Elastic Net Logistic Regression Research Articles

Assessment of Circulating Protein Signatures for Kidney Transplantation in Pediatric Recipients.

Identification and use of non-invasive biomarkers for kidney transplantation monitoring is an unmet need. A total of 121 biobanked sera collected from 111 unique kidney transplant (KT) patients (children and adolescent) and 10 age-matched healthy normal controls were used to profile serum proteins using semi-quantitative proteomics. The proteomics data were analyzed to identify panels of serum proteins that were specific to various transplant injuries, which included acute rejection (AR), BK virus nephropathy (BKVN), and chronic allograft nephropathy (CAN). Gene expression data from matching peripheral blood mononuclear cells were interrogated to investigate the association between soluble serum proteins and altered gene expression of corresponding genes in different injury phenotypes. Analysis of the proteomics data identified from different patient phenotypes, with criteria of false discovery rate <0.05 and at least twofold changes in either direction, resulted in a list of 10 proteins that distinguished KT injury from no injury. Similar analyses to identify proteins specific to chronic injury, acute injury, and AR after kidney transplantation identified 22, 6, and 10 proteins, respectively. Elastic-Net logistic regression method was applied on the 137 serum proteins to classify different transplant injuries. This algorithm has identified panels of 10 serum proteins specific for AR, BKVN, and CAN with classification rates 93, 93, and 95%, respectively. The identified proteins could prove to be potential surrogate biomarkers for routine monitoring of the injury status of pediatric KT patients.

Widespread structural brain involvement in ALS is not limited to the C9orf72 repeat expansion

BackgroundIn patients with a C9orf72 repeat expansion (C9+), a neuroimaging phenotype with widespread structural cerebral changes has been found. We aimed to investigate the specificity of this neuroimaging phenotype in...

COPD Exacerbation Biomarkers Validated Using Multiple Reaction Monitoring Mass Spectrometry.

BackgroundAcute exacerbations of chronic obstructive pulmonary disease (AECOPD) result in considerable morbidity and mortality. However, there are no objective biomarkers to diagnose AECOPD.MethodsWe used multiple reaction monitoring mass spectrometry to quantify 129 distinct proteins in plasma samples from patients with COPD. This analytical approach was first performed in a biomarker cohort of patients hospitalized with AECOPD (Cohort A, n = 72). Proteins differentially expressed between AECOPD and convalescent states were chosen using a false discovery rate <0.01 and fold change >1.2. Protein selection and classifier building were performed using an elastic net logistic regression model. The performance of the biomarker panel was then tested in two independent AECOPD cohorts (Cohort B, n = 37, and Cohort C, n = 109) using leave-pair-out cross-validation methods.ResultsFive proteins were identified distinguishing AECOPD and convalescent states in Cohort A. Biomarker scores derived from this model were significantly higher during AECOPD than in the convalescent state in the discovery cohort (p<0.001). The receiver operating characteristic cross-validation area under the curve (CV-AUC) statistic was 0.73 in Cohort A, while in the replication cohorts the CV-AUC was 0.77 for Cohort B and 0.79 for Cohort C.ConclusionsA panel of five biomarkers shows promise in distinguishing AECOPD from convalescence and may provide the basis for a clinical blood test to diagnose AECOPD. Further validation in larger cohorts is necessary for future clinical translation.

SP 1. Parkinsonian rest tremor can be detected based on frequency domain features of the cerebral tremor network

