[A hybrid attention temporal sequential network for sleep stage classification].

Sleep stage classification is a necessary fundamental method for the diagnosis of sleep diseases, which has attracted extensive attention in recent years. Traditional methods for sleep stage classification, such as manual marking methods and machine learning algorithms, have the limitations of low efficiency and defective generalization. Recently, deep neural networks have shown improved results by the capability of learning complex pattern in the sleep data. However, these models ignore the intra-temporal sequential information and the correlation among all channels in each segment of the sleep data. To solve these problems, a hybrid attention temporal sequential network model is proposed in this paper, choosing recurrent neural network to replace traditional convolutional neural network, and extracting temporal features of polysomnography from the perspective of time. Furthermore, intra-temporal attention mechanism and channel attention mechanism are adopted to achieve the fusion of the intra-temporal representation and the fusion of channel-correlated representation. And then, based on recurrent neural network and inter-temporal attention mechanism, this model further realized the fusion of inter-temporal contextual representation. Finally, the end-to-end automatic sleep stage classification is accomplished according to the above hybrid representation. This paper evaluates the proposed model based on two public benchmark sleep datasets downloaded from open-source website, which include a number of polysomnography. Experimental results show that the proposed model could achieve better performance compared with ten state-of-the-art baselines. The overall accuracy of sleep stage classification could reach 0.801, 0.801 and 0.717, respectively. Meanwhile, the macro average F1-scores of the proposed model could reach 0.752, 0.728 and 0.700. All experimental results could demonstrate the effectiveness of the proposed model.

A hybrid self-attention deep learning framework for multivariate sleep stage classification.

BACKGROUND: Sleep is a complex and dynamic biological process characterized by different sleep patterns. Comprehensive sleep monitoring and analysis using multivariate polysomnography (PSG) records has achieved significant efforts to prevent sleep-related disorders. To alleviate the time consumption caused by manual visual inspection of PSG, automatic multivariate sleep stage classification has become an important research topic in medical and bioinformatics. RESULTS: We present a unified hybrid self-attention deep learning framework, namely HybridAtt, to automatically classify sleep stages by capturing channel and temporal correlations from multivariate PSG records. We construct a new multi-view convolutional representation module to learn channel-specific and global view features from the heterogeneous PSG inputs. The hybrid attention mechanism is designed to further fuse the multi-view features by inferring their dependencies without any additional supervision. The learned attentional representation is subsequently fed through a softmax layer to train an end-to-end deep learning model. CONCLUSIONS: We empirically evaluate our proposed HybridAtt model on a benchmark PSG dataset in two feature domains, referred to as the time and frequency domains. Experimental results show that HybridAtt consistently outperforms ten baseline methods in both feature spaces, demonstrating the effectiveness of HybridAtt in the task of sleep stage classification.

FusionAtt: Deep Fusional Attention Networks for Multi-Channel Biomedical Signals.

Recently, pervasive sensing technologies have been widely applied to comprehensive patient monitoring in order to improve clinical treatment. Various types of biomedical signals collected by different sensing channels provide different aspects of patient health information. However, due to the uncertainty and variability in clinical observation, not all the channels are relevant and important to the target task. Thus, in order to extract informative representations from multi-channel biosignals, channel awareness has become a key enabler for deep learning in biosignal processing and has attracted increasing research interest in health informatics. Towards this end, we propose FusionAtt-a deep fusional attention network that can learn channel-aware representations of multi-channel biosignals, while preserving complex correlations among all the channels. FusionAtt is able to dynamically quantify the importance of each biomedical channel, and relies on more informative ones to enhance feature representation in an end-to-end manner. We empirically evaluated FusionAtt in two clinical tasks: multi-channel seizure detection and multivariate sleep stage classification. Experimental results showed that FusionAtt consistently outperformed the state-of-the-art models in four different evaluation measurements, demonstrating the effectiveness of the proposed fusional attention mechanism.

A Review of Ultrasound Tissue Characterization with Mean Scatterer Spacing.

