Absolute Pearson Correlation Research Articles

Integrative interpretation of cardiopulmonary exercise tests for cardiovascular risk stratification: a machine learning approach

Abstract Background and objectives Cardiopulmonary exercise testing (CPET) remains underutilized for cardiovascular (CV) risk assessment. Integrative interpretation of CPET results may improve characterization of cardiorespiratory fitness and assessment of CV risk. Therefore, we explored the clinical value of CPET-based phenomapping for CV risk stratification. Methods We retrospectively retrieved clinical data from 2280 patients with diverse CV risk (47.9% female) who underwent maximal CPET by cycle ergometry. We derived 18 key CPET indices as well as data on fatal and non-fatal CV events (median follow-up time: 5.3 years). Next, an unsupervised clustering algorithm (Gaussian Mixture modelling) subdivided the cohort in sex-specific phenogroups solely based on differences in CPET metrics. Clinical characteristics were compared across CPET phenogroups and their multivariable-adjusted association with future CV events was determined. Results 8 of the 18 CPET metrics were excluded from clustering due to high collinearity (i.e. an absolute Pearson correlation coefficient >0.8). Based on the remaining 10 CPET variables, the clustering algorithm subdivided the cohort in four male and four female CPET phenogroups. In both males and females, the phenogroups differed significantly in age, BMI, spirometric and CPET measurements and prevalence of hypertension, intake of antihypertensive and lipid-lowering drugs, CV disease and CV surgery (P<0.05 for trends). During follow-up, 278 males and 109 females experienced a CV event (43.9 and 16.2 events per 1000 years of follow-up, respectively). Compared to male phenogroup 1, male phenogroups 3 and 4 presented higher risk for incident CV events (multivariable-adjusted hazard ratio: 1.97 and 1.62; P≤0.027). Significant differences in risk for future CV events between the female phenogroups disappeared after adjustment for confounders. Conclusion Integrative CPET-based phenogrouping adequately stratified patients according to their risk for future CV events. Such phenomapping may facilitate comprehensive evaluation of CPET results and steer future CV risk management.Study flow and findings

A Novel Quality Control Algorithm for Medical Image Segmentation Based on Fuzzy Uncertainty

Deep learning methods have achieved excellent performance in medical image segmentation. However, the practical application of deep learning based segmentation models is limited in clinical settings due to the lack of reliable information about the segmentation quality. In this paper, we propose a novel quality control algorithm based on fuzzy uncertainty to quantify the quality of the predicted segmentation results as part of the model inference process. Firstly, test-time augmentation and Monte Carlo dropout are applied simultaneously to capture both the data and model uncertainties of the trained image segmentation model. Then a fuzzy set is generated to describe the captured uncertainty with the assistance of the linear Euclidean distance transform algorithm. Finally, the fuzziness of the generated fuzzy set is adopted to calculate an image-level segmentation uncertainty and therefore to infer the segmentation quality. Extensive experiments using five medical image segmentation applications on the detection of skin lesion, nuclei, lung, breast and cell are conducted to evaluate the proposed algorithm. The experimental results show that the estimated image-level uncertainties using the proposed method have strong correlations with the segmentation qualities measured by the Dice coefficient, resulting in absolute Pearson correlation coefficients of 0.60-0.92. Our method outperforms other five state-of-the-art quality control methods in classifying the segmentation results into good and poor quality groups (area under the ROC of greater than 0.92, while other methods are below 0.85). The code is available on Github ( <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/qiaolin1992/Fuzzy-Uncertainty</uri> ).

