Showing papers presented at "IEEE International Symposium on Medical Measurements and Applications in 2013"

KINECT applications for the physical rehabilitation

Abstract: This report presents the results of KINECT applications used in physical rehabilitation tests. Aoyama Gakuin and Kitasato universities collaborated on this project, which is supported by SCOPE. The applications, following standard tests, are for the timed “Up & Go Test”, the timed “10-Meter Walk Test” and for a Joint “Range of Motion” Measurement”; test results are given. The implementation, evaluation and advantages of a proposed “Real-time ROM Measurement” are also given. The proposed KINECT application will be useful for enhancement of KINECT technical capabilities and for further advancements in medical care.

Correcting Smartphone orientation for accelerometer-based analysis

Abstract: A method was developed for rotating a Smartphone accelerometer coordinate system from an offset to a predetermined three-dimensional position to improve accelerometer-based activity identification. A quaternion-based rotation matrix was constructed from an axis-angle pair, produced via algebraic manipulations of the gravity acceleration components in the device's body-fixed frame of reference with the desired position of the vector. The rotation matrix is constructed during quiet standing and then applied to all subsequent accelerometer readings thereafter, transforming their values in this new fixed frame. This method provides a consistent accelerometer orientation between people, thereby reducing Smartphone orientation variability that can adversely affect activity classification algorithms.

Evaluation of the correlation between blood pressure and pulse transit time

Abstract: The arterial blood pressure is an essential physiological parameter for health monitoring. Most blood measurement devices determine the systolic and diastolic arterial blood pressure through the inflation and the deflation of a cuff. This method is uncomfortable to the user and may cause anxiety which in turns can affect the blood pressure (white coat syndrome). This paper investigates a cuff-less non-intrusive approach to measure arterial blood pressure that is suitable for continuous measurement. The approach is based on measuring the delay between the R-peak of the electrocardiogram (ECG) signal and the peaks of the finger photoplethysmograph (PPG) signal. The results of this paper show a high correlation between the blood pressure and the pulse transit time (PTT). Different polynomial regressions are applied for further estimation. The paper uses actual ECG, PPG and blood pressure measurements extracted from the MIMIC database that contains clinical signal data reflecting real measurements. The simulation results verify that the delay (PTT) between the R-peak of the ECG signal and the peaks of the finger PPG signal have a high correlation with arterial blood pressure and can be used as an indicator of the arterial blood pressure.

A Polynomial filtering model for enhancement of mammogram lesions

Abstract: This paper presents a preliminary analysis of a class of non-linear filters for enhancement of mammogram lesions. A non-linear filtering approach employing polynomial model of non-linearity is designed by second order truncation of Volterra series expansion. The proposed filter response is a linear combination of Type-0 and Type-II Volterra filters. Type-0 filter provides contrast enhancement, suppressing the ill-effects of background noise. On the other hand, Type-II filter employs edge enhancement. The objective analysis of the proposed filter is carried out by estimating values of quality parameters like CEM and PSNR on mammograms from MIAS and DDSM databases.

Mutidimensional analysis of EEG features using advanced spectral estimates for diagnosis accuracy

Abstract: Electroencephalogram (EEG) is a source of interesting information if one is able to extract them according to appropriate techniques. The conditions of individual under EEG test is a key issue. In general, EEG feature extraction can be associated to other information like Electrocardiogram (ECG), ergospirometry and electromyogram (EMG). However, in some cases, a multidimensional representation is used; bispectrum is an example of such a representation. HOS (high order statistics), for instance, include the bispectrum and the trispectrum (third and fourth order statistics, respectively). Advanced estimate spectral analysis can reveal new information encompassed in EEG signals. That is the reason the author propose an algorithm based on DSD (Decimated Signal Diagonalization) that is able of processing exponentially dumped signals like those that regard EEG features. The version proposed here is a multidimensional one.

Real-time biosignal quality analysis of ambulatory ECG for detection of myocardial ischemia

Abstract: This work is part of an on-going effort to implement real-time biosignal quality analysis for electrocardiograms (ECG). Biosignal quality analysis can provide an estimate of the signal-to-noise ratio (SNR), which can be used to gate alarms for myocardial ischemia, as indicated by ST deviations in ambulatory ECG. Currently, the false alarm rate is high due to contaminants in the ECG, such as motion artifact. In the proposed algorithm, ECG data are segmented into 16-beat analysis windows. Within an analysis window, the heart's PQRST complexes are segmented, aligned, and averaged to form an estimate of the true PQRST complex output by the heart. A SNR is estimated for each beat by comparing the PQRST complex of each beat to the average PQRST complex; the lowest SNR across all beats in the analysis window is the SNR assigned to the analysis window. Performance of the algorithm is evaluated using ECG data contaminated with various levels of motion artifact. The algorithm provides SNR estimates that are correlated with true SNR (r = 0.89). Leveraging the repeatability of the PQRST complex, it derives these estimates solely from the recorded ECG and does not require a noise free period or any supplementary signals. Results indicate that this algorithm may be a viable method for gating myocardial alarms for reduction of the false alarm rate.