Tissues exhibiting quasi-periodic structures can be modeled as a collection of diffuse scatterers and coherent scatterers. The mean scatterer spacing (MSS) of coherent and quasi-periodic components is directly related to tissue microstructure and has become an important quantitative ultrasound (QUS) parameter in the characterization of quasi-periodic tissues. In this paper, a review of the literature on the development of MSS as a QUS parameter was conducted. First, a unified theoretical background of MSS estimates was provided. Then, the application of MSS estimates was summarized with respect to liver, spleen, breast, bone, muscle, and other tissues. MSS estimation techniques were applied to (a) the diagnosis of hepatitis, liver fibrosis and cirrhosis, and lesions in tissues such as liver, breast, and spleen; (b) the differentiation between benign and malignant breast tumors, and the grading of breast cancer; (c) the detection of cancellous bone; and (d) the monitoring of the efficacy of treatments such as thermal ablation, with various levels of success. Future developments were also discussed in terms of real-time implementation of MSS estimates, local MSS estimation, relationship of MSS to other QUS parameters, combination of MSS with other QUS parameters, in vivo validation of MSS estimates, MSS parametric imaging, and three-dimensional ultrasound tissue characterization.

An efficient hierarchical video coding scheme combining visual perception characteristics.

Different visual perception characteristic saliencies are the key to constitute the low-complexity video coding framework. A hierarchical video coding scheme based on human visual systems (HVS) is proposed in this paper. The proposed scheme uses a joint video coding framework consisting of visual perception analysis layer (VPAL) and video coding layer (VCL). In VPAL, effective visual perception characteristics detection algorithm is proposed to achieve visual region of interest (VROI) based on the correlation between coding information (such as motion vector, prediction mode, etc.) and visual attention. Then, the interest priority setting for VROI according to visual perception characteristics is completed. In VCL, the optional encoding method is developed utilizing the visual interested priority setting results from VPAL. As a result, the proposed scheme achieves information reuse and complementary between visual perception analysis and video coding. Experimental results show that the proposed hierarchical video coding scheme effectively alleviates the contradiction between complexity and accuracy. Compared with H.264/AVC (JM17.0), the proposed scheme reduces 80% video coding time approximately and maintains a good video image quality as well. It improves video coding performance significantly.

An adaptive motion estimation scheme for video coding.

The unsymmetrical-cross multihexagon-grid search (UMHexagonS) is one of the best fast Motion Estimation (ME) algorithms in video encoding software. It achieves an excellent coding performance by using hybrid block matching search pattern and multiple initial search point predictors at the cost of the computational complexity of ME increased. Reducing time consuming of ME is one of the key factors to improve video coding efficiency. In this paper, we propose an adaptive motion estimation scheme to further reduce the calculation redundancy of UMHexagonS. Firstly, new motion estimation search patterns have been designed according to the statistical results of motion vector (MV) distribution information. Then, design a MV distribution prediction method, including prediction of the size of MV and the direction of MV. At last, according to the MV distribution prediction results, achieve self-adaptive subregional searching by the new estimation search patterns. Experimental results show that more than 50% of total search points are dramatically reduced compared to the UMHexagonS algorithm in JM 18.4 of H.264/AVC. As a result, the proposed algorithm scheme can save the ME time up to 20.86% while the rate-distortion performance is not compromised.

Graph-based EEG approach for depression prediction: integrating time-frequency complexity and spatial topology.

Depression has become the prevailing global mental health concern. The accuracy of traditional depression diagnosis methods faces challenges due to diverse factors, making primary identification a complex task. Thus, the imperative lies in developing a method that fulfills objectivity and effectiveness criteria for depression identification. Current research underscores notable disparities in brain activity between individuals with depression and those without. The Electroencephalogram (EEG), as a biologically reflective and easily accessible signal, is widely used to diagnose depression. This article introduces an innovative depression prediction strategy that merges time-frequency complexity and electrode spatial topology to aid in depression diagnosis. Initially, time-frequency complexity and temporal features of the EEG signal are extracted to generate node features for a graph convolutional network. Subsequently, leveraging channel correlation, the brain network adjacency matrix is employed and calculated. The final depression classification is achieved by training and validating a graph convolutional network with graph node features and a brain network adjacency matrix based on channel correlation. The proposed strategy has been validated using two publicly available EEG datasets, MODMA and PRED+CT, achieving notable accuracy rates of 98.30 and 96.51%, respectively. These outcomes affirm the reliability and utility of our proposed strategy in predicting depression using EEG signals. Additionally, the findings substantiate the effectiveness of EEG time-frequency complexity characteristics as valuable biomarkers for depression prediction.

Magneto-acousto-electrical tomography based on frequency response compensated linearly frequency-modulated signal stimulation.

