Murat Akcakaya

Noninvasive Brain–Computer Interfaces for Augmentative and Alternative Communication

Brain-computer interfaces (BCIs) promise to provide a novel access channel for assistive technologies, including augmentative and alternative communication (AAC) systems, to people with severe speech and physical impairments (SSPI). Research on the subject has been accelerating significantly in the last decade and the research community took great strides toward making BCI-AAC a practical reality to individuals with SSPI. Nevertheless, the end goal has still not been reached and there is much work to be done to produce real-world-worthy systems that can be comfortably, conveniently, and reliably used by individuals with SSPI with help from their families and care givers who will need to maintain, setup, and debug the systems at home. This paper reviews reports in the BCI field that aim at AAC as the application domain with a consideration on both technical and clinical aspects.

MIMO Radar Detection and Adaptive Design Under a Phase Synchronization Mismatch

We consider the problem of target detection for multi-input multi-output radar with widely separated antennas in the presence of a phase synchronization mismatch between the transmitter and receiver pairs. Such mismatch often occurs due to imperfect knowledge of the locations and local oscillator characteristics of the antennas. First, we introduce a data model using a von Mises distribution to represent the phase error terms. Then, we employ an expectation-maximization algorithm to estimate the error distribution parameter, target returns, and noise variance. We develop a generalized likelihood ratio test target detector using these estimates. Based on the mutual information between the radar measurements and received target returns (and hence the phase error), we propose an algorithm to adaptively distribute the total transmitted energy among the transmitters. Using Monte Carlo simulations, we demonstrate that the adaptive energy allocation, increase in the phase information, and realistic measurement modeling improve the detection performance.

MIMO Radar Sensitivity Analysis for Target Detection

We consider the effect of imperfect separability in the received signals on the detection performance of multi-input multi-output (MIMO) radar with widely separated antennas. The mutual orthogonality among the received signals is often assumed but cannot be achieved in practice for all Doppler and delay pairs. We introduce a data model considering the correlation among the data from different transmitter-receiver pairs as unknown parameters. Based on the expectation maximization algorithm, we propose a method to estimate the target, correlation, and noise parameters. We then use the estimates of these parameters to develop a statistical decision test. Employing the asymptotic statistical characteristics and the numerical performance of the test, we analyze the sensitivity of the MIMO radar with respect to changes in the cross-correlation levels of the measurements. We demonstrate the effect of the increase in the correlation among the received signals from different transmitters on the detection performance.

Adaptive MIMO Radar Design and Detection in Compound-Gaussian Clutter

Multiple-input multiple-output (MIMO) radars with widely separated transmitters and receivers are useful to discriminate a target from clutter using the spatial diversity of the scatterers in the illuminated scene. We consider the detection of targets in compound-Gaussian clutter, describing heavy-tailed clutter distributions fitting high-resolution and/or low-grazing-angle radars in the presence of sea or foliage clutter. First we introduce a data model using the inverse gamma distribution to represent the clutter texture. Then we apply the parameter-expanded expectation-maximization (PX-EM) algorithm to estimate the clutter texture and speckle as well as the target parameters. We develop a statistical decision test using these estimates and demonstrate its statistical characteristics. Based on the statistical characteristics of this test, we propose an algorithm to adaptively distribute the transmitted energy among the transmitters and maximize the detection performance. We demonstrate the advantages of the MIMO setup and adaptive energy allocation in target detection in the presence of compound-Gaussian clutter using Monte Carlo (MC) simulations.

MIMO radar detection of targets in compound-Gaussian clutter

Multiple-input multiple-output (MIMO) radars with widely-separated transmitters and receivers are useful to discriminate a target from clutter using the spatial diversity of the scatterers in the illuminated scene. We consider the detection of targets in compound-Gaussian clutter fitting for example scatterers with heavy-tailed distributions for high-resolution and/or low-grazing-angle radars in the presence of sea or foliage clutter. First, we introduce a data model using the inverse gamma distribution to represent the clutter texture. Then, we apply the parameter-expanded expectation-maximization (PX-EM) algorithm to estimate the clutter texture and speckle, as well as the target parameters.We develop a generalized likelihood ratio (GLR) test target detector using the estimates and show the advantages of MIMO using Monte Carlo simulations.