Recent research has revealed several features of electrophysiological signals, which change upon manifestation of Parkinsonian rest tremor. These observations spur hope that the onset and the presence of tremor can be inferred by online analysis of brain signals, enabling demand-driven deep brain stimulation for tremor. It is an open question, however, whether it is possible to detect the presence of tremor given only a short data epoch. Furthermore, it is unclear which signals need to be considered for this task. In this study, we re-analyze data from 18 subthalamic nuclei (STN) of 12 tremor-dominant Parkinson patients in whom magnetoencephalography (MEG) and STN local field potentials (LFPs) were recorded simultaneously. Patients rested while tremor waxed and waned spontaneously. Epochs were attributed either to the tremor condition or to the tremor-free condition based on inspection of forearm electromyograms. MEG source activity was extracted from the cerebral tremor network proposed by Timmermann et al. (2003). Data were segmented into 4s epochs and power and connectivity features were computed for each epoch. These features were used as input to elastic net logistic regression to decode the tremor condition. Feature importance was evaluated through analysis of activation patterns and re-classification with single features or a subset of features. When testing and training was performed in different subjects, the decoder accuracy was 70%. Accuracy could be substantially improved by performing cross-validated training within subjects (81 ± 12%). An analysis of feature importance revealed that power features were much more informative than connectivity features. Among others, STN and primary motor cortex beta power as well as thalamic gamma power were found to be particularly discriminative. Considering only STN power features yielded an accuracy of 77 ± 12% in within-subject classification, which was increased to 86 ± 11% when adding MEG power features. A similar accuracy could be obtained when using MEG power features only (85 ± 12%). We conclude that Parkinsonian rest tremor can be distinguished from tremor-free periods with low error rate, provided that training data are available for each subject. Furthermore, connectivity features contribute little to tremor detection on a timescale of a few seconds. Nevertheless, it is of advantage to record cortical, cerebellar and thalamic sources in addition to the STN, as their power features provide additional information. Future research will integrate the current and related findings to design an algorithm for online stimulation control, incorporating estimates of the tremor state and patient-specific parameters set by the clinician.

Combining multiple anatomical MRI measures improves Alzheimer's disease classification.

Several anatomical MRI markers for Alzheimer's disease (AD) have been identified. Hippocampal volume, cortical thickness, and grey matter density have been used successfully to discriminate AD patients from controls. These anatomical MRI measures have so far mainly been used separately. The full potential of anatomical MRI scans for AD diagnosis might thus not yet have been used optimally. In this study, we therefore combined multiple anatomical MRI measures to improve diagnostic classification of AD. For 21 clinically diagnosed AD patients and 21 cognitively normal controls, we calculated (i) cortical thickness, (ii) cortical area, (iii) cortical curvature, (iv) grey matter density, (v) subcortical volumes, and (vi) hippocampal shape. These six measures were used separately and combined as predictors in an elastic net logistic regression. We made receiver operating curve plots and calculated the area under the curve (AUC) to determine classification performance. AUC values for the single measures ranged from 0.67 (cortical thickness) to 0.94 (grey matter density). The combination of all six measures resulted in an AUC of 0.98. Our results demonstrate that the different anatomical MRI measures contain complementary information. A combination of these measures may therefore improve accuracy of AD diagnosis in clinical practice. Hum Brain Mapp 37:1920-1929, 2016. © 2016 Wiley Periodicals, Inc.

単極脳波信号を用いた子どもの集中状態判別

SUMMARY In this paper, we aim to classify two classes in children by using single-channel electroencephalogram (EEG). EEG has been used to define neural patterns and to adjust the wide applicability to a larger population of healthy and diseased users. Specialized EEG devices have recently developed as for compact and portable measurement system using them in the real environment. If there is a multiplex state estimation system with EEG through a specialized EEG device, it would be a powerful tool for neuroscience studies and clinical applications. We first focused on the state of concentration; therefore, two kinds of single-channel EEG signals (during meditation and concentration) from 10 children were measured. Recordings were processed to remove artifacts, and then extracted their periodic or nonperiodic features by three methods (Fourier transform, wavelet transform, and empirical mode decomposition). Elastic net logistic regression constructed predictive models to classify two classes of the optimized extracted features. A model showed 0.988 area under the receiver-operating characteristic curve when wavelet transform was selected as feature extraction method. We next construct a multiplex state estimation system. Finally, we will make portable applications using a specialized EEG device that include the multiplex model and encourage children to develop the child's sense.