Evaluating parallelized support vector regression and nearest neighbor regression with different input variations for estimating daily global solar radiation of the humid subtropical region in China

AbstractIndirectly estimating global solar radiation is strongly nonlinear and needs to be addressed by machine learning. Sequentially developing a machine learning model can be very time consuming. Moreover, whether and how the exogenous meteorological, geographical and temporal variables affect regression accuracy still has not been well understood. This paper evaluates parallelized support vector regression (SVR) and nearest neighbor regression (NNR) models for estimating daily global solar radiation of the humid subtropical region in China using existing Python libraries on a multi-core central processing unit (CPU) and a graphical processing unit (GPU). Seven input variations are studied. Two variations are commonly adopted in literature, four variations contain meteorological, geographical and/or temporal features with bounded Pearson correlation coefficients (PCCs), and the other variation simply include all the available features. Experimental results demonstrate that: SVR and NNR are equally powerful for nonlinear regression, and the variation comprising features with absolute PCCs no less than 0.3 (i.e. just all the meteorological features) is able to achieve most accurate estimation; the GPU-parallelized SVR model can accelerate parameter calibration and prediction; compared with the CPU-parallelized and GPU-parallelized SVR models, the GPU-parallelized NNR model is much more efficient and rather more scalable with the increment of the number of data samples; and the CPU-parallelized NNR model consumes quite less parameter calibration time than the GPU-parallelized NNR model, owing to different methods adopted for determining distances and significant time wasted by the GPU-parallelized NNR model on repeatedly calculating required information during cross-validation.

An Adaptive Channel Selection and Graph ResNet Based Algorithm for Motor Imagery Classification

In Brain-Computer interface (BCI) applications, achieving accurate control relies heavily on the classification accuracy and efficiency of motor imagery electroencephalogram (EEG) signals. However, factors such as mutual interference between multi-channel signals, inter-individual variability, and noise interference in the channels pose challenges to motor imagery EEG signal classification. To address these problems, this paper proposes an Adaptive Channel Selection algorithm aimed at optimizing classification accuracy and Information Translate Rate (ITR). First, C3, C4, and Cz are selected as key channels based on neurophysiological evidence and extensive experimental studies. Next, the channel selection is fine-tuned using spatial location and absolute Pearson correlation coefficients. By analyzing the relationship between EEG channels and key channels, the most relevant channel combination is determined for each subject, reducing confounding information and improving classification accuracy. To validate the method, the SHU Dataset and the PhysioNet Dataset are used in experiments. The Graph ResNet classification model is employed to extract features from the selected channel combinations using deep learning techniques. Experimental results show that the average classification accuracy is improved by 5.36% and 9.19%, and the Information Translate Rate is improved by 29.24% and 26.75%, respectively, compared to a single channel combination.

Estimating Coffee Plant Yield Based on Multispectral Images and Machine Learning Models

The coffee plant is one of the main crops grown in Brazil. However, strategies to estimate its yield are questionable given the characteristics of this crop; in this context, robust techniques, such as those based on machine learning, may be an alternative. Thus, the aim of the present study was to estimate the yield of a coffee crop using multispectral images and machine learning algorithms. Yield data from a same study area in 2017, 2018 and 2019, Sentinel 2 images, Random Forest (RF) algorithms, Support Vector Machine (SVM), Neural Network (NN) and Linear Regression (LR) were used. Statistical analysis was performed to assess the absolute Pearson correlation and coefficient of determination values. The Sentinel 2 satellite images proved to be favorable in estimating coffee yield. Despite the low spatial resolution in estimating agricultural variables below the canopy, the presence of specific bands such as the red edge, mid infrared and the derived vegetation indices, act as a countermeasure. The results show that the blue band and green normalized difference vegetation index (GNDVI) exhibit greater correlation with yield. The NN algorithm performed best and was capable of estimating yield with 23% RMSE, 20% MAPE and R² 0.82 using 85% of the training and 15% of the validation data of the algorithm. The NN algorithm was also more accurate (27% RMSE) in predicting yield.

