Rafik A. Goubran

Noise cancellation using parallel adaptive filters

The development of adaptive noise cancellers based on the reference microphone approach proposed by B. Widrow et al. (1975) is presented. It is shown that the use of multiple reference microphones improves the performance of noise cancelers in a multipath, multisource noise environment commonly found in an automobile. Cancellation based on subbands of the reference signals and the structure of the canceler are shown to improve the ability of the canceler to track nonstationary noise processes. The authors compare various parallel multireference filter structures using noise recorded in an automobile to show that a two-stage canceler using a precorrelation stage followed by a beamforming stage provides noise reduction at all frequencies in the telephone audio band with a maximum reduction of 28 dB and an average reduction of 16 dB. >

Metrics to monitor performance of patients with mild cognitive impairment using computer based games

Mild Cognitive Impairment (MCI) results in a noticeable decrease in cognitive abilities but does not interfere with regular daily functions. Some patients with MCI may be able to return to age-appropriate cognitive abilities, but some will progress to dementia. Clinical interviews, tests and measurements are used to identify and monitor cognitive decline. Timely detection of cognitive decline provides an opportunity for early intervention and can improve care and quality of life. Periodic monitoring of the cognitive status of MCI patients in their own environment is a significant and cost effective aid in patient care. This paper presents a methodology and performance metrics for remote monitoring of cognitive ability using computer based Sudoku and word search games. We collected game data for 2 MCI patients from the Bruyére Memory Program over a period of weeks. Their game performance was then compared to results from cognitive testing using the Montreal Cognitive Assessment (MoCA) and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). The patient who performed better on the games also had higher scores on the cognitive testing.

Driver identification using vehicle acceleration and deceleration events from naturalistic driving of older drivers

Driving is a common task that involves skill and individual preferences that can differ between drivers. The unique driving behaviours can be beneficial for differentiating drivers of shared vehicles and identifying differences between older drivers with normal and declining driving abilities. This paper presents a method for identifying individual drivers based on motor vehicle acceleration and deceleration events from their natural driving behaviour. We provide a novel approach to driver identification based on classification using multiple in-vehicle sensor signals collected in naturalistic conditions with anonymized driving locations. The dataset consists of thousands of trips from a selection of 14 stable-health older drivers (70 years and older) from their first year of the Candrive research study. We trained separate multiclass linear discriminant analysis classifiers to recognize unique patterns in their acceleration and deceleration events to predict the identity of the driver out of a group of drivers. For five different drivers, the acceleration and deceleration events were used to distinguish between drivers at 34% and 30% average accuracy, respectively. By taking a majority vote among the events, the accuracy improved to 61%, exceeding by about three times the null model of random guessing. This performance improvement continues when expanding the group from 2 to 14 drivers. The analysis shows potential for identifying drivers by the patterns in their driving maneuvers such as turning and stopping.

Cognition assessment: A framework for episodic measurement

This paper presents a framework for episodic cognition assessment to augment the current discrete assessment methods used to diagnose Mild Cognitive Impairment (MCI)/dementia. Physicians currently rely on patient history and clinical assessments to identify changes in memory or functional cognitive ability as indications of MCI/dementia in conjunction with exclusionary tests. They are limited in these discrete assessments by their accuracy and infrequent occurrence. The episodic cognition assessment framework proposed will allow better ongoing information about the true well-being of the patient through assessment of memory and functional cognitive ability between appointments. The paper presents three emerging cognitive assessment techniques through the application of sensors, signal processing and data analytics: [i]Electroencephalogram (EEG) / Event Related Potential (ERP) with Neuropsychological Behavioural Testing (NBT) that has the potential to provide a novel bio-marker for MCI/dementia; [ii]Driving and [iii] Computer game use as two Instrumental Activities of Daily Living (IADL). The latter are presented within the episodic assessment framework as both are complex functional cognitive tasks and provide rich insight into the patients ability and changes in ability. The paper shows results of the measures possible from an instrumented word search game and for driving using location tracking sensors. It shows data for an example participant and how the framework would allow these results to be combined.