Objective. In magneto-acousto-electrical tomography (MAET), linearly frequency-modulated (LFM) signal stimulation uses much lower peak voltage than the spike pulse stimulation, lengthening the operation life of the transducer. However, due to the uneven frequency responses of the transducer, the low-noise amplifier (LNA), and the bandpass filter (BPF), MAET using LFM signal stimulation suffers from longitudinal resolution loss. In this paper, frequency response compensated linearly frequency-modulated (FRC-LFM) signal stimulation is investigated to resolve the problem.Approach. The physical model of measurement of the frequency responses of the transducer and the cascading module of the detection electrodes, the LNA, and the BPF is constructed. The frequency responses are approximated by fitting a curve to the measurement data. The frequency response compensation function is set to the reciprocal of the product of the frequency responses. The digital FRC-LFM signal is generated in MATLAB and converted to analog signal through an arbitrary waveform generator. Two groups of MAET experiments are designed to confirm the performance of the FRC-LFM signal stimulation. Pure agar phantom with rectangular through-holes and agar phantom with pork tissue inclusion serve as the samples.Main results. The pulse-compressed magneto-acousto-electrical signal obtained using FRC-LFM stimulation has narrower main-lobe than that obtained using LFM excitation, although the signal to noise pulse interference ratio of the former is little lower than that of the latter, which is due to the limitation of the power amplifier. FRC-LFM also proves to be an effective method to utilize the frequency outside the working band of the transducer in MAET.Significance. The method in this study compensates for the longitudinal resolution loss due to the uneven frequency responses. Combining with high-capability power amplifier and high-performance LNA, the MAET using FRC-LFM signal stimulation can potentially achieve high longitudinal resolution and high sensitivity, advancing MAET toward the clinical application.

Magneto-acousto-electrical tomography using nonlinearly frequency-modulated ultrasound.

Objective. In this study, nonlinearly frequency-modulated (NLFM) ultrasound was applied to magneto-acousto-electrical tomography (MAET) to increase the dynamic range of detection.Approach. Generation of NLFM signals using window function method-based on the principle of stationary phase-and piecewise linear frequency modulation method-based on the genetic algorithm-was discussed. The MAET experiment systems using spike, linearly frequency-modulated (LFM), or NLFM pulse stimulation were constructed, and three groups of MAET experiments on saline agar phantom samples were carried out to verify the performance-respectively the sensitivity, the dynamic range, and the longitudinal resolution of detection-of MAET using NLFM ultrasound in comparison to that using LFM ultrasound. Based on the above experiments, a pork sample was imaged by ultrasound imaging method, spike MAET method, LFM MAET method, and NLFM MAET method, to compare the imaging accuracy.Main results. The experiment results showed that, through sacrificing very little main-lobe width of pulse compression or equivalently the longitudinal resolution, the MAET using NLFM ultrasound achieved higher signal-to-interference ratio (and therefore higher detection sensitivity), lower side-lobe levels of pulse compression (and therefore larger dynamic range of detection), and large anti-interference capability, compared to the MAET using LFM ultrasound.Significance. The applicability of the MAET using NLFM ultrasound was proved in circumferences where sensitivity and dynamic range of detection were mostly important and slightly lower longitudinal resolution of detection was acceptable. The study furthered the scheme of using coded ultrasound excitation toward the clinical application of MAET.

Low-complexity saliency detection algorithm for fast perceptual video coding.

A low-complexity saliency detection algorithm for perceptual video coding is proposed; low-level encoding information is adopted as the characteristics of visual perception analysis. Firstly, this algorithm employs motion vector (MV) to extract temporal saliency region through fast MV noise filtering and translational MV checking procedure. Secondly, spatial saliency region is detected based on optimal prediction mode distributions in I-frame and P-frame. Then, it combines the spatiotemporal saliency detection results to define the video region of interest (VROI). The simulation results validate that the proposed algorithm can avoid a large amount of computation work in the visual perception characteristics analysis processing compared with other existing algorithms; it also has better performance in saliency detection for videos and can realize fast saliency detection. It can be used as a part of the video standard codec at medium-to-low bit-rates or combined with other algorithms in fast video coding.

A temporal multi-scale hybrid attention network for sleep stage classification.

