Nuri F. Ince

Beta-Band Activity during Motor Planning Reflects Response Uncertainty

It has been known for many years that the power of beta-band oscillatory activity in motor-related brain regions decreases during the preparation and execution of voluntary movements. However, it is not clear yet whether the amplitude of this desynchronization is modulated by any parameter of the motor task. Here, we examined whether the degree of uncertainty about the upcoming movement direction modulated beta-band desynchronization during motor preparation. To this end, we recorded whole-head neuromagnetic signals while human subjects performed an instructed-delay reaching task with one, two, or three possible target directions. We found that the reduction of power of beta-band activity (16–28 Hz) during motor preparation was scaled relative to directional uncertainty. Furthermore, we show that the change of beta-band power correlates with the change of latency of response associated with response uncertainty. Finally, we show that the main source of beta-band desynchronization was located in the peri-Rolandic region. The results establish directional uncertainty as an important determinant of beta-band power during motor preparation and indicate that neural activity in the sensorimotor cortex during motor preparation covaries with directional uncertainty.

High Accuracy Decoding of Movement Target Direction in Non-Human Primates Based on Common Spatial Patterns of Local Field Potentials

Background The current development of brain-machine interface technology is limited, among other factors, by concerns about the long-term stability of single- and multi-unit neural signals. In addition, the understanding of the relation between potentially more stable neural signals, such as local field potentials, and motor behavior is still in its early stages. Methodology/Principal Findings We tested the hypothesis that spatial correlation patterns of neural data can be used to decode movement target direction. In particular, we examined local field potentials (LFP), which are thought to be more stable over time than single unit activity (SUA). Using LFP recordings from chronically implanted electrodes in the dorsal premotor and primary motor cortex of non-human primates trained to make arm movements in different directions, we made the following observations: (i) it is possible to decode movement target direction with high fidelity from the spatial correlation patterns of neural activity in both primary motor (M1) and dorsal premotor cortex (PMd); (ii) the decoding accuracy of LFP was similar to the decoding accuracy obtained with the set of SUA recorded simultaneously; (iii) directional information varied with the LFP frequency sub-band, being greater in low (0.3–4 Hz) and high (48–200 Hz) frequency bands than in intermediate bands; (iv) the amount of directional information was similar in M1 and PMd; (v) reliable decoding was achieved well in advance of movement onset; and (vi) LFP were relatively stable over a period of one week. Conclusions/Significance The results demonstrate that the spatial correlation patterns of LFP signals can be used to decode movement target direction. This finding suggests that parameters of movement, such as target direction, have a stable spatial distribution within primary motor and dorsal premotor cortex, which may be used for brain-machine interfaces.

Adapting subject specific motor imagery EEG patterns in space–time–frequency for a brain computer interface

Abstract In this paper we propose a new technique that adaptively extracts subject specific motor imagery related EEG patterns in the space–time–frequency plane for single trial classification. The proposed approach requires no prior knowledge of reactive frequency bands, their temporal behavior or cortical locations. For a given electrode array, it finds all these parameters by constructing electrode adaptive time–frequency segmentations that are optimized for discrimination. This is accomplished first by segmenting the EEG along the time axis with Local Cosine Packets. Next the most discriminant frequency subbands are selected in each time segment with a frequency axis clustering algorithm to achieve time and frequency band adaptation individually. Finally the subject adapted features are sorted according to their discrimination power to reduce dimensionality and the top subset is used for final classification. We provide experimental results for 5 subjects of the BCI competition 2005 dataset IVa to show the superior performance of the proposed method. In particular, we demonstrate that by using a linear support vector machine as a classifier, the classification accuracy of the proposed algorithm varied between 90.5% and 99.7% and the average classification accuracy was 96%.

Selection of optimal programming contacts based on local field potential recordings from subthalamic nucleus in patients with Parkinson's disease.

