Analysis of variations of correlation dimension and nonlinear interdependence for the prediction of pediatric myoclonic seizures – A preliminary study

Predicting Severity of Huntington's Disease With Wearable Sensors

The Unified Huntington's Disease Rating Scale (UHDRS) is the primary clinical assessment tool for rating motor function in patients with Huntington's disease (HD). However, the UHDRS and similar rating scales (e.g., UPDRS) are both subjective and limited to in-office assessments that must be administered by a trained and experienced rater. An objective, automated method of quantifying disease severity would facilitate superior patient care and could be used to better track severity over time. We conducted the present study to evaluate the feasibility of using wearable sensors, coupled with machine learning algorithms, to rate motor function in patients with HD. Fourteen participants with symptomatic HD and 14 healthy controls participated in the study. Each participant wore five adhesive biometric sensors applied to the trunk and each limb while completing brief walking, sitting, and standing tasks during a single office visit. A two-stage machine learning method was employed to classify participants by HD status and to predict UHDRS motor subscores. Linear discriminant analysis correctly classified all participants' HD status except for one control subject with abnormal gait (96.4% accuracy, 92.9% sensitivity, and 100% specificity in leave-one-out cross-validation). Two regression models accurately predicted individual UHDRS subscores for gait, and dystonia within a 10% margin of error. Our regression models also predicted a composite UHDRS score-a sum of left and right arm rigidity, total chorea, total dystonia, bradykinesia, gait, and tandem gait subscores-with an average error below 15%. Machine learning classifiers trained on brief in-office datasets discriminated between controls and participants with HD, and could accurately predict selected motor UHDRS subscores. Our results could enable the future use of biosensors for objective HD assessment in the clinic or remotely and could inform future studies for the use of this technology as a potential endpoint in clinical trials.

Real-time epileptic seizure prediction based on online monitoring of pre-ictal features

Reliable prediction of epileptic seizures is of prime importance as it can drastically change the quality of life for patients. This study aims to propose a real-time low computational approach for the prediction of epileptic seizures and to present an efficient hardware implementation of this approach for portable prediction systems. Three levels of feature extraction are performed to characterize the pre-ictal activities of the EEG signal. In the first-level, the line length algorithm is applied to the pre-ictal region. The features obtained in the first-level are mathematically integrated to extract the second-level features and then the line lengths of the second-level features are calculated to obtain our third-level feature. The third-level information is compared with predefined threshold levels to make a decision on whether the extracted characteristics are relevant to a seizure occurrence or not. The validity of this algorithm was tested by EEG recordings in the CHB-MIT database (97 seizures, 834.224 h) for 19 epileptic patients. The results showed that the average sensitivity was 90.62%, the specificity was 88.34%, the accuracy was 88.76% with the average false prediction rate as low as 0.0046 h^-1, and the average prediction time was 23.3 min. The low computational complexity is the superiority of the proposed approach, which provides a technologically simple but accurate way of predicting epileptic seizures and enables hardware implantable devices. Graphical abstract Proposed seizure prediction algorithm and its features.