Sleep is crucial for human health. Automatic sleep stage classification based on polysomnogram (PSG) is meaningful for the diagnosis of sleep disorders, which has attracted extensive attention in recent years. Most existing methods could not fully consider the different transitions of sleep stages and fit the visual inspection of sleep experts simultaneously. To this end, we propose a temporal multi-scale hybrid attention network, namely TMHAN, to automatically achieve sleep staging. The temporal multi-scale mechanism incorporates short-term abrupt and long-term periodic transitions of the successive PSG epochs. Furthermore, the hybrid attention mechanism includes 1-D local attention, 2-D global attention, and 2-D contextual sparse multi-head self-attention for three kinds of sequence-level representations. The concatenated representation is subsequently fed into a softmax layer to train an end-to-end model. Experimental results on two benchmark sleep datasets show that TMHAN obtains the best performance compared with several baselines, demonstrating the effectiveness of our model. In general, our work not only provides good classification performance, but also fits the actual sleep staging processes, which makes contribution for the combination of deep learning and sleep medicine.

X-ray Cherenkov-luminescence tomography reconstruction with a three-component deep learning algorithm: Swin transformer, convolutional neural network, and locality module.

Significance: X-ray Cherenkov-luminescence tomography (XCLT) produces fast emission data from megavoltage (MV) x-ray scanning, in which the excitation location of molecules within tissue is reconstructed. However standard filtered backprojection (FBP) algorithms for XCLT sinogram reconstruction can suffer from insufficient data due to dose limitations, so there are limits in the reconstruction quality with some artifacts. We report a deep learning algorithm for XCLT with high image quality and improved quantitative accuracy. Aim: To directly reconstruct the distribution of emission quantum yield for x-ray Cherenkov-luminescence tomography, we proposed a three-component deep learning algorithm that includes a Swin transformer, convolution neural network, and locality module model. Approach: A data-to-image model x-ray Cherenkov-luminescence tomography is developed based on a Swin transformer, which is used to extract pixel-level prior information from the sinogram domain. Meanwhile, a convolutional neural network structure is deployed to transform the extracted pixel information from the sinogram domain to the image domain. Finally, a locality module is designed between the encoder and decoder connection structures for delivering features. Its performance was validated with simulation, physical phantom, and in vivo experiments. Results: This approach can better deal with the limits to data than conventional FBP methods. The method was validated with numerical and physical phantom experiments, with results showing that it improved the reconstruction performance mean square error ( > 94.1 % ), peak signal-to-noise ratio ( > 41.7 % ), and Pearson correlation ( > 19 % ) compared with the FBP algorithm. The Swin-CNN also achieved a 32.1% improvement in PSNR over the deep learning method AUTOMAP. Conclusions: This study shows that the three-component deep learning algorithm provides an effective reconstruction method for x-ray Cherenkov-luminescence tomography.

Selfrec-Net: self-supervised deep learning approach for the reconstruction of Cherenkov-excited luminescence scanned tomography.

As an emerging imaging technique, Cherenkov-excited luminescence scanned tomography (CELST) can recover a high-resolution 3D distribution of quantum emission fields within tissue using X-ray excitation for deep penetrance. However, its reconstruction is an ill-posed and under-conditioned inverse problem because of the diffuse optical emission signal. Deep learning based image reconstruction has shown very good potential for solving these types of problems, however they suffer from a lack of ground-truth image data to confirm when used with experimental data. To overcome this, a self-supervised network cascaded by a 3D reconstruction network and the forward model, termed Selfrec-Net, was proposed to perform CELST reconstruction. Under this framework, the boundary measurements are input to the network to reconstruct the distribution of the quantum field and the predicted measurements are subsequently obtained by feeding the reconstructed result to the forward model. The network was trained by minimizing the loss between the input measurements and the predicted measurements rather than the reconstructed distributions and the corresponding ground truths. Comparative experiments were carried out on both numerical simulations and physical phantoms. For singular luminescent targets, the results demonstrate the effectiveness and robustness of the proposed network, and comparable performance can be attained to a state-of-the-art deep supervised learning algorithm, where the accuracy of the emission yield and localization of the objects was far superior to iterative reconstruction methods. Reconstruction of multiple objects is still reasonable with high localization accuracy, although with limits to the emission yield accuracy as the distribution becomes more complex. Overall though the reconstruction of Selfrec-Net provides a self-supervised way to recover the location and emission yield of molecular distributions in murine model tissues.

Total variation regularization for bioluminescence tomography with the split Bregman method.

