scholarly journals Fixel-Based Analysis and Free Water Corrected DTI Evaluation of HIV-Associated Neurocognitive Disorders

2021 ◽  
Vol 12 ◽  
Author(s):  
Alan Finkelstein ◽  
Abrar Faiyaz ◽  
Miriam T. Weber ◽  
Xing Qiu ◽  
Md Nasir Uddin ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cognitive Impairment ◽  
Cross Section ◽  
Free Water ◽  
Neurocognitive Disorders ◽  
Fiber Density ◽  
Consistent Finding ◽  
Machine Learning Classifiers ◽  
Learning Classifiers ◽  
Hiv Associated Neurocognitive Disorders

Background: White matter (WM) damage is a consistent finding in HIV-infected (HIV+) individuals. Previous studies have evaluated WM fiber tract-specific brain regions in HIV-associated neurocognitive disorders (HAND) using diffusion tensor imaging (DTI). However, DTI might lack an accurate biological interpretation, and the technique suffers from several limitations. Fixel-based analysis (FBA) and free water corrected DTI (fwcDTI) have recently emerged as useful techniques to quantify abnormalities in WM. Here, we sought to evaluate FBA and fwcDTI metrics between HIV+ and healthy controls (HIV−) individuals. Using machine learning classifiers, we compared the specificity of both FBA and fwcDTI metrics in their ability to distinguish between individuals with and without cognitive impairment in HIV+ individuals.Methods: Forty-two HIV+ and 52 HIV– participants underwent MRI exam, clinical, and neuropsychological assessments. FBA metrics included fiber density (FD), fiber bundle cross section (FC), and fiber density and cross section (FDC). We also obtained fwcDTI metrics such as fractional anisotropy (FAT) and mean diffusivity (MDT). Tract-based spatial statistics (TBSS) was performed on FAT and MDT. We evaluated the correlations between MRI metrics with cognitive performance and blood markers, such as neurofilament light chain (NfL), and Tau protein. Four different binary classifiers were used to show the specificity of the MRI metrics for classifying cognitive impairment in HIV+ individuals.Results: Whole-brain FBA showed significant reductions (up to 15%) in various fiber bundles, specifically the cerebral peduncle, posterior limb of internal capsule, middle cerebellar peduncle, and superior corona radiata. TBSS of fwcDTI metrics revealed decreased FAT in HIV+ individuals compared to HIV– individuals in areas consistent with those observed in FBA, but these were not significant. Machine learning classifiers were consistently better able to distinguish between cognitively normal patients and those with cognitive impairment when using fixel-based metrics as input features as compared to fwcDTI metrics.Conclusion: Our findings lend support that FBA may serve as a potential in vivo biomarker for evaluating and monitoring axonal degeneration in HIV+ patients at risk for neurocognitive impairment.

scholarly journals Fixel-Based Analysis and Free Water Corrected DTI Evaluation of HIV Associated Neurocognitive Disorders

2021 ◽  
Author(s):  
Alan Finkelstein ◽  
Abrar Faiyaz ◽  
Miriam T. Weber ◽  
Xing Qiu ◽  
Md Nasir Uddin ◽  
...  
Keyword(s):  
Cognitive Impairment ◽  
Cross Section ◽  
Diffusion Tensor ◽  
Free Water ◽  
Neurocognitive Disorders ◽  
Cerebral Peduncle ◽  
Fiber Density ◽  
Adaptive Boosting ◽  
Consistent Finding ◽  
Hiv Associated Neurocognitive Disorders