Dimensional characteristics of the major types of amphibole mineral particles and the implications for carcinogenic risk assessment

Context: Though some significant advances have been made in recent decades to evaluate the importance of size and morphology (habit) of elongate mineral particles (EMPs), further research is needed to better understand the role of each dimensional metric in determining the levels of cancer potency. Objective: To determine dimensional parameters most relevant for predicting cancer potency of durable elongate particles, specifically amphibole and durable silicate minerals generally. Methods: A database on dimensional and other relevant characteristics of elongate amphibole mineral particles was created, containing particle-by-particle information for 128 099 particles. Integral statistical characteristics on dimensionality of various amphibole types and morphological habits of EMPs were calculated, compared, and correlated with published mesothelioma and lung cancer potency factors. Results: The highest absolute Pearson correlation (r = 0.97, r 2 = 0.94, p < 0.05) was achieved between mesothelioma potency (RM) and specific surface area. The highest correlation with adjusted lung cancer potency was found with particle aspect ratio (AR) (r = 0.80, r 2 = 0.64, p < 0.05). Cluster analysis demonstrates that fractions of thin fibers (width less than 0.15 and 0.25 µm) also closely relate both to lung cancer and RM. Asbestiform and non-asbestiform populations of amphiboles significantly differ by dimensionality and carcinogenic potency. Conclusions: Dimensional parameters and morphological habits of EMPs are the main drivers for the observable difference in cancer potency among amphibole populations.

A Hybrid Leaf Area Index Estimation Method of Dioscorea Polystachya Turczaninow Using Sentinel-2 Vegetation Indices

<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Dioscorea polystachya</i> Turczaninow is an herbaceous vine plant distributed in China, and its rhizome named Chinese yam is a famous traditional Chinese medicine for treating diabetes and other diseases. However, a large region monitoring method of its growth status is lacking, which is important for Chinese yam yield estimation. Therefore, this study proposed a leaf area index (LAI) estimation algorithm for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Dioscorea polystachya</i> Turczaninow using a hybrid method and Sentinel-2 vegetation indices. First, two feature selection algorithms the gradient boosting regression tree (GBRT) and absolute Pearson correlation coefficient (APCC), were combined with field-measured data and radiation transfer model simulated data to generate four different feature important ranking groups. Then, a hybrid feature selection algorithm was used to determine the best feature subsets under each ranking group, and GBRT regression and least absolute shrinkage and selection operator (LASSO) were used to develop the LAI estimation models. Finally, the best LAI estimation model for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Dioscorea polystachya</i> Turczaninow was determined based on validation accuracy. The results indicated that the field-measured data were more reliable than the simulated data for feature selection, and the best LAI estimation model was the LASSO model using nine selected vegetation indices, which achieved the performance with RMSE of 0.391 and MAE of 0.310. The proposed method could provide real-time LAI estimates for future Chinese yam yield prediction.

Deep learning-based computer-aided heart sound analysis in children with left-to-right shunt congenital heart disease

ObjectiveThe purpose of this study was to explore a new algorithm model capable of leverage deep learning to screen and diagnose specific types of left-to-right shunt congenital heart disease (CHD) in children. MethodsUsing deep learning, screening models were constructed to identify 884 heart sound recordings from children with left-to-right shunt CHD. The most suitable model for each type was summarized and compared with expert auscultation. An exploratory analysis was conducted to assess whether there were correlations between heart sounds and left ventricular ejection fraction (LVEF), pulmonary artery pressure, and malformation size. ResultsThe residual convolution recurrent neural network (RCRnet) classification model had higher accuracy than other models with respect to atrial septal defect (ASD), ventricular septum defect (VSD), patent ductus arteriosus (PDA) and combined CHD, and the best auscultation sites were determined to be the 4th, 5th, 2nd and 3rd auscultation areas, respectively. The diagnostic results of this model were better than those derived from expert auscultation, with sensitivity values of 0.932–1.000, specificity values of 0.944–0.997, precision values of 0.888–0.997 and accuracy values of 0.940–0.994. Absolute Pearson correlation coefficient values between heart sounds of the four types of CHD and LVEF, right ventricular systolic pressure (RVSP) and malformation size were all less than 0.3. ConclusionsThe RCRnet model can preliminarily determine types of left-to-right shunt CHD and improve diagnostic efficiency, which may provide a new choice algorithmic CHD screening in children.

