Intra-day signal instabilities affect decoding performance in an intracortical neural interface system

Perge, János A.; Homer, Mark; Malik, W.Q.; Cash, Sydney S.; Eskandar, Emad N.; Friehs, Gerhard; Donoghue, John P.; Hochberg, Leigh R.

doi:10.1088/1741-2560/10/3/036004

Cited by 207 publications

(216 citation statements)

References 65 publications

(118 reference statements)

Supporting

Mentioning

210

Contrasting

Unclassified

Order By: Relevance

“…This suggests that self-recalibrating classifiers can be simplified to capture only between-day changes, as was done in this work. Recently Perge et al investigated intracortical BCI signal stability within the context of decoding continuous control signals (i.e., regression) in closed loop human experiments (Perge et al, 2013). While they do not explicitly compare within-day changes to between-day changes in BCI signals, they do report reduced decoding accuracy due to within-day changes in neural spike rates.…”

Section: Discussionmentioning

confidence: 99%

Self-recalibrating classifiers for intracortical brain–computer interfaces

et al. 2014

View full text Add to dashboard Cite

Objective Intracortical brain-computer interface (BCI) decoders are typically retrained daily to maintain stable performance. Self-recalibrating decoders aim to remove the burden this may present in the clinic by training themselves autonomously during normal use but have only been developed for continuous control. Here we address the problem for discrete decoding (classifiers). Approach We recorded threshold crossings from 96-electrode arrays implanted in the motor cortex of two rhesus macaques performing center-out reaches in 7 directions over 41 and 36 separate days spanning 48 and 58 days in total for offline analysis. Main results We show that for the purposes of developing a self-recalibrating classifier, tuning parameters can be considered as fixed within days and that parameters on the same electrode move up and down together between days. Further, drift is constrained across time, which is reflected in the performance of a standard classifier which does not progressively worsen if it is not retrained daily, though overall performance is reduced by more than 10% compared to a daily retrained classifier. Two novel self-recalibrating classifiers produce a ~15% increase in classification accuracy over that achieved by the non-retrained classifier to nearly recover the performance of the daily retrained classifier. Significance We believe that the development of classifiers that require no daily retraining will accelerate the clinical translation of BCI systems. Future work should test these results in a closed loop setting.

show abstract

Section: Discussionmentioning

confidence: 99%

Self-recalibrating classifiers for intracortical brain–computer interfaces

et al. 2014

View full text Add to dashboard Cite

show abstract

“…Recently, an invasive BMI was built for a monkey to control a lower-limb exoskeleton in real time (Vouga et al 2017). However, these approaches face the risk of surgical complications and infections, short-term and long-term signal instabilities that degrade neural decoding of intent (Perge et al 2013), and the challenge of maintaining stable chronic recordings (Meng et al 2016). The added risks associated with the testing of BMI systems for lower-limb applications have perhaps precluded the development of invasive BMIs for lower limb applications in humans.…”

Section: Introductionmentioning

confidence: 99%

Brain–machine interfaces for controlling lower-limb powered robotic systems

et al. 2018

View full text Add to dashboard Cite

Objective. Lower-limb, powered robotics systems such as exoskeletons and orthoses have emerged as novel robotic interventions to assist or rehabilitate people with walking disabilities. These devices are generally controlled by certain physical maneuvers, for example pressing buttons or shifting body weight. Although effective, these control schemes are not what humans naturally use. The usability and clinical relevance of these robotics systems could be further enhanced by brain–machine interfaces (BMIs). A number of preliminary studies have been published on this topic, but a systematic understanding of the experimental design, tasks, and performance of BMI-exoskeleton systems for restoration of gait is lacking. Approach. To address this gap, we applied standard systematic review methodology for a literature search in PubMed and EMBASE databases and identified 11 studies involving BMI-robotics systems. The devices, user population, input and output of the BMIs and robot systems respectively, neural features, decoders, denoising techniques, and system performance were reviewed and compared. Main results. Results showed BMIs classifying walk versus stand tasks are the most common. The results also indicate that electroencephalography (EEG) is the only recording method for humans. Performance was not clearly presented in most of the studies. Several challenges were summarized, including EEG denoising, safety, responsiveness and others. Significance. We conclude that lower-body powered exoskeletons with automated gait intention detection based on BMIs open new possibilities in the assistance and rehabilitation fields, although the current performance, clinical benefits and several key challenging issues indicate that additional research and development is required to deploy these systems in the clinic and at home. Moreover, rigorous EEG denoising techniques, suitable performance metrics, consistent trial reporting, and more clinical trials are needed to advance the field.