Detecting Cognitive Ability Changes in Patients With Moderate Dementia Using a Modified “Whack-a-Mole” Game

This paper presents the results from a one-year study of 12 patients with moderate dementia in an adult day program who played a novel whack-a-mole game-based measurement instrument for cognitive behavior and performance. The ongoing measurement of cognition and changes associated with dementia is a challenge for healthcare providers. Measurement methods based on a tablet-based instrument are proposed. Partnership with the adult day program greatly eased recruitment: all but 1 eligible participant joined our study, compared to one in five, or lower, for previous studies with similar populations. There are three unique aspects to the design of our game: first, it has two distinct targets requiring different actions, which increases the cognitive processing for the users; second, each level is systematically more difficult; third, it records and analyzes player performance. The results show that the patients’ game performance improves over the first few weeks; this indicates that they are learning the game and retaining ability gains from week-to-week, suggesting some procedural learning is still intact. Over the year, 4 participants showed cognitive decline, 4 were stable and 3 improved based on their Minimental State Exam (MMSE) score. Two measures are proposed based on level progression within the sessions and mole hit performance. The level progression measure identifies declining participants with one false negative (FN) and one false positive error. The mole hit performance measure identifies declining participants with one FN error. These results demonstrate the potential for the proposed instrument to provide an ongoing measurement as an alternative for the repeated application of the MMSE.

Naturalistic Driving: A Framework and Advances in Using Big Data

Driving is an activity that facilitates physical, cognitive, and social stimulation in older adults, ultimately leading to better physical and cognitive health. However, aging is associated with declines in vision, physical health, and cognitive health, all of which can affect driving ability. One way of assessing driving ability is with the use of sensors in the older adult’s own vehicle. This paper provides a framework for driving assessment and addresses how naturalistic driving studies can assist in such assessments. The framework includes driving characteristics (how much driving, speed, position, type of road), actions and reactions (lane changes, intersections, passing, merging, traffic lights, pedestrians, other vehicles), destinations (variety and distance, sequencing and route planning), and driving conditions (time of day and season). Data from a subset of Ottawa drivers from the Candrive study is used to illustrate the use of naturalistic driving data. Challenges in using naturalistic driving big data and the changing technology in vehicles are discussed.

Monitoring cognitive ability in patients with moderate dementia using a modified “whack-a-mole”

This paper presents results from the first 2 months of a 1-year study of 12 moderate dementia patients that participate in a weekly adult day program within a local community-care access center. The 12 patients are using a tablet-based whack-a-mole game, instrumented to record the users behavior; this record is analyzed to extract indicators, as potential proxies of cognitive ability. Our partnership with the adult day program greatly eased recruitment: all but 1 eligible participant joined our study. The measurements recorded by the game include the detailed user progression through the game levels. There are two unique aspects to the design of our game: first, it includes two distinct targets requiring different actions, which increases the cognitive processing in the tap task for the users; second, each level is systematically more difficult than the last. The results show that the patients performance within the game improves over the first few weeks; this indicates that they are learning the game and retaining ability gains from week to week, which is unexpected in dementia patients. Subsequently they appear to reach a performance plateau, with consistent performance from one week to the next. The performance levels are compared to their MMSE Total score and MMSE Orientation for Time sub-score and they are shown to have a maximum correlation of 0.465 and 0.654 respectively. These results demonstrate the potential for the whack-a-mole game to provide an ongoing measurement alternative for the MMSE and specifically the Orientation for Time sub-score that is a predictor of future decline.