Audio Enhancement of Physical Models of Musical Instruments Using Optimal Correction Factors: The Recorder Case

A simulation of a musical instrument is considered to be a successful one when there is a good resemblance between the model’s synthesized sound and the real instrument’s sound. In this work, we propose the integration of physical modeling (PM) methods with an optimization process to regulate a generated digital signal. Its goal is to find a new set of values of the PM’s parameters’ that would lead to a synthesized signal matching as much as possible to reference signals corresponding to the physical musical instrument. The reference signals can be: (a) described by their acoustic characteristics (e.g., fundamental frequencies, inharmonicity, etc.) and/or (b) the signals themselves (e.g., impedances, recordings, etc.). We put this method into practice for a commercial recorder, simulated using the digital waveguides’ PM technique. The reference signals, in our case, are the recorded signals of the physical instrument. The degree of similarity between the synthesized (PM) and the recorded signal (musical instrument) is calculated by the signals’ linear cross-correlation. Our results show that the adoption of the optimization process resulted in more realistic synthesized signals by (a) enhancing the degree of similarity between the synthesized and the recorded signal (the average absolute Pearson Correlation Coefficient increased from 0.13 to 0.67), (b) resolving mistuning issues (the average absolute deviation of the synthesized from the recorded signals’ pitches reduced from 40 cents to the non-noticeable level of 2 cents) and (c) similar sound color characteristics and matched overtones (the average absolute deviation of the synthesized from the recorded signals’ first five partials reduced from 41 cents to 2 cents).

Region- and degeneration dependent stiffness distribution in intervertebral discs derived by shear wave elastography

Information on the local stiffness characteristics of the intervertebral disc (IVD) is crucial for the understanding of its structure-function properties in health and disease and may improve numerical modeling. Previous studies have attempted to map local tissue stiffness by sectioning the disc and performing mechanical testing on these discrete tissue units, which is technically challenging and may bias the results. Shear wave elastography (SWE) represents a nondestructive alternative that can provide spatially continuous elasticity estimates. We investigated the feasibility of SWE for human intervertebral disc elasticity mapping in a laboratory setting.To this end, global spinal segment mechanical behavior was determined in 6 loading directions and served as ground truth data for the validation of the approach. Subsequently, the cranial spinal vertebra was removed and shear wave elastographic scans of the IVD were acquired. SWE-measurements were reconstructed into three-dimensional elastographic maps, discretized into distinct IVD regions and correlated with global segment mechanical parameters. SWE-derived Young’s modulus estimates were compared among different regions and as a function of their state of degeneration.We found annulus shear wave speed to be moderately correlated with segment mechanical behavior irrespective of the loading direction whereas shear wave speed in the nucleus pulposus showed a very weak association (mean (SD) absolute Pearson correlation coefficients: 0.51 (0.14) and 0.17 (0.12), respectively). Young’s modulus mapping of the intervertebral disc revealed stiffness to be highest in the ventral annulus with a stiffness decrease both circumferentially towards the dorsal aspect as well as towards the center of the disc. SWE hence provides a valid alternative to disc sectioning and piecewise mechanical testing.

Classification of Adolescent Major Depressive Disorder Via Static and Dynamic Connectivity.

This paper introduces an approach for classifying adolescents suffering from MDD using resting-state fMRI. Accurate diagnosis of MDD involves interviews with adolescent patients and their parents, symptom rating scales based on Diagnostic and Statistical Manual of Mental Disorders (DSM), behavioral observation as well as the experience of a clinician. Discovering predictive biomarkers for diagnosing MDD patients using functional magnetic resonance imaging (fMRI) scans can assist the clinicians in their diagnostic assessments. This paper investigates various static and dynamic connectivity measures extracted from resting-state fMRI for assisting with MDD diagnosis. First, absolute Pearson correlation matrices from 85 brain regions are computed and they are used to calculate static features for predicting MDD. A predictive sub-network extracted using sub-graph entropy classifies adolescent MDD vs. typical healthy controls with high accuracy, sensitivity and specificity. Next, approaches utilizing dynamic connectivity are employed to extract tensor based, independent component based and principal component based subject specific attributes. Finally, features from static and dynamic approaches are combined to create a feature vector for classification. A leave-one-out cross-validation method is used for the final predictor performance. Out of 49 adolescents with MDD and 33 matched healthy controls, a support vector machine (SVM) classifier using a radial basis function (RBF) kernel using differential sub-graph entropy combined with dynamic connectivity features classifies MDD vs. healthy controls with an accuracy of 0.82 for leave-one-out cross-validation. This classifier has specificity and sensitivity of 0.79 and 0.84, respectively.