show abstract

“…Evidence suggests that mechanical mismatch is an important reason leading to abrupt and chronically unstable interfaces within the brain 4,22 . For example, motion of skull-affixed rigid probes in chronic experiments can induce shear stresses and lead to tissue scarring 13,23 , and thereby compromise the stability of recorded signals on the time scale of weeks to months 4,24, 25 .…”

mentioning

confidence: 99%

Three-dimensional macroporous nanoelectronic networks as minimally invasive brain probes

et al. 2015

View full text Add to dashboard Cite

Abstract. Direct electrical recording and stimulation of neural activity using microfabricated silicon and metal micro-wire probes have contributed extensively to basic neuroscience and therapeutic applications; however, the dimensional and mechanical mismatch of these probes with the brain tissue limits their stability in chronic implants and decreases the neuron-device contact. Here, we demonstrate the realization of a 3D macroporous nanoelectronic brain probe that combines ultra-flexibility and subcellular feature sizes to overcome these limitations. Built-in strains controlling the local geometry of the macroporous devices are designed to optimize the neuron/probe interface and to promote integration with the brain tissue while introducing minimal mechanical perturbation. The ultra-flexible probes were implanted frozen into rodent brains and used to record multiplexed local field potentials (LFPs) and single-unit action potentials from the somatosensory cortex. Significantly, histology analysis revealed filling-in of neural tissue through the macroporous network and attractive neuron-probe interactions, consistent with long-term biocompatibility of the device.Currently, there is intense interest in the development of materials and electronic devices that can extend and/or provide new capabilities for probing neural circuitry and afford long-term minimally-invasive brain-electronics interfaces 1,2,3,4 . Conventional brain probes have contributed extensively to basic neuroscience 5,6 and therapeutic applications 7,8,9,10 , although they suffer from chronic stability and poor neuron-device contacts 4,11,12,13 . Recent studies of smaller 14,15 and more flexible 16,17 probes suggest that addressing size and mechanical factors could help overcome current limitations. 3The most common neural electrical probes are fabricated from metal 18 and silicon 19,20 , materials that have very different structural and mechanical properties compared to brain tissue 21 .Evidence suggests that mechanical mismatch is an important reason leading to abrupt and chronically unstable interfaces within the brain 4,22 . For example, motion of skull-affixed rigid probes in chronic experiments can induce shear stresses and lead to tissue scarring 13,23 , and thereby compromise the stability of recorded signals on the time scale of weeks to months 4,24, 25 .More recent work has shown that flexible probes fabricated on polymer substrates 12,17 and smaller-sized probes 11,14 can reduce deleterious tissue response. More generally, there has also been effort developing flexible bioelectronics 26, 27, 28 and nanoscale devices for single cell recording 29, 30 . We have also shown that 3D macroporous electronic device arrays can function as a scaffold for and allow for 3D interpenetration of cultured neuron cell networks without an adverse effect on cell viability 31 , and such networks can be injected by syringe through needles into materials, including brain tissue 32 . In the latter case, it remains challenging to make electrical in...

show abstract

Intra-day signal instabilities affect decoding performance in an intracortical neural interface system

Cited by 207 publications

References 65 publications

Self-recalibrating classifiers for intracortical brain–computer interfaces

Self-recalibrating classifiers for intracortical brain–computer interfaces

Brain–machine interfaces for controlling lower-limb powered robotic systems

Three-dimensional macroporous nanoelectronic networks as minimally invasive brain probes

Contact Info

Product

Resources

About