NeuralTree: A 256-Channel 0.227-µJ/Class Versatile Neural Activity Classification and Closed-Loop Neuromodulation SoC.

Closed-loop neural interfaces with on-chip machine learning can detect and suppress disease symptoms in neurological disorders or restore lost functions in paralyzed patients. While high-density neural recording can provide rich neural activity information for accurate disease-state detection, existing systems have low channel counts and poor scalability, which could limit their therapeutic efficacy. This work presents a highly scalable and versatile closed-loop neural interface SoC that can overcome these limitations. A 256-channel time-division multiplexed (TDM) front-end with a two-step fast-settling mixed-signal DC servo loop (DSL) is proposed to record high-spatial-resolution neural activity and perform channel-selective brain-state inference. A tree-structured neural network (NeuralTree) classification processor extracts a rich set of neural biomarkers in a patient- and disease-specific manner. Trained with an energy-aware learning algorithm, the NeuralTree classifier detects the symptoms of underlying disorders (e.g., epilepsy and movement disorders) at an optimal energy-accuracy trade-off. A 16-channel high-voltage (HV) compliant neurostimulator closes the therapeutic loop by delivering charge-balanced biphasic current pulses to the brain. The proposed SoC was fabricated in 65nm CMOS and achieved a 0.227µJ/class energy efficiency in a compact area of 0.014mm2/channel. The SoC was extensively verified on human electroencephalography (EEG) and intracranial EEG (iEEG) epilepsy datasets, obtaining 95.6%/94% sensitivity and 96.8%/96.9% specificity, respectively. In-vivo neural recordings using soft µECoG arrays and multi-domain biomarker extraction were further performed on a rat model of epilepsy. In addition, for the first time in literature, on-chip classification of rest-state tremor in Parkinson's disease (PD) from human local field potentials (LFPs) was demonstrated.

Migraine classification using somatosensory evoked potentials.

OBJECTIVE: The automatic detection of migraine states using electrophysiological recordings may play a key role in migraine diagnosis and early treatment. Migraineurs are characterized by a deficit of habituation in cortical information processing, causing abnormal changes of somatosensory evoked potentials. Here, we propose a machine learning approach to utilize somatosensory evoked potential-based biomarkers for migraine classification in a noninvasive setting. METHODS: Forty-two migraine patients, including 29 interictal and 13 ictal, were recruited and compared with 15 healthy volunteers of similar age and gender distribution. The right median nerve somatosensory evoked potentials were collected from all subjects. State-of-the-art machine learning algorithms including random forest, extreme gradient-boosting trees, support vector machines, K-nearest neighbors, multilayer perceptron, linear discriminant analysis, and logistic regression were used for classification and were built upon somatosensory evoked potential features in time and frequency domains. A feature selection method was employed to assess the contribution of features and compare it with previous clinical findings, and to build an optimal feature set by removing redundant features. RESULTS: Using a set of relevant features and different machine learning models, accuracies ranging from 51.2% to 72.4% were achieved for the healthy volunteers-ictal-interictal classification task. Following model and feature selection, we successfully separated the three groups of subjects with an accuracy of 89.7% for the healthy volunteers-ictal, 88.7% for healthy volunteers-interictal, 80.2% for ictal-interictal, and 73.3% for healthy volunteers-ictal-interictal classification tasks, respectively. CONCLUSION: Our proposed model suggests the potential use of somatosensory evoked potentials as a prominent and reliable signal in migraine classification. This non-invasive somatosensory evoked potential-based classification system offers the potential to reliably separate migraine patients in ictal and interictal states from healthy controls.

Three-dimensional focusing through scattering media using conjugate adaptive optics with remote focusing (CAORF).

The small correction volume for conventional wavefront shaping methods limits their application in biological imaging through scattering media. We demonstrate large volume wavefront shaping through a scattering layer with a single correction by conjugate adaptive optics and remote focusing (CAORF). The remote focusing module can maintain the conjugation between the adaptive optical (AO) element and the scattering layer during three-dimensional scanning. This new configuration provides a wider correction volume by better utilization of the memory effect in a fast three-dimensional laser scanning microscope. Our results show that the proposed system can provide 10 times wider axial field of view compared with a conventional conjugate AO system when 16,384 segments are used on a spatial light modulator. We also demonstrate three-dimensional fluorescence imaging, multi-spot patterning through a scattering layer and two-photon imaging through mouse skull tissue.

Numerical studies of focal modulation microscopy in high-NA system.

High spatial resolution with deep imaging penetration depth is the main advantage of focal modulation microscopy (FMM). This paper investigates effects of polarization on FMM in a high-NA system based on vectorial diffraction theory. Compared with confocal microscopy, FMM shows a 20.1% improvement in axial resolution. The performance of different polarization patterns is also discussed numerically. The study on polarization modulation may provide a new way to obtain a tighter focal spot.

Stripe-shaped apertures in confocal microscopy.