Unpacking Prevalence and Dichotomy in Quick Sequential Organ Failure Assessment and Systemic Inflammatory Response Syndrome Parameters: Observational Data-Driven Approach Backed by Sepsis Pathophysiology.

BackgroundConsidering morbidity, mortality, and annual treatment costs, the dramatic rise in the incidence of sepsis and septic shock among intensive care unit (ICU) admissions in US hospitals is an increasing concern. Recent changes in the sepsis definition (sepsis-3), based on the quick Sequential Organ Failure Assessment (qSOFA), have motivated the international medical informatics research community to investigate score recalculation and information retrieval, and to study the intersection between sepsis-3 and the previous definition (sepsis-2) based on systemic inflammatory response syndrome (SIRS) parameters.ObjectiveThe objective of this study was three-fold. First, we aimed to unpack the most prevalent criterion for sepsis (for both sepsis-3 and sepsis-2 predictors). Second, we intended to determine the most prevalent sepsis scenario in the ICU among 4 possible scenarios for qSOFA and 11 possible scenarios for SIRS. Third, we investigated the multicollinearity or dichotomy among qSOFA and SIRS predictors.MethodsThis observational study was conducted according to the most recent update of Medical Information Mart for Intensive Care (MIMIC-III, Version 1.4), the critical care database developed by MIT. The qSOFA (sepsis-3) and SIRS (sepsis-2) parameters were analyzed for patients admitted to critical care units from 2001 to 2012 in Beth Israel Deaconess Medical Center (Boston, MA, USA) to determine the prevalence and underlying relation between these parameters among patients undergoing sepsis screening. We adopted a multiblind Delphi method to seek a rationale for decisions in several stages of the research design regarding handling missing data and outlier values, statistical imputations and biases, and generalizability of the study.ResultsAltered mental status in the Glasgow Coma Scale (59.28%, 38,854/65,545 observations) was the most prevalent sepsis-3 (qSOFA) criterion and the white blood cell count (53.12%, 17,163/32,311 observations) was the most prevalent sepsis-2 (SIRS) criterion confronted in the ICU. In addition, the two-factored sepsis criterion of high respiratory rate (≥22 breaths/minute) and altered mental status (28.19%, among four possible qSOFA scenarios besides no sepsis) was the most prevalent sepsis-3 (qSOFA) scenario, and the three-factored sepsis criterion of tachypnea, high heart rate, and high white blood cell count (12.32%, among 11 possible scenarios besides no sepsis) was the most prevalent sepsis-2 (SIRS) scenario in the ICU. Moreover, the absolute Pearson correlation coefficients were not significant, thereby nullifying the likelihood of any linear correlation among the critical parameters and assuring the lack of multicollinearity between the parameters. Although this further bolsters evidence for their dichotomy, the absence of multicollinearity cannot guarantee that two random variables are statistically independent.ConclusionsQuantifying the prevalence of the qSOFA criteria of sepsis-3 in comparison with the SIRS criteria of sepsis-2, and understanding the underlying dichotomy among these parameters provides significant inferences for sepsis treatment initiatives in the ICU and informing hospital resource allocation. These data-driven results further offer design implications for multiparameter intelligent sepsis prediction in the ICU.