BACKGROUNDIn the United States, the most commonly used surgical treatment for patients with Parkinsons disease is the implantation of deep brain stimulation (DBS) electrodes within the subthalamic nucleus. However, DBS device programming remains difficult and is a possible source of decreased efficacy. OBJECTIVEWe investigated the relationship between local field potential (LFP) activities in the subthalamic nucleus and the therapeutic response to programming. METHODSWe recorded LFPs with macroelectrodes placed unilaterally for DBS in 4 PD patients, 3 weeks after implantation, before the start of log-term DBS. Power-frequency spectra were calculated for each of 7 possible electrode contacts or contact pairs, over multiple 5- to 10-minute quiet waking epochs and over 30-second epochs during hand movements. Subsequently, DBS devices were programmed, with testing to determine which electrode contacts or contact pairs demonstrated optimal therapeutic efficacy. RESULTSFor each patient, the contact pair found to provide optimal efficacy was associated with the highest energy in the β (13–32 Hz) and γ (48–220 Hz) bands during postoperative LFP recordings at rest and during hand movements. Activities in other frequency bands did not show significant correlations between LFP power and optimal electrode contacts. CONCLUSIONPostoperative subband analysis of LFP recordings in β and γ frequency ranges may be used to select optimal electrode contacts. These results indicate that LFP recordings from implanted DBS electrodes can provide important clues to guide the optimization of DBS therapy in individual patients.

Detection of early morning daily activities with static home and wearable wireless sensors

This paper describes a flexible, cost-effective, wireless in-home activity monitoring system for assisting patients with cognitive impairments due to traumatic brain injury (TBI). The system locates the subject with fixed home sensors and classifies early morning bathroom activities of daily living with a wearable wireless accelerometer. The system extracts time- and frequency-domain features from the accelerometer data and classifies these features with a hybrid classifier that combines Gaussian mixture models and a finite state machine. In particular, the paper establishes that despite similarities between early morning bathroom activities of daily living, it is possible to detect and classify these activities with high accuracy. It also discusses system training and provides data to show that with proper feature selection, accurate detection and classification are possible for any subject with no subject specific training.

Classification of schizophrenia with spectro-temporo-spatial MEG patterns in working memory.

OBJECTIVE To investigate whether temporo-spatial patterns of brain oscillations extracted from multichannel magnetoencephalogram (MEG) recordings in a working memory task can be used successfully as a biometric marker to discriminate between healthy control subjects and patients with schizophrenia. METHODS Five letters appearing sequentially on a screen had to be memorized. The letters constituted a word in one condition and a pronounceable non-word in the other. Power changes of 248 channel MEG data were extracted in frequency sub-bands and a two-step filter and search algorithm was used to select informative features that discriminated patients and controls. RESULTS The discrimination between patients and controls was greater in the word condition than in the non-word condition. Furthermore, in the word condition, the most discriminant patterns were extracted in delta (1-4 Hz), alpha (12-16 Hz) and beta (16-24 Hz) frequency bands. These features were located in the left dorso-frontal, occipital and left fronto-temporal, respectively. CONCLUSION The analysis of the oscillatory patterns of MEG recordings in the working memory task provided a high level of correct classification of patients and controls. SIGNIFICANCE We show, using a newly developed algorithm, that the temporo-spatial patterns of brain oscillations can be used as biometric marker that discriminate schizophrenia patients and healthy controls.

Temporospatial Characterization of Brain Oscillations (TSCBO) Associated with Subprocesses of Verbal Working Memory in Schizophrenia

The studies of the neural correlates of verbal working memory in schizophrenia are somewhat inconsistent. This could be related to experimental paradigms that engage differentially working memory components or methodological limitations in terms of characterization of brain activity. Magnetoencephalographic recordings were obtained on 10 schizophrenia patients and 11 healthy controls while performing a modified Sternberg paradigm to investigate subprocesses of verbal working memory. A new method for temporospatial characterization of brain oscillations was applied to whole head recordings and a 1–48 Hz frequency range. Patients differed from controls in event-related synchronization/desynchronization (ERS/ERD) patterns during the encode phase, the mid-maintain phase, and the end of the maintain phase. During the encode phase, patients did not show 1–4 Hz ERS in the left anterior frontal and left parietal lobes. In the mid-maintain phase, the left anterior frontal and left parietal lobes 1–4 Hz ERS, and the bilateral occipital lobes 8–32 Hz ERS were not observed in patients. At the end of the maintain phase, patients did not exhibit 12–48 Hz ERD in the left frontal and parietal lobes. The behavioral data showed reduced primacy effect In schizophrenia, the encode and maintain subprocesses were associated with less ERS and less ERD, respectively. These ERS/ERD abnormalities had specificity in terms of frequency and spatial location. Less ERD reflects reduced complexity of the neural activity, while reduced ERS reflects failure of the neural systems to resume idle state. The impaired primacy effect appears related to specific ERS/ERD patterns in the encode and maintain phases.