Variation in acceleration driving patterns as a measure of older adult health status

The evaluation of older adult in-car driving data and identification of appropriate driving measures could lead to a better understanding of unsafe driving behaviors and associations with older adult health status. In this paper, the coefficient of variation was examined as a measure for identifying differences related to acceleration driving event variability between 12 physically, cognitively, and general health stable older adult drivers and 16 older adult drivers with declining physical, cognitive, or general health status over a one year period. Declining older adult drivers had significantly lower acceleration coefficient of variation than stable drivers for acceleration event subsets with final velocities between 60 and 100 km/hr, changes in velocities between 15 and 30 km/hr and between 60 and 75 km/hr, and changes in heading between 70 and 180 degrees, which may indicate a reduced adaptation of acceleration driving patterns. Further investigation of acceleration coefficient of variation is warranted to determine whether the decreased acceleration variability in declining drivers is due to physical, cognitive, or general health declines, and to determine the accuracy of this measure at identifying older adults with declining health and those with unsafe driving.

EEG/ERP: Within Episodic Assessment Framework for Cognition

This work explores the potential for electroencephalography (EEG)-based event-related potential (ERP) measurements to be included in the framework for episodic assessment. In current discrete clinical cognitive assessment methods used to diagnose mild cognitive impairment (MCI)/dementia, physicians see patients annually or biannually. They rely on patient’s history and clinical cognitive assessments to identify changes in memory or functional ability as indications of MCI/dementia in conjunction with exclusionary tests. They are limited in these discrete assessments by their accuracy and infrequent occurrence. The episodic assessment framework proposed will allow better ongoing information about the true well-being of the patient through assessment of memory and functional cognitive ability on a weekly or even daily basis between appointments. This paper identifies features of EEG/ERP measurements during Neuropsychological Behavioral Testing that could allow the future inclusion in the framework using now available consumer EEG devices. A pilot group of 32 participants (17 healthy, 15 MCI) was studied using a 1-back test, while their brain activity was measured using EEG. Features of the ERP thus generated were identified and measured. These ERP features and associated behavioral measurements from the 1-back test were analyzed to identify features both alone and in combination that allowed the correct classification of the participants using a repeated leave-one-out train and test method. A feature pair (1-back response time and CPz correlation) was identified that provides the best two-feature performance (1 false positive/1 false negative error) with a third feature (response required – Pz P1 to N1 latency) providing additional improvement. Given these results, it is possible that episodic ERP measurements may help with cognitive impairment diagnosis in the future.

EVOKED RESPONSE POTENTIALS: WHICH FEATURES HELP DISTINGUISH MCI FROM HEALTHY CONTROLS?

and region-specific progression of neuropathology. Additionally, synaptic dysfunction is evident in early development of pathology. However, there is no consensus about the extent to which basal synaptic transmission (BST) and synaptic plasticity are affected in these models.Methods:We used extracellular field potential recordings to study BST, short-term plasticity (PTP, post-tetanic potentiation; PPF, paired-pulse facilitation) and long-term potentiation (LTP) at the Schaffer collateral-CA1 pyramidal cell synapses in young and old rTg4510 (2-3 and 6-7 month old) and old APP/ PS1 mice (8-10 month old). Results:We find that old but not young rTg4510 mice exhibit a correlated reduction in pre-synaptic fiber volley (FV) amplitude (w50%) and field excitatory post-synaptic potential (fEPSP) slope (w40%) compared to WT, consistent with hippocampal neurodegeneration. We also find that BST per se is not altered in this model, since fEPSP slope, controlled for FV amplitude, remained unchanged. In contrast, old APP/PS1 mice did not show reduced FV amplitude compared to WT, while fEPSP slope was reduced by w34%, suggesting a deficit in BST. PTP was reduced in old APP/PS1 mice compared to WT, but not in old rTg4510 mice. PPF was unchanged in old APP/PS1 compared to WT, but was reduced in old rTg4510 mice. LTP, induced with high-frequency stimulation, was reduced in old rTg4510 and APP/PS1 mice. Conclusions: Our data suggest that APP/PS1 mice show reduced BST. In rTg4510 mice, early onset of neurodegeneration may mask BST dysfunction. However, PPF was reduced in rTg4510 mice suggesting some pre-synaptic alteration in remaining neurons. In APP/PS1 mice LTP reduction may be due to induction deficits, since they exhibit both reduced BST and PTP. The basis of the LTP deficit in rTg4510 mice remains unclear.