0.05) but not in mismatch and core pixels. vi) OC decreased CBF in normal (P<0.05) but not in mismatch and core pixels. T(2)⁎-weighted MRI of OC has the potential to offer unique clinically relevant data.
    Brain research 11/2011; 1425:132-41. DOI:10.1016/j.brainres.2011.09.052 · 2.84 Impact Factor
Show more