A multi-frequency and multi-GNSS method for the retrieval of the ionospheric TEC and intraday variability of receiver DCBs

As one of the important factors influencing the ionospheric total electron content (TEC) estimation accuracy, receiver differential code biases (DCBs) should be properly removed from global navigation satellite system (GNSS) measurements. The intraday variability in receiver DCBs (rDCBs), which is usually ignored in the commonly used ionospheric observable retrieval procedure, has been identified as one of the major errors degrading the accuracy of TEC estimation. The modified carrier-to-code leveling (MCCL) method can be adopted to eliminate the impact of the rDCB variability on the retrieval of the ionospheric TEC from dual-frequency (DF) GNSS observations. In this contribution, we extend the MCCL method from two aspects. First, the DF MCCL method is adapted to the multi-frequency (MF) case, in which DF, triple-frequency or even arbitrary-frequency observations can be readily processed to simultaneously estimate both the ionospheric TEC and rDCB variations. Second, the MCCL method is refined to enable the handling of GLONASS data by accounting for the effects of code inter-frequency biases induced by the frequency division multiple access (FDMA) technology. Based on the test results, the retrieval accuracy of the ionospheric TEC using our proposed method can be improved from 9.47 TECu to 2.67 TECu in the presence of significant intraday rDCB variations. We discovered that the maximum difference in the rDCB variations of the same satellite system between different frequency bands can be as large as 10 ns. The dependence of multi-GNSS and MF rDCB variations on the ambient temperature is further verified in this study. The results show that the temperature dependence of rDCB varies among different satellite systems and frequency bands. Compared to the Galileo, GPS and GLONASS satellite systems, the Beidou system (BDS) rDCB estimates exhibit a stronger correlation with the measured temperature. The percentages of stations with the mean absolute Pearson correlation coefficient value above 0.8 are 27.17% for GPS, 30.58% for GLONASS, 43.78% for BDS and 33.9% for Galileo, respectively.

Application and comparison of machine learning models for predicting quality assurance outcomes in radiation therapy treatment planning

Abstract The purpose of this study was to develop and evaluate machine learning models for predicting quality assurance (QA) outcomes of volumetric modulated arc radiation therapy (VMAT) treatment plans. A dataset of 500 VMAT treatment plans and diode-array QA measurements were collected for this study. Gamma passing rates (GPRs) were computed using a 3%/3 mm dose difference and distance-to-agreement gamma criterion with local normalization. 241 complexity metrics and plan parameters were extracted from each treatment plan and their relative importance for accurately predicting GPRs was assessed and compared using feature selection methods via forests of Extra-Trees, mutual information, and linear regression. Hyperparameters of different machine learning models – which included linear models, support vector machines (SVMs), tree-based models, and neural networks – were tuned using cross-validation on the training data (80%/20% training/testing split). Features were weakly correlated with GPRs, with the small aperture score (SAS) at 50 mm having the largest absolute Pearson correlation coefficient (0.38; p

Model complexity affects species distribution projections under climate change

AbstractAimStatistical species distribution models (SDMs) are the most common tool to predict the impact of climate change on biodiversity. They can be tuned to fit relationships at various levels of complexity (defined here as parameterization complexity, number of predictors, and multicollinearity) that may co‐determine whether projections to novel climatic conditions are useful or misleading. Here, we assessed how model complexity affects the performance of model extrapolations and influences projections of species ranges under future climate change.LocationEurope.Taxon34 European tree species.MethodsWe sampled three replicates of predictor sets for all combinations of 10 levels (n = 3–12) of environmental variables (climate, terrain, soil) and 10 levels of multicollinearity. We used these sets for each species to fit four SDM algorithms at three levels of parameterization complexity. The &gt;100,000 resulting SDM fits were then evaluated under environmental block cross‐validation and projected to environmental conditions for 2061–2080 considering four climate models and two emission scenarios. Finally, we investigated the relationships of model design with model performance and projected distributional changes.ResultsModel complexity affected both model performance and projections of species distributional change. Fits of intermediate parameterization complexity performed best, and more complex parameterizations were associated with higher projected loss of current ranges. Model performance peaked at 10–11 variables but increasing number of variables had no consistent effect on distributional change projections. Multicollinearity had a low impact on model performance but distinctly increased projected loss of current ranges.Main conclusionsSDM‐based climate change impact assessments should be based on ensembles of projections, varying SDM algorithms as well as parameterization complexity, besides emission scenarios and climate models. The number of predictor variables should be kept reasonably small and the classical threshold of maximum absolute Pearson correlation of 0.7 restricts collinearity‐driven effects in projections of species ranges.