An ECG monitoring system using conductive fabric

Abstract: In this work we evaluated the quality of the ECG signals obtained from conductive fabric dry electrodes. The purpose of this work was to evaluate if lead I ECG collected with dry conductive fabric electrodes is suitable for applications where no special preparation of electrodes specific for ECG monitoring is allowed. An example of such application is ECG-assisted blood pressure (BP) monitoring device where the user of the BP device should only follow standard procedures of BP measurements. In this paper, we present a system that we developed for the acquisition of ECG signals and their transfer to the PC, and we evaluate the quality of the ECG signals from different electrodes placed at biceps and wrist or touched by fingers. In our experiments we compared signals obtained using gel Ag/AgCl, dry contact electrodes made of golden plates and conductive fabric-based electrodes. Conductive fabric-based electrodes are capable of collecting ECG with accuracy comparable to the accuracy of the signal collected by gel electrodes.

Photoplethysmogram-based Blood pressure evaluation using Kalman filtering and Neural Networks

Abstract: The paper deals with the accurate evaluation of the Blood pressure (BP) by an Artificial Neural Network (ANN) and the Photoplethysmogram (PPG) signal. The proposed method allows evaluating the blood pressure for each heart beat, without using a cuff or invasive tool. For each heart beat, a fixed number of features, which characterize the PPG pulse, are extracted and given as the input to the ANN. A systolic, diastolic and mean BP are obtained as the output. The improvement of the BP evaluation accuracy is obtained by removing artifacts from the references used to train the ANN. The filtering of the reference inputs is performed with Kalman based filter in order to take into account the variability of the human pulse rate and cardiovascular system. Preliminary experimental results confirm the suitability of the proposal and asses the BP evaluation accuracy within 5±8 mmHg.

Novel method and real-time system for detecting the Cardiac Defense Response based on the ECG

Abstract: This paper describes a novel integrated system for detecting human emotions using the electrocardiogram (ECG) signal. We describe two different algorithms to detect the heartbeat (QRS complex) in the ECG signal. Their output is used by a third algorithm used to detect a basic emotion response, known as Cardiac Defense Response (CDR), which precedes the emotion of fear. We implemented these algorithms in a real-time system for mobile devices using the SPINE Android programming framework. The proposed system has been validated on 40 subjects during controlled experiments and recognized correctly the activation of the CDR mechanism reaching an overall accuracy of 65%.

Bed occupancy measurements using under mattress pressure sensors for long term monitoring of community-dwelling older adults

Abstract: There is a growing demand for systems that support independent living into advanced age. Technologies that monitor changes in the amount of time older adults spend in bed have the potential to detect critical changes in mobility and support earlier health intervention. Although under mattress sensors have been used previously, processing algorithms were designed for short term monitoring. The objective of this paper was to develop an algorithm and determine optimal sampling rate to obtain bed occupancy characteristics over the longer term. Under mattress sensors were installed in the home of an older adult and data collected over a 3 month period. A processing algorithm was developed to extract bed occupancy information including time in bed, number of bed exits and time of first morning exit. Data were compared using various sampling rates and processing times. Findings indicate that the ideal down sample time for the application was 5 seconds (0.2Hz) and that computational time requirements could be reduced significantly without sacrificing the ability to accurately measure bed occupancy. Features of bed occupancy were plotted and patterns discovered which may be of interest to health clinicians and sleep researchers.

A wireless-based home rehabilitation system for monitoring 3D movements

Abstract: This paper deals with a home rehabilitation system, based on wireless motion sensors, for assisting therapists in assessing recovery of patients with motor dysfunctions. Measurements in home rehabilitation may complement the subjective perception of therapist with objective observations that lead to quantitative evaluation. The goal of this work is to discuss design and implementation of a complete system capable of bringing a real-time 3D reconstruction of human posture to therapist's desk, starting from exercises performed at patient's house. Attention has been paid to reducing costs and clutter of home-side equipments as much as possible.

Biosignal quality analysis of surface EMG using a correlation coefficient test for normality

Abstract: A correlation test of normality is applied to surface electromyography (sEMG) signals to detect and quantify contaminants. Three contaminants were examined: power line interference, motion artifact, and electrocardiogram (ECG) interference. sEMG data from both simulations and human subjects were artificially contaminated at various signal-to-noise ratios (SNR). For each contaminant, lower SNR values were associated with a lower Pearson correlation coefficient; however, the value of the Pearson correlation coefficient did not correspond to the same SNR across contaminant types. The correlation test of normality can be a useful method for detecting contaminants in sEMG, when the type of contaminant is unknown (e.g., for automatic verification sEMG acquisition setups or automatic rejection of contaminated sEMG signals).

