Identifying treatment non-responders based on pre-treatment gait characteristics - A machine learning approach.

Background: Paediatric movement disorders such as cerebral palsy often negatively impact walking behaviour. Although clinical gait analysis is usually performed to guide therapy decisions, not all respond positively to their assigned treatment. Identifying these individuals based on their pre-treatment characteristics could guide clinicians towards more appropriate and personalized interventions. Using routinely collected pre-treatment gait and anthropometric features, we aimed to assess whether standard machine learning approaches can be effective in identifying patients at risk of negative treatment outcomes. Methods: Observational data of 119 patients with movement disorders were retrospectively extracted from a local clinical database, comprising sagittal joint angles and spatiotemporal parameters, derived from motion capture data pre- and post-treatment (physiotherapy, orthosis, botulin toxin injections, or surgery). Participants were labelled based on their change in gait profile score (GPS, non-responders with a decline in GPS of <1.6° vs. responders). Their pre-treatment features (sagittal joint angles, spatiotemporal parameters, anthropometrics) were used to train a support vector machine classifier with 5-fold cross-validation and Bayesian optimization within a MATLAB-based Classification Learner App. Results: An average accuracy of 88.2 ± 0.5 % was achieved for identifying participants whose gait will not respond to treatment, with 64 % true negative rate and an area under the curve of 88 %. Conclusion: Overall, a classical machine learning model was able to identify patients at risk of not responding to treatment, based on gait features and anthropometrics collected prior to treatment. The output of such a model could function as a warning signal, notifying clinicians that a certain individual might not respond well to the standard of care and that a more personalized intervention might be needed.

Generative Adversarial Networks-Based Data Augmentation for Brain-Computer Interface.

The performance of a classifier in a brain-computer interface (BCI) system is highly dependent on the quality and quantity of training data. Typically, the training data are collected in a laboratory where the users perform tasks in a controlled environment. However, users' attention may be diverted in real-life BCI applications and this may decrease the performance of the classifier. To improve the robustness of the classifier, additional data can be acquired in such conditions, but it is not practical to record electroencephalogram (EEG) data over several long calibration sessions. A potentially time- and cost-efficient solution is artificial data generation. Hence, in this study, we proposed a framework based on the deep convolutional generative adversarial networks (DCGANs) for generating artificial EEG to augment the training set in order to improve the performance of a BCI classifier. To make a comparative investigation, we designed a motor task experiment with diverted and focused attention conditions. We used an end-to-end deep convolutional neural network for classification between movement intention and rest using the data from 14 subjects. The results from the leave-one subject-out (LOO) classification yielded baseline accuracies of 73.04% for diverted attention and 80.09% for focused attention without data augmentation. Using the proposed DCGANs-based framework for augmentation, the results yielded a significant improvement of 7.32% for diverted attention ( ) and 5.45% for focused attention ( ). In addition, we implemented the method on the data set IVa from BCI competition III to distinguish different motor imagery tasks. The proposed method increased the accuracy by 3.57% ( ). This study shows that using GANs for EEG augmentation can significantly improve BCI performance, especially in real-life applications, whereby users' attention may be diverted.

A Vital Signs Telemonitoring Programme Improves the Dynamic Prediction of Readmission Risk in Patients with Heart Failure.

Heart failure (HF) is a leading cause of hospital readmissions. There is great interest in approaches to efficiently predict emerging HF-readmissions in the community setting. We investigate the possibility of leveraging streaming telemonitored vital signs data alongside readily accessible patient profile information for predicting evolving 30-day HF-related readmission risk. We acquired data within a non-randomized controlled study that enrolled 150 HF patients over a 1-year post-discharge telemonitoring and telesupport programme. Using the sequential data and associated ground truth readmission outcomes, we developed a recurrent neural network model for dynamic risk prediction. The model detects emerging readmissions with sensitivity > 71%, specificity > 75%, AUROC ~80%. We characterize model performance in relation to telesupport based nurse assessments, and demonstrate strong sensitivity improvements. Our approach enables early stratification of high-risk patients and could enable adaptive targeting of care resources for managing patients with the most urgent needs at any given time.

Inter-subject transfer learning with an end-to-end deep convolutional neural network for EEG-based BCI.

OBJECTIVE: Despite the effective application of deep learning (DL) in brain-computer interface (BCI) systems, the successful execution of this technique, especially for inter-subject classification, in cognitive BCI has not been accomplished yet. In this paper, we propose a framework based on the deep convolutional neural network (CNN) to detect the attentive mental state from single-channel raw electroencephalography (EEG) data. APPROACH: We develop an end-to-end deep CNN to decode the attentional information from an EEG time series. We also explore the consequences of input representations on the performance of deep CNN by feeding three different EEG representations into the network. To ensure the practical application of the proposed framework and avoid time-consuming re-training, we perform inter-subject transfer learning techniques as a classification strategy. Eventually, to interpret the learned attentional patterns, we visualize and analyse the network perception of the attention and non-attention classes. MAIN RESULTS: The average classification accuracy is 79.26%, with only 15.83% of 120 subjects having an accuracy below 70% (a generally accepted threshold for BCI). This is while with the inter-subject approach, it is literally difficult to output high classification accuracy. This end-to-end classification framework surpasses conventional classification methods for attention detection. The visualization results demonstrate that the learned patterns from the raw data are meaningful. SIGNIFICANCE: This framework significantly improves attention detection accuracy with inter-subject classification. Moreover, this study sheds light on the research on end-to-end learning; the proposed network is capable of learning from raw data with the least amount of pre-processing, which in turn eliminates the extensive computational load of time-consuming data preparation and feature extraction.

Personalized features for attention detection in children with Attention Deficit Hyperactivity Disorder.

Measuring attention from electroencephalogram (EEG) has found applications in the treatment of Attention Deficit Hyperactivity Disorder (ADHD). It is of great interest to understand what features in EEG are most representative of attention. Intensive research has been done in the past and it has been proven that frequency band powers and their ratios are effective features in detecting attention. However, there are still unanswered questions, like, what features in EEG are most discriminative between attentive and non-attentive states? Are these features common among all subjects or are they subject-specific and must be optimized for each subject? Using Mutual Information (MI) to perform subject-specific feature selection on a large data set including 120 ADHD children, we found that besides theta beta ratio (TBR) which is commonly used in attention detection and neurofeedback, the relative beta power and theta/(alpha+beta) (TBAR) are also equally significant and informative for attention detection. Interestingly, we found that the relative theta power (which is also commonly used) may not have sufficient discriminative information itself (it is informative only for 3.26% of ADHD children). We have also demonstrated that although these features (relative beta power, TBR and TBAR) are the most important measures to detect attention on average, different subjects have different set of most discriminative features.