Malware Detection Using Nonparametric Bayesian Clustering and Classification Techniques

Computer security requires statistical methods to quickly and accurately flag malicious programs. This article proposes a nonparametric Bayesian approach for classifying programs as benign or malicious and simultaneously clustering malicious programs. The analysis is based on the dynamic trace (DT) of instructions under the first-order Markov assumption. Each row of the trace’s transition matrix is modeled using the Dirichlet process mixture (DPM) model. The DPM model clusters programs within each class (malicious or benign), and produces the posterior probability of being a malware which is used for classification. The novelty of the model is using this clustering algorithm to improve the classification accuracy. The simulation study shows that the DPM model outperforms the elastic net logistic (ENL) regression and the support vector machine (SVM) in classification performance under most of the scenarios, and also outperforms the spectral clustering method for grouping similar malware. In an analysis of real malicious and benign programs, the DPM model gives significantly better classification performance than the ENL model, and competitive results to the SVM. More importantly, the DPM model identifies clusters of programs during the classification procedure which is useful for reverse engineering.

Impact of Present-on-admission Indicators on Risk-adjusted Hospital Mortality Measurement

Benchmarking performance across hospitals requires proper adjustment for differences in baseline patient and procedural risk. Recently, a Risk Stratification Index was developed from Medicare data, which used all diagnosis and procedure codes associated with each stay, but did not distinguish present-on-admission (POA) diagnoses from hospital-acquired diagnoses. We sought to (1) develop and validate a risk index for in-hospital mortality using only POA diagnoses, principal procedures, and secondary procedures occurring before the date of the principal procedure (POARisk) and (2) compare hospital performance metrics obtained using the POARisk model with those obtained using a similarly derived model which ignored the timing of diagnoses and procedures (AllCodeRisk). We used the 2004-2009 California State Inpatient Database to develop, calibrate, and prospectively test our models (n = 24 million). Elastic net logistic regression was used to estimate the two risk indices. Agreement in hospital performance under the two respective risk models was assessed by comparing observed-to-expected mortality ratios; acceptable agreement was predefined as the AllCodeRisk-based observed-to-expected ratio within ± 20% of the POARisk-based observed-to-expected ratio for more than 95% of hospitals. After recalibration, goodness of fit (i.e., model calibration) within the 2009 data was excellent for both models. C-statistics were 0.958 and 0.981, respectively, for the POARisk and AllCodeRisk models. The AllCodeRisk-based observed-to-expected ratio was within ± 20% of the POARisk-based observed-to-expected ratio for 89% of hospitals, which was slightly lower than the predefined limit of agreement. Consideration of POA coding meaningfully improved hospital performance measurement. The POARisk model should be used for risk adjustment when POA data are available.

Identifying Neuroimaging and Proteomic Biomarkers for MCI and AD via the Elastic Net.

Multi-modal neuroimaging and biomarker data provide exciting opportunities to enhance our understanding of phenotypic characteristics associated with complex disorders. This study focuses on integrative analysis of structural MRI data and proteomic data from an RBM panel to examine their predictive power and identify relevant biomarkers in a large MCI/AD cohort. MRI data included volume and thickness measures of 98 regions estimated by FreeSurfer. RBM data included 146 proteomic analytes extracted from plasma and serum. A sparse learning model, elastic net logistic regression, was proposed to classify AD and MCI, and select disease-relevant biomarkers. A linear support vector machine coupled with feature selection was employed for comparison. Combining RBM and MRI data yielded improved prediction rates: HC vs AD (91.9%), HC vs MCI (90.5%) and MCI vs AD (86.5%). Elastic net identified a small set of meaningful imaging and proteomic biomarkers. The elastic net has great power to optimize the sparsity of feature selection while maintaining high predictive power. Its application to multi-modal imaging and biomarker data has considerable potential for discovering biomarkers and enhancing mechanistic understanding of AD and MCI.