Iron-Insensitive Quantitative Assessment of Subcortical Gray Matter Demyelination in Multiple Sclerosis Using the Macromolecular Proton Fraction.

Fast macromolecular proton fraction mapping is a recent quantitative MR imaging method for myelin assessment. The objectives of this study were to evaluate the macromolecular proton fraction as a measure of demyelination in subcortical GM structures in multiple sclerosis and assess a potential relationship between demyelination and excess iron deposition using the macromolecular proton fraction and T2* mapping. Macromolecular proton fraction and T2* maps were obtained from 12 healthy controls, 18 patients with relapsing-remitting MS, and 12 patients with secondary-progressive MS using 3T MR imaging. Parameter values in the caudate nucleus, globus pallidus, putamen, substantia nigra, and thalamus were compared between groups and correlated to clinical data. The macromolecular proton fraction in all subcortical structures and T2* in the globus pallidus, putamen, and caudate nucleus demonstrated a significant monotonic decrease from controls to patients with relapsing-remitting MS and from those with relapsing-remitting MS to patients with secondary-progressive MS. The macromolecular proton fraction in all subcortical structures significantly correlated with the Expanded Disability Status Scale and MS Functional Composite scores with absolute Pearson correlation coefficient (r) values in a range of 0.4-0.6. Significant correlations (r = -0.4 to -0.6) were also identified between the macromolecular proton fraction and the 9-Hole Peg Test, indicating a potential relationship with nigrostriatal pathway damage. Among T2* values, weak significant correlations with clinical variables were found only in the putamen. The macromolecular proton fraction did not correlate with T2* in any of the studied anatomic structures. The macromolecular proton fraction provides an iron-insensitive measure of demyelination. Myelin loss in subcortical GM structures in MS is unrelated to excess iron deposition. Subcortical GM demyelination is more closely associated with the disease phenotype and disability than iron overload.

Proteomic Profiling of 128 Proteins In 194 Acute Lymphocytic Leukemia (ALL) Patients Using Reverse Phase Proteins Arrays (RPPA) Reveals Recurrent Proteins Expression Signatures with Prognostic Implications