Performance analysis of the Ormia ochracea’s coupled ears

The Ormia ochracea is able to locate a crickets mating call despite the small distance between its ears compared with the wavelength. This phenomenon has been explained by the mechanical coupling between the ears. In this paper, it is first shown that the coupling enhances the differences in times of arrival and frequency responses of the ears to the incoming source signals. Then, the accuracy of estimating directions of arrival (DOAs) by the O. ochracea is analyzed by computing the Cramér-Rao bound (CRB). The differential equations of the mechanical model are rewritten in state space and its frequency response is calculated. Using the spectral properties of the system, the CRB for multiple stochastic sources with unknown directions and spectra is asymptotically computed. Numerical examples compare the CRB for the coupled and the uncoupled cases, illustrating the effect of the coupling on reducing the errors in estimating the DOAs.

Language-Model Assisted Brain Computer Interface for Typing: A Comparison of Matrix and Rapid Serial Visual Presentation

Noninvasive electroencephalography (EEG)-based brain-computer interfaces (BCIs) popularly utilize event-related potential (ERP) for intent detection. Specifically, for EEG-based BCI typing systems, different symbol presentation paradigms have been utilized to induce ERPs. In this manuscript, through an experimental study, we assess the speed, recorded signal quality, and system accuracy of a language-model-assisted BCI typing system using three different presentation paradigms: a 4 × 7 matrix paradigm of a 28-character alphabet with row-column presentation (RCP) and single-character presentation (SCP), and rapid serial visual presentation (RSVP) of the same. Our analyses show that signal quality and classification accuracy are comparable between the two visual stimulus presentation paradigms. In addition, we observe that while the matrix-based paradigm can be generally employed with lower inter-trial-interval (ITI) values, the best presentation paradigm and ITI value configuration is user dependent. This potentially warrants offering both presentation paradigms and variable ITI options to users of BCI typing systems.

Moving towards a real-time system for automatically recognizing stereotypical motor movements in individuals on the autism spectrum using wireless accelerometry

This paper extends previous work automatically detecting stereotypical motor movements (SMM) in individuals on the autism spectrum. Using three-axis accelerometer data obtained through wearable wireless sensors, we compare recognition results for two different classifiers -- Support Vector Machine and Decision Tree -- in combination with different feature sets based on time-frequency characteristics of accelerometer data. We use data collected from six individuals on the autism spectrum who participated in two different studies conducted three years apart in classroom settings, and observe an average accuracy across all participants over time ranging from 81.2% (TPR: 0.91; FPR: 0.21) to 99.1% (TPR: 0.99; FPR: 0.01) for all combinations of classifiers and feature sets. We also provide analyses of kinematic parameters associated with observed movements in an attempt to explain classifier-feature specific performance. Based on our results, we conclude that real-time, person-dependent, adaptive algorithms are needed in order to accurately and consistently measure SMM automatically in individuals on the autism spectrum over time in real-word settings.

A Bayesian Framework for Intent Detection and Stimulation Selection in SSVEP BCIs

Currently, many Brain Computer Interfaces (BCI) classifiers output point estimates of user intent which make it difficult to incorporate context prior information or assign a principled confidence measurement to a decision. We propose a Bayesian framework to extend current Steady State Visually Evoked Potential (SSVEP) classifiers to a maximum a posteriori (MAP) classifiers by using a Kernel Density Estimate (KDE) to learn the distribution of features conditioned on stimulation class. To demonstrate our framework we extend Canonical Correlation Analysis (CCA) and Power Spectral Density (PSD) style methods. Traditionally, in either example, the class is estimated as the class associated with the maximum feature. Our framework increases performance by relaxing the assumption that a stimulation classs sample often maximizes its class-associated feature. Further, by leveraging the KDE, we present a method which estimates the performance of a classifier under different stimulation frequency sets. Using this, we optimize the selection of stimulation frequencies from those present in a training set.

A Robust Fusion Algorithm for Sensor Failure

Accurate multimodal and multisensor detection of a target phenomenon requires knowledge of probabilistic sensor characteristics to determine an appropriate fusion rule which optimizes an objective of interest, traditionally the expected Bayesian risk. However, a particular sensor characteristic can change online, introducing unaccounted additional risk to the fusion rule that was based on assumed sensor specifications. To mitigate such changes, we propose a sensor-failure-robust fusion rule assuming that only first order characteristics of a probabilistic sensor failure model are known. Under this failure model, we compute the expected Bayesian risk and minimize this risk to develop the proposed fusion method.