An embedded intracranial seizure monitor for objective outcome measurements and rhythm identification.

Providing clinicians with objective outcomes of neuromodulation therapy is a key unmet need, especially in emerging areas such as epilepsy and mood disorders. These diseases have episodic behavior and circadian/multidien rhythm characteristics that are difficult to capture in short clinical follow-ups. This work presents preliminary validation evidence for an implantable neuromodulation system with integrated physiological event monitoring, with an initial focus on seizure tracking for epilepsy. The system was developed to address currently unmet requirements for patients undergoing neuromodulation therapy for neurological disorders, specifically the ability to sense physiological data during stimulation and track events with seconds-level granularity. The system incorporates an interactive software tool to enable optimal configuration of the signal processing chain on an embedded implantable device (the Picostim-DyNeuMo Mk-2) including data ingestion from the device loop recorder, event labeling, generation of filter and classification parameters, as well as summary statistics. When the monitor parameters are optimized, the user can wirelessly update the system for chronic event tracking. The simulated performance of the device was assessed using an in silico model with human data to predict the real-time device performance at tracking recorded seizure activity. The in silico performance was then compared against its performance in an in vitro model to capture the full environmental constraints such as sensing during stimulation at the tissue electrode interface. In vitro modeling demonstrated comparable results to the in silico model, providing verification of the software tool and model. This study provides validation evidence of the suitability of the proposed system for tracking longitudinal seizure activity. Given its flexibility, the event monitor can be adapted to track other disorders with episodic and rhythmic symptoms represented by bioelectrical behavior.Clinical relevance-An implantable neuromodulation system is presented that enables chronic tracking of physiological events in disease. This physiological monitor provides the basis for longitudinal assessments of therapy outcomes for patients, such as those with epilepsy where objective identification of patient seizure activity and rhythms might help guide therapy optimization. The system is configurable for other disease states such as Parkinson's disease and mood disorders.

MorpheusNet: Resource efficient sleep stage classifier for embedded on-line systems.

Sleep Stage Classification (SSC) is a labor-intensive task, requiring experts to examine hours of electrophysiological recordings for manual classification. This is a limiting factor when it comes to leveraging sleep stages for therapeutic purposes. With increasing affordability and expansion of wearable devices, automating SSC may enable deployment of sleep-based therapies at scale. Deep Learning has gained increasing attention as a potential method to automate this process. Previous research has shown accuracy comparable to manual expert scores. However, previous approaches require sizable amount of memory and computational resources. This constrains the ability to classify in real time and deploy models on the edge. To address this gap, we aim to provide a model capable of predicting sleep stages in real-time, without requiring access to external computational sources (e.g., mobile phone, cloud). The algorithm is power efficient to enable use on embedded battery powered systems. Our compact sleep stage classifier can be deployed on most off-the-shelf microcontrollers (MCU) with constrained hardware settings. This is due to the memory footprint of our approach requiring significantly fewer operations. The model was tested on three publicly available data bases and achieved performance comparable to the state of the art, whilst reducing model complexity by orders of magnitude (up to 280 times smaller compared to state of the art). We further optimized the model with quantization of parameters to 8 bits with only an average drop of 0.95% in accuracy. When implemented in firmware, the quantized model achieves a latency of 1.6 seconds on an Arm Cortex-M4 processor, allowing its use for on-line SSC-based therapies.

Computationally efficient neural network classifiers for next generation closed loop neuromodulation therapy - a case study in epilepsy.

This work explores the potential utility of neural network classifiers for real- time classification of field-potential based biomarkers in next-generation responsive neuromodulation systems. Compared to classical filter-based classifiers, neural networks offer an ease of patient-specific parameter tuning, promising to reduce the burden of programming on clinicians. The paper explores a compact, feed - forward neural network architecture of only dozens of units for seizure-state classification in refractory epilepsy. The proposed classifier offers comparable accuracy to filter- classifiers on clinician-labeled data, while reducing detection latency. As a trade-off to classical methods, the paper focuses on keeping the complexity of the architecture minimal, to accommodate the on-board computational constraints of implantable pulse generator systems. Clinical relevance-A neural network-based classifier is presented for responsive neurostimulation, with comparable accuracy to classical methods at reduced latency.