AbstractAbstract 3230 Background.Having previously shown that protein expression signatures (ProExpSig), based on the activation state of cell cycle, apoptosis and signal transduction (STP) regulating proteins, existed and were prognostic in AML, we extended this to evaluate protein expression patterns in ALL. Methodology.We have generated RPPA using protein derived from the leukemia-enriched fraction of 283 primary ALL samples from 216 cases with the goal of providing comprehensive proteomic based classification of ALL. Included are 253 samples from 194 newly diagnosed ALL patients (179 Marrow (BM), 95 blood (PB) samples, 41 with concurrent PB and BM specimens), and 14 relapse (REL) samples from these same cases. Phenotypes included 163 pre-B, 22 T-Cell and 9 Burkitt's cases. There were 42 PH1+ cases but only the 28 treated with TKIs were included for outcome analysis. All protein preps were made from fresh cells on the day of collection as we previously observed that protein expression changes dramatically in ALLs cells with cryopreservation. Present as controls were 16 CD34+ BM and 9 normal PB lymphocyte samples. Samples are printed as 5 serial 1:2 dilutions in duplicate using an Aushon 2470 Arrayer. Each array has a total of 6912 dots printed. Slides were probed with 128 antibodies (ABs) against apoptosis, cell cycle, signaling (STP), regulating proteins, integrins, phosphatases etc including 90 vs. total protein, 33 vs. phos-specific sites and 5 vs. caspase or PARP cleavage sites. Spot intensities were quantified using MicroVigene software. Data was analysed using R, with loading control and topographical background normalization being utilized. Results.Proteins were clustered using absolute Pearson correlation and Ward linkage. Based on the Gap statistic there appeared to be 15 constellations of protein with correlated expression. Many formed expected associations e.g., cleaved caspases, phosphorylated STATs, activated STPs. Based on the combination of expression of these 15 constellations, 7 distinct ProExpSig were defined containing between 18 and 37 cases. Phenotype was highly unbalanced with T cell and Burkitt's cases significantly overrepresented and B cell underrepresented in Signature 3 (O/E 5.7,4.2 and .3) and T cell ALL underrepresented in signatures 2, 4,5 and 6 (O/E .47, .76, .4 and .28). PH+ disease was strongly associated with signatures 5 (18 of 22) and 2 (11 of 375 cases). The PH+ dominated Sig5 was characterized by high expression of Myc, ARC, Survivin as well as high levels of activated STPs and Stats. The Ph+ cases in Sig 2 and 5 differed in that Sig5 had high expression of phospho PKCα, GAB2 and Stat 3 and 6, while Sig 2 had higher expression of proteins in the mTOR and TSC2 axis Almost all t(4:11) cases (9 of 11) were in Sig6, which was characterized by high levels of expression of most proteins excluding those that were highly expressed in the PH+ cases. Cases with Burkitt's were characterized by high expression of cyclins D1 and D3, pRB and by high levels of cleaved caspases. The WBC was markedly higher in Sig 3 and 6. The vast majority n=174) achieved remission with 9 primary refractory and 10 induction deaths occurring, so a signature predicting primary resistance was not observed. It was readily evident that there was a group of favorable signatures (1,2, 4 and 7) and another distinctly unfavorable group of signatures (3, 5, and 6). The favorable group had superior remission duration (median 172 vs. 96 weeks, p= 0.35), event free survival (median 132 vs. 72 weeks, p=0.03) and overall survival (median 180 vs. 93 weeks, P=0.04) compared to the unfavorable signatures. Conclusions.ALL is characterized by 7 distinct protein expression signatures which form a favorable and an unfavorable prognostic group. The adverse ProExpSig are associated with significantly shorter remission duration and survival. Differences within these ProExpSig can be used to direct the rational application of various targeted therapies to those patients most likely to benefit from them thereby improving outcome. Disclosures:No relevant conflicts of interest to declare.

The Relationship between Visual Field and Retinal Nerve Fiber Layer Measurements in Patients with Multiple Sclerosis

To investigate the relationship between visual function, measured by standard automated perimetry (SAP), and retinal nerve fiber layer (RNFL) thickness, measured by optical coherence tomography (OCT), in patients with multiple sclerosis (MS). SAP and RNFL thickness were measured in patients with MS in 28 eyes with the last optic neuritis (ON) >or=6 months prior (ON group) and 33 eyes without ON history (non-ON group). Abnormal overall or quadrant RNFL thickness was defined by measured values below 5% of the norm. A whole visual field or a sector of the field was classified as abnormal by using cluster criteria on total-deviation plots. Agreement between SAP and OCT results in classifying eyes/sectors was presented as a percentage of observed agreement, along with the AC1 statistic, which corrects for chance agreement. Regression analyses were performed relating several SAP parameters and RNFL thickness in the ON group. ON eyes showed more loss of visual sensitivity (MD, P = 0.02) and more loss of RNFL thickness (P < 0.0001) than did non-ON eyes. SAP and OCT agreed in 86% (AC1 = 0.78) of eyes and 69% (AC1 = 0.38) of sectors in the ON group and 61% (AC1 = 0.33) of eyes and 66% (AC1 = 0.48) of sectors in the non-ON group. Overall RNFL thickness was related to MD (dB) by a simple exponential function (R(2) = 0.48), supporting a linear relationship between these measures when both are expressed on linear scales. Absolute Pearson correlation coefficients for overall RNFL thickness and several SAP parameters ranged from 0.51 to 0.69. Good agreement between SAP and OCT was found in ON eyes but not in non-ON eyes or in individual sectors in either group. The findings in this study provide further support for the utility of combining structural and functional testing in clinical research on patients with MS, as well as in future neuroprotection trials for which the anterior visual pathways in patients with MS and optic neuritis may be used as a model.