Background: White matter (WM) damage is a consistent finding in HIV infected (HIV+) individuals. Previous studies have evaluated WM fiber tract specific brain regions in HIV-associated neurocognitive disorders (HAND) using conventional diffusion tensor imaging (DTI). However, DTI might lack an accurate biological interpretation, and the technique suffers from several limitations. Fixel-based analysis (FBA) and free water corrected DTI (fwcDTI) have recently emerged as useful techniques to quantify abnormalities in WM. Here, we sought to evaluate FBA and fwcDTI metrics between HIV+ and HIV- individuals. Using machine learning classifiers, we compared the specificity of both FBA and fwcDTI metrics in their ability to distinguish between individuals with and without cognitive impairment. Methods: Forty-two HIV+ and 52 HIV- participants underwent MRI exam, clinical and neuropsychological assessments. FBA metrics included fiber density (FD), fiber bundle cross-section (FC), and fiber density and cross-section (FDC). We also obtained fwcDTI metrics such as fractional anisotropy (FAT) and mean diffusivity (MDT). Tract-based spatial statistics (TBSS) was performed on FAT and MDT. We evaluated the correlations between MRI metrics with cognitive performance and blood markers, such as neurofilament light chain NfL, and Tau protein. Statistical significance was evaluated at α = 0.05. Four different binary classifiers were used to show the specificity of the MRI metrics for classifying cognitive impairment in HIV+ individuals. Results: Whole brain FBA showed significant reductions (up to 15%) in various fiber bundles, specifically, the cerebral peduncle, posterior limb of the internal capsule, middle cerebellar peduncle and superior corona radiata. TBSS of fwcDTI metrics revealed decreased FAT (by 1-2%) in HIV+ individuals compared to HIV- individuals in areas consistent with those observed in FBA, but these were not significant. An adaptive boosting binary classifier reliably distinguished between cognitively normal patients and those with cognitive impairment with 80% precision and 78% recall. The mean area under the curve (AUC) was 0.805 ± 0.093 for the receiver operatic characteristic (ROC) curve, and 0.859 ± 0.133 for the precision-recall curve, using five-fold cross-validation. Conclusion: Our findings lend support that FBA may serve as a potential in vivo biomarker for evaluating and monitoring patients with HIV at risk for neurocognitive impairment, and emphasizes the importance of utilizing more complex diffusion models in evaluating HIV infection.

scholarly journals Assessing the Effect of Training Sampling Design on the Performance of Machine Learning Classifiers for Land Cover Mapping Using Multi-Temporal Remote Sensing Data and Google Earth Engine

2021 ◽  
Vol 13 (8) ◽  
pp. 1433
Author(s):  
Shobitha Shetty ◽  
Prasun Kumar Gupta ◽  
Mariana Belgiu ◽  
S. K. Srivastav
Keyword(s):  
Machine Learning ◽  
Remote Sensing ◽  
Random Sampling ◽  
Sampling Design ◽  
Remote Sensing Data ◽  
Google Earth ◽  
Machine Learning Classifiers ◽  
Learning Classifiers ◽  
Multi Temporal ◽  
Google Earth Engine

Machine learning classifiers are being increasingly used nowadays for Land Use and Land Cover (LULC) mapping from remote sensing images. However, arriving at the right choice of classifier requires understanding the main factors influencing their performance. The present study investigated firstly the effect of training sampling design on the classification results obtained by Random Forest (RF) classifier and, secondly, it compared its performance with other machine learning classifiers for LULC mapping using multi-temporal satellite remote sensing data and the Google Earth Engine (GEE) platform. We evaluated the impact of three sampling methods, namely Stratified Equal Random Sampling (SRS(Eq)), Stratified Proportional Random Sampling (SRS(Prop)), and Stratified Systematic Sampling (SSS) upon the classification results obtained by the RF trained LULC model. Our results showed that the SRS(Prop) method favors major classes while achieving good overall accuracy. The SRS(Eq) method provides good class-level accuracies, even for minority classes, whereas the SSS method performs well for areas with large intra-class variability. Toward evaluating the performance of machine learning classifiers, RF outperformed Classification and Regression Trees (CART), Support Vector Machine (SVM), and Relevance Vector Machine (RVM) with a >95% confidence level. The performance of CART and SVM classifiers were found to be similar. RVM achieved good classification results with a limited number of training samples.

scholarly journals A review of infant cry analysis and classification

2021 ◽  
Vol 2021 (1) ◽  
Author(s):  
Chunyan Ji ◽  
Thosini Bamunu Mudiyanselage ◽  
Yutong Gao ◽  
Yi Pan
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Signal Analysis ◽  
Future Research ◽  
Prosodic Features ◽  
Infant Cry ◽  
Machine Learning Classification ◽  
Machine Learning Classifiers ◽  
Learning Classifiers ◽  
Processing Techniques

AbstractThis paper reviews recent research works in infant cry signal analysis and classification tasks. A broad range of literatures are reviewed mainly from the aspects of data acquisition, cross domain signal processing techniques, and machine learning classification methods. We introduce pre-processing approaches and describe a diversity of features such as MFCC, spectrogram, and fundamental frequency, etc. Both acoustic features and prosodic features extracted from different domains can discriminate frame-based signals from one another and can be used to train machine learning classifiers. Together with traditional machine learning classifiers such as KNN, SVM, and GMM, newly developed neural network architectures such as CNN and RNN are applied in infant cry research. We present some significant experimental results on pathological cry identification, cry reason classification, and cry sound detection with some typical databases. This survey systematically studies the previous research in all relevant areas of infant cry and provides an insight on the current cutting-edge works in infant cry signal analysis and classification. We also propose future research directions in data processing, feature extraction, and neural network classification fields to better understand, interpret, and process infant cry signals.

Sign in / Sign up

Export Citation Format

Share Document