Selection of spectro-temporal patterns in multichannel MEG with support vector machines for schizophrenia classification

We present a new framework for the diagnosis of schizophrenia based on the spectro-temporal patterns selected by a support vector machine from multichannel magnetoencephalogram (MEG) recordings in a verbal working memory task. In the experimental paradigm, five letters appearing sequentially on a screen were memorized by subjects. The letters constituted a word in one condition and a pronounceable nonword in the other. Power changes were extracted as features in frequency subbands of 248 channel MEG data to form a rich feature dictionary. A support vector machine has been used to select a small subset of features with recursive feature elimination technique (SVM-RFE) and the reduced subset was used for classification. We note that the discrimination between patients and controls in the word condition was higher than in the non-word condition (91.8% vs 83.8%). Furthermore, in the word condition, the most discriminant patterns were extracted in delta (1–4 Hz), theta (4–8Hz) and alpha (12–16 Hz) frequency bands. We note that these features were located around the left frontal, left temporal and occipital areas, respectively. Our results indicate that the proposed approach can quantify discriminative neural patterns associated to a functional task in spatial, spectral and temporal domain. Moreover these features provide interpretable information to the medical expert about physiological basis of the illness and can be effectively used as a biometric marker to recognize schizophrenia in clinical practice.

Classification of hazelnut kernels by using impact acoustic time-frequency patterns

Hazelnuts with damaged or cracked shells are more prone to infection with aflatoxin producing molds (Aspergillus flavus). These molds can cause cancer. In this study, we introduce a new approach that separates damaged/cracked hazelnut kernels from good ones by using time-frequency features obtained from impact acoustic signals. The proposed technique requires no prior knowledge of the relevant time and frequency locations. In an offline step, the algorithm adaptively segments impact signals from a training data set in time using local cosine packet analysis and a Kullback-Leibler criterion to assess the discrimination power of different segmentations. In each resulting time segment, the signal is further decomposed into subbands using an undecimated wavelet transform. The most discriminative subbands are selected according to the Euclidean distance between the cumulative probability distributions of the corresponding subband coefficients. The most discriminative subbands are fed into a linear discriminant analysis classifier. In the online classification step, the algorithm simply computes the learned features from the observed signal and feeds them to the linear discriminant analysis (LDA) classifier. The algorithm achieved a throughput rate of 45 nuts/s and a classification accuracy of 96% with the 30 most discriminative features, a higher rate than those provided with prior methods.

Averaged acoustic emission events for accurate damage localization

Localizing micro cracks in critical components is crucial in the field of continuous structural health monitoring. In this paper, we utilize several signal processing and machine learning techniques such as hierarchical clustering and support vector machines (SVM) to process multisensor acoustic emission (AE) data generated by the inception and propagation of cracks. We present preliminary laboratory results that explore the pairwise event correlation of AE waveforms generated in the process of controlled crack propagation, and use these characteristics for clustering AE. By averaging the AE events within each cluster obtained from hierarchical clustering, we compute super-acoustics with higher signal to noise ratio (SNR) and use them in the second step of our analysis for calculating the time of arrival information (TOA) for crack localization. We utilize a SVM classifier to recognize the so called P-waves in the presence of noise by using features extracted from the frequency domain for accurate earliest arrival detection. Preliminary results show that our method has the potential to be a component of a structural health monitoring system based on acoustic emissions for instance for bridges.