Regularization methods have been broadly applied to bioluminescence tomography (BLT) to obtain stable solutions, including l2 and l1 regularizations. However, l2 regularization can oversmooth reconstructed images and l1 regularization may sparsify the source distribution, which degrades image quality. In this paper, the use of total variation (TV) regularization in BLT is investigated. Since a nonnegativity constraint can lead to improved image quality, the nonnegative constraint should be considered in BLT. However, TV regularization with a nonnegativity constraint is extremely difficult to solve due to its nondifferentiability and nonlinearity. The aim of this work is to validate the split Bregman method to minimize the TV regularization problem with a nonnegativity constraint for BLT. The performance of split Bregman-resolved TV (SBRTV) based BLT reconstruction algorithm was verified with numerical and in vivo experiments. Experimental results demonstrate that the SBRTV regularization can provide better regularization quality over l2 and l1 regularizations.

Efficient sparse reconstruction algorithm for bioluminescence tomography based on duality and variable splitting.

Bioluminescence tomography (BLT) can three-dimensionally and quantitatively resolve the molecular processes in small animals in vivo. In this paper, we propose a BLT reconstruction algorithm based on duality and variable splitting. By using duality and variable splitting to obtain a new equivalent constrained optimization problem and updating the primal variable as the Lagrangian multiplier in the dual augmented Lagrangian problem, the proposed method can obtain fast and stable source reconstruction even without the permissible source region and multispectral measurements. Numerical simulations on a mouse atlas and in vivo mouse experiments were conducted to validate the effectiveness and potential of the method.

A Depression Prediction Algorithm Based on Spatiotemporal Feature of EEG Signal.

Depression has gradually become the most common mental disorder in the world. The accuracy of its diagnosis may be affected by many factors, while the primary diagnosis seems to be difficult to define. Finding a way to identify depression by satisfying both objective and effective conditions is an urgent issue. In this paper, a strategy for predicting depression based on spatiotemporal features is proposed, and is expected to be used in the auxiliary diagnosis of depression. Firstly, electroencephalogram (EEG) signals were denoised through the filter to obtain the power spectra of the three corresponding frequency ranges, Theta, Alpha and Beta. Using orthogonal projection, the spatial positions of the electrodes were mapped to the brainpower spectrum, thereby obtaining three brain maps with spatial information. Then, the three brain maps were superimposed on a new brain map with frequency domain and spatial characteristics. A Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU) were applied to extract the sequential feature. The proposed strategy was validated with a public EEG dataset, achieving an accuracy of 89.63% and an accuracy of 88.56% with the private dataset. The network had less complexity with only six layers. The results show that our strategy is credible, less complex and useful in predicting depression using EEG signals.

Deep-learning based image reconstruction for MRI-guided near-infrared spectral tomography.

Non-invasive near-infrared spectral tomography (NIRST) can incorporate the structural information provided by simultaneous magnetic resonance imaging (MRI), and this has significantly improved the images obtained of tissue function. However, the process of MRI guidance in NIRST has been time consuming because of the needs for tissue-type segmentation and forward diffuse modeling of light propagation. To overcome these problems, a reconstruction algorithm for MRI-guided NIRST based on deep learning is proposed and validated by simulation and real patient imaging data for breast cancer characterization. In this approach, diffused optical signals and MRI images were both used as the input to the neural network, and simultaneously recovered the concentrations of oxy-hemoglobin, deoxy-hemoglobin, and water via end-to-end training by using 20,000 sets of computer-generated simulation phantoms. The simulation phantom studies showed that the quality of the reconstructed images was improved, compared to that obtained by other existing reconstruction methods. Reconstructed patient images show that the well-trained neural network with only simulation data sets can be directly used for differentiating malignant from benign breast tumors.

An adaptive regularization parameter choice strategy for multispectral bioluminescence tomography.