We have theoretically verified that, compared with the aperture shapes of previous research, combining two stripe-shaped apertures in a confocal microscope with a finite-sized pinhole improves the axial resolution to a certain extent. Because different stripe shapes cause different effects, we also investigated the relationships among resolution, shapes, pinhole size, and the signal-to-background ratio.

Fast and accurate decoding of finger movements from ECoG through Riemannian features and modern machine learning techniques.

Objective.Accurate decoding of individual finger movements is crucial for advanced prosthetic control. In this work, we introduce the use of Riemannian-space features and temporal dynamics of electrocorticography (ECoG) signal combined with modern machine learning (ML) tools to improve the motor decoding accuracy at the level of individual fingers.Approach.We selected a set of informative biomarkers that correlated with finger movements and evaluated the performance of state-of-the-art ML algorithms on the brain-computer interface (BCI) competition IV dataset (ECoG, three subjects) and a second ECoG dataset with a similar recording paradigm (Stanford, nine subjects). We further explored the temporal concatenation of features to effectively capture the history of ECoG signal, which led to a significant improvement over single-epoch decoding in both classification (p < 0.01) and regression tasks (p < 0.01).Main results.Using feature concatenation and gradient boosted trees (the top-performing model), we achieved a classification accuracy of 77.0% in detecting individual finger movements (six-class task, including rest state), improving over the state-of-the-art conditional random fields by 11.7% on the three BCI competition subjects. In continuous decoding of movement trajectory, our approach resulted in an average Pearson's correlation coefficient (r) of 0.537 across subjects and fingers, outperforming both the BCI competition winner and the state-of-the-art approach reported on the same dataset (CNN + LSTM). Furthermore, our proposed method features a low time complexity, with only<17.2 s required for training and<50 ms for inference. This enables about 250× speed-up in training compared to previously reported deep learning method with state-of-the-art performance.Significance.The proposed techniques enable fast, reliable, and high-performance prosthetic control through minimally-invasive cortical signals.

Closed-Loop Neural Prostheses With On-Chip Intelligence: A Review and a Low-Latency Machine Learning Model for Brain State Detection.

The application of closed-loop approaches in systems neuroscience and therapeutic stimulation holds great promise for revolutionizing our understanding of the brain and for developing novel neuromodulation therapies to restore lost functions. Neural prostheses capable of multi-channel neural recording, on-site signal processing, rapid symptom detection, and closed-loop stimulation are critical to enabling such novel treatments. However, the existing closed-loop neuromodulation devices are too simplistic and lack sufficient on-chip processing and intelligence. In this paper, we first discuss both commercial and investigational closed-loop neuromodulation devices for brain disorders. Next, we review state-of-the-art neural prostheses with on-chip machine learning, focusing on application-specific integrated circuits (ASIC). System requirements, performance and hardware comparisons, design trade-offs, and hardware optimization techniques are discussed. To facilitate a fair comparison and guide design choices among various on-chip classifiers, we propose a new energy-area (E-A) efficiency figure of merit that evaluates hardware efficiency and multi-channel scalability. Finally, we present several techniques to improve the key design metrics of tree-based on-chip classifiers, both in the context of ensemble methods and oblique structures. A novel Depth-Variant Tree Ensemble (DVTE) is proposed to reduce processing latency (e.g., by 2.5× on seizure detection task). We further develop a cost-aware learning approach to jointly optimize the power and latency metrics. We show that algorithm-hardware co-design enables the energy- and memory-optimized design of tree-based models, while preserving a high accuracy and low latency. Furthermore, we show that our proposed tree-based models feature a highly interpretable decision process that is essential for safety-critical applications such as closed-loop stimulation.

ResOT: Resource-Efficient Oblique Trees for Neural Signal Classification.

Classifiers that can be implemented on chip with minimal computational and memory resources are essential for edge computing in emerging applications such as medical and IoT devices. This paper introduces a machine learning model based on oblique decision trees to enable resource-efficient classification on a neural implant. By integrating model compression with probabilistic routing and implementing cost-aware learning, our proposed model could significantly reduce the memory and hardware cost compared to state-of-the-art models, while maintaining the classification accuracy. We trained the resource-efficient oblique tree with power-efficient regularization (ResOT-PE) on three neural classification tasks to evaluate the performance, memory, and hardware requirements. On seizure detection task, we were able to reduce the model size by 3.4× and the feature extraction cost by 14.6× compared to the ensemble of boosted trees, using the intracranial EEG from 10 epilepsy patients. In a second experiment, we tested the ResOT-PE model on tremor detection for Parkinson's disease, using the local field potentials from 12 patients implanted with a deep-brain stimulation (DBS) device. We achieved a comparable classification performance as the state-of-the-art boosted tree ensemble, while reducing the model size and feature extraction cost by 10.6× and 6.8×, respectively. We also tested on a 6-class finger movement detection task using ECoG recordings from 9 subjects, reducing the model size by 17.6× and feature computation cost by 5.1×. The proposed model can enable a low-power and memory-efficient implementation of classifiers for real-time neurological disease detection and motor decoding.