Innovative System for Evaluation and Rehabilitation of Human Imbalance.

<b>Introduction:</b> Mobile posturography is based on wearable inertial sensors; it allows to test static stability (static posturography) and gait disturbances. </br> </br> <b>Aim:</b> The aim of this work was to present the results of research on the innovative MEDIPOST system used for diagnosis and rehabilitation of balance disorders. </br> </br> <b>Material and methods:</b> Fourteen articles published in influenced foreign journals were presented and discussed. The deve-lopment and construction of the device was preceded by a literature review and methodological work. The Dizziness Handi-cap Inventory (DHI) questionnaire was translated and validated. The methodology of posturography with head movements with a frequency of 0.3 Hz was also developed in the group with chronic vestibular disorders. Simultaneous measurements were performed (static posturogrphy vs. MEDIPOST) in the CTSIB-M (Modified Clinical Test of Sensory Interaction in Balance) test in healthy subjects and patients with unilateral peripheral dysfunction.</br> </br> <b>Results:</b> In the posturography with head movements the improvement of sensitivity (67 to 74%) and specificity (65 to 71%) was noted. In the CTSIB-M test the intraclass correlation coefficients for both methods were 0.9. The greatest differences between examinations were observed for the mean angular velocity in the tests on the foam (trials no. 3 and 4), in particular on the foam with eyes closed (trial no. 4 - sensitivity 86.4%, specificity 87.7%). Two functional tests were analyzed: the Swap Seats test and the 360 degree turn test. In the former, the results are studied from 6 sensors - 86% of the true positives and 73% of the true negatives for the fall/ no-fall group classification. The second test differentiates people with vestibular impairment and healthy people. It can be analyzed with 1 (sensitivity 80%) and 6 sensors (sensitivity 86%, specificity 84%). Currently, the MEDIPOST device is in the development and certification phase.

Detection of balance disorders using rotations around vertical axis and an artificial neural network.

Vestibular impairments affect patients' movements and can result in difficulties with daily life activities. The main aim of this study is to answer the question whether a simple and short test such as rotation about a vertical axis can be an objective method of assessing balance dysfunction in patients with unilateral vestibular impairments. A 360˚ rotation test was performed using six MediPost devices. The analysis was performed in three ways: (1) the analytical approach based only on data from one sensor; (2) the analytical approach based on data from six sensors; (3) the artificial neural network (ANN) approach based on data from six sensors. For approaches 1 and 2 best results were obtained using maximum angular velocities (MAV) of rotation and rotation duration (RD), while approach 3 used 11 different features. The following sensitivities and specificities were achieved: for approach 1: MAV-80% and 60%, RD-69% and 74%; for approach 2: 61% and 85% and RD-74% and 56%; for approach 3: 88% and 84%. The ANN-based six-sensor approach revealed the best sensitivity and specificity among parameters studied, however one-sensor approach might be a simple screening test used e.g. for rehabilitation purposes.

Fully Automatic Fall Risk Assessment Based on a Fast Mobility Test.

This paper presents a fall risk assessment approach based on a fast mobility test, automatically evaluated using a low-cost, scalable system for the recording and analysis of body movement. This mobility test has never before been investigated as a sole source of data for fall risk assessment. It can be performed in a very limited space and needs only minimal additional equipment, yet provides large amounts of information, as the presented system can obtain much more data than traditional observation by capturing minute details regarding body movement. The readings are provided wirelessly by one to seven low-cost micro-electro-mechanical inertial measurement units attached to the subject's body segments. Combined with a body model, these allow segment rotations and translations to be computed and for body movements to be recreated in software. The subject can then be automatically classified by an artificial neural network based on selected values in the test, and those with an elevated risk of falls can be identified. Results obtained from a group of 40 subjects of various ages, both healthy volunteers and patients with vestibular system impairment, are presented to demonstrate the combined capabilities of the test and system. Labelling of subjects as fallers and non-fallers was performed using an objective and precise sensory organization test; it is an important novelty as this approach to subject labelling has never before been used in the design and evaluation of fall risk assessment systems. The findings show a true-positive ratio of 85% and true-negative ratio of 63% for classifying subjects as fallers or non-fallers using the introduced fast mobility test, which are noticeably better than those obtained for the long-established Timed Up and Go test.

Real-time prediction of acute cardiovascular events using hardware-implemented Bayesian networks.

This paper presents a decision support system that aims to estimate a patient׳s general condition and detect situations which pose an immediate danger to the patient׳s health or life. The use of this system might be especially important in places such as accident and emergency departments or admission wards, where a small medical team has to take care of many patients in various general conditions. Particular stress is laid on cardiovascular and pulmonary conditions, including those leading to sudden cardiac arrest. The proposed system is a stand-alone microprocessor-based device that works in conjunction with a standard vital signs monitor, which provides input signals such as temperature, blood pressure, pulseoxymetry, ECG, and ICG. The signals are preprocessed and analysed by a set of artificial intelligence algorithms, the core of which is based on Bayesian networks. The paper focuses on the construction and evaluation of the Bayesian network, both its structure and numerical specification.