PURPOSE: Bioluminescence tomography (BLT) provides an effective tool for monitoring physiological and pathological activities in vivo. However, the measured data in bioluminescence imaging are corrupted by noise. Therefore, regularization methods are commonly used to find a regularized solution. Nevertheless, for the quality of the reconstructed bioluminescent source obtained by regularization methods, the choice of the regularization parameters is crucial. To date, the selection of regularization parameters remains challenging. With regards to the above problems, the authors proposed a BLT reconstruction algorithm with an adaptive parameter choice rule. METHODS: The proposed reconstruction algorithm uses a diffusion equation for modeling the bioluminescent photon transport. The diffusion equation is solved with a finite element method. Computed tomography (CT) images provide anatomical information regarding the geometry of the small animal and its internal organs. To reduce the ill-posedness of BLT, spectral information and the optimal permissible source region are employed. Then, the relationship between the unknown source distribution and multiview and multispectral boundary measurements is established based on the finite element method and the optimal permissible source region. Since the measured data are noisy, the BLT reconstruction is formulated as l(2) data fidelity and a general regularization term. When choosing the regularization parameters for BLT, an efficient model function approach is proposed, which does not require knowledge of the noise level. This approach only requests the computation of the residual and regularized solution norm. With this knowledge, we construct the model function to approximate the objective function, and the regularization parameter is updated iteratively. RESULTS: First, the micro-CT based mouse phantom was used for simulation verification. Simulation experiments were used to illustrate why multispectral data were used rather than monochromatic data. Furthermore, the study conducted using an adaptive regularization parameter demonstrated our ability to accurately localize the bioluminescent source. With the adaptively estimated regularization parameter, the reconstructed center position of the source was (20.37, 31.05, 12.95) mm, and the distance to the real source was 0.63 mm. The results of the dual-source experiments further showed that our algorithm could localize the bioluminescent sources accurately. The authors then presented experimental evidence that the proposed algorithm exhibited its calculated efficiency over the heuristic method. The effectiveness of the new algorithm was also confirmed by comparing it with the L-curve method. Furthermore, various initial speculations regarding the regularization parameter were used to illustrate the convergence of our algorithm. Finally, in vivo mouse experiment further illustrates the effectiveness of the proposed algorithm. CONCLUSIONS: Utilizing numerical, physical phantom and in vivo examples, we demonstrated that the bioluminescent sources could be reconstructed accurately with automatic regularization parameters. The proposed algorithm exhibited superior performance than both the heuristic regularization parameter choice method and L-curve method based on the computational speed and localization error.

Quantification of blood flow index in diffuse correlation spectroscopy using long short-term memory architecture.

Diffuse correlation spectroscopy (DCS) is a noninvasive technique that derives blood flow information from measurements of the temporal intensity fluctuations of multiply scattered light. Blood flow index (BFI) and especially its variation was demonstrated to be approximately proportional to absolute blood flow. We investigated and assessed the utility of a long short-term memory (LSTM) architecture for quantification of BFI in DCS. Phantom and in vivo experiments were established to measure normalized intensity autocorrelation function data. Improved accuracy and faster computational time were gained by the proposed LSTM architecture. The results support the notion of using proposed LSTM architecture for quantification of BFI in DCS. This approach would be especially useful for continuous real-time monitoring of blood flow.

Exercise Training Increases Resting Calf Muscle Oxygen Metabolism in Patients with Peripheral Artery Disease.

Exercise training can mitigate symptoms of claudication (walking-induced muscle pain) in patients with peripheral artery disease (PAD). One adaptive response enabling this improvement is enhanced muscle oxygen metabolism. To explore this issue, we used arterial-occlusion diffuse optical spectroscopy (AO-DOS) to measure the effects of exercise training on the metabolic rate of oxygen (MRO2) in resting calf muscle. Additionally, venous-occlusion DOS (VO-DOS) and frequency-domain DOS (FD-DOS) were used to measure muscle blood flow (F) and tissue oxygen saturation (StO2), and resting calf muscle oxygen extraction fraction (OEF) was calculated from MRO2, F, and blood hemoglobin. Lastly, the venous/arterial ratio (γ) of blood monitored by FD-DOS was calculated from OEF and StO2. PAD patients who experience claudication (n = 28) were randomly assigned to exercise and control groups. Patients in the exercise group received 3 months of supervised exercise training. Optical measurements were obtained at baseline and at 3 months in both groups. Resting MRO2, OEF, and F, respectively, increased by 30% (12%, 44%) (p < 0.001), 17% (6%, 45%) (p = 0.003), and 7% (0%, 16%) (p = 0.11), after exercise training (median (interquartile range)). The pre-exercise γ was 0.76 (0.61, 0.89); it decreased by 12% (35%, 6%) after exercise training (p = 0.011). Improvement in exercise performance was associated with a correlative increase in resting OEF (R = 0.45, p = 0.02).