How speech processing can help with beat-to-beat heart rate estimation in ballistocardiograms

Abstract: Unobtrusive sensors for monitoring patients' heart rate and other physiological rhythms often produce signals which are challenging to analyze from a signal processing point of view. This is particularly true for bed-mounted ballistocardiography (BCG) sensors. In this paper, we discuss how methods, which are commonly used for pitch tracking in speech processing, can be applied to the problem of beat-to-beat heart rate estimation from BCGs. In particular, we introduce a novel adaptive window autocorrelation method in order to improve the estimation of individual beat-to-beat intervals. Furthermore, we discuss the application of dynamic programming to the extraction of a smooth beat-to-beat interval series from time-varying correlograms. The proposed methods are evaluated with respect to RR-intervals obtained from a reference ECG. Based on the recordings of 5 subjects, the proposed adaptive window method achieved a mean beat-to-beat heart rate interval error of 0.75%.

Relationships of kinematics indexes with amplitude and velocity of upper arm reaching movement

Abstract: Arm reaching movements (ARM) are mainly used in rehabilitative settings, as robot mediated therapies, virtual reality and motion capture systems in exergaming. ARM typical consist of gaussian-like bell shaped velocity profile. No previous paper addressed in details their relationships with amplitude and speed of the movement and aim of the paper is to study existing correlations between quantitative kinematics indexes (KI) and to describe their relationships with the amplitude and the velocity of the movement in normal subjects. We studied about one thousands of horizontal and vertical ARM of 10 normal subjects at four amplitude's values (15-20-25- 30°) and at three different target velocity (20-30- 40°/s), quantitatively evaluated by symmetry, skewness, kurtosis and smoothness indexes. KI showed extremely low correlations to each other, clearly describing different ARM features, higher correlations with the velocity than with the amplitude of the movement, and in vertical antigravitary rather than horizontal ARM. Among all KI, smoothness showed the higher correlations' values, following an exponential growth relationship both with velocity and amplitude in horizontal and vertical ARM.

Dynamic sign language and voice recognition for smart home interactive application

Abstract: This paper presents a system for recognition of dynamic sign language and voice recognition for smart home interactive applications. We use the Bag-of-Features and a local part model approach for bare hand dynamic gesture recognition from video. We use a dense sampling to extract local 3D multiscale whole-part features. We adopted three-dimensional histograms of a gradient orientation (3D HOG) descriptor to represent features. The K-means++ method was applied to cluster the visual words. Dynamic hand gesture classification was conducted by using a Bag-of-features (BOF) and non-linear support vector machine (SVM) methods. As the BOF does not track the order of events we use a multiscale local part model to preserve temporal context. Initial experimental results show a higher level of recognition. A voice recognition system is then used to translate voice commands complementing the hand gesture commands for the human-intuitive control of a personal service android robot for smart home or long-term healthcare environment applications.

Temperature and measurement changes over time for F-Scan sensors

Abstract: Plantar pressure measurement is an important tool for understanding foot and gait biomechanics. F-Scan is a popular device for measuring in-shoe plantar pressures; however, the validity of the F-Scan force measurements has been questioned. Therefore, a study was performed to analyze changes in plantar pressure and temperature over time. One participant was fitted with two F-Scan sensors before step calibration. Single leg standing trials were captured for each limb while the subject stood on a force plate, then the subject performed multiple trials of level ground walking. Sensor temperatures were measured immediately after each set of walking trials. This procedure was repeated every 10 minutes for 140 minutes. Total force values decreased over time, with the largest decrease in total force occurring in the first 60 minutes. Sensor temperature increased during the first hour and then leveled off. Centre of pressure trajectories were similar over 140 minutes, indicating that cell pressures change similarly over time. This study showed that F-Scan is appropriate for evaluating pressure profiles and centre of pressure shape but additional considerations are required when using total force as an outcome measure.

Cognitive change measurement through driving navigation ability sensing and analysis

Abstract: This paper explores the detection of cognitive change in individuals by sensing a high cognition task (driving) The paper proposes algorithms for the analysis of a set of training trips by a driver to create baseline attributes and features for measurement of baseline navigational performance Algorithms are proposed for the measurement of subsequent trips through comparison to the baseline performance attributes and the paper shows that trips with common coping mechanisms for cognitive decline can be identified and classified Common coping mechanisms include use of familiar routes by backtracking to home or reduction in trip complexity through reduction in the variety of stops or in the number of stops are all identified In addition, algorithms are proposed that identify changes in the navigation ability by indicating routing mistakes or poor choices The paper shows that the measurement of patient performance can be compared to gold standard Google Maps based routing and navigation choices providing a baseline for a patient's cognitive performance and that cognitive change could be detected in behavior change relative to this baseline including less efficient trip planning, reduced trip complexity or less optimal navigation through use of inefficient but more familiar routes as coping mechanisms

Validation of a home rehabilitation system for range of motion measurements of limb functions

Abstract: Home rehabilitation systems have already made their way into widely accepted medical practices, and their further diffusion will help limit healthcare-related costs and improve treatment conditions. For a therapy to successfully rely on a rehabilitation system, this should provide support that can be proven to meet treatment-specific purposes. This work deals with the validation of a home rehabilitation system designed to assist therapists in limb motor dysfunction treatment by providing joint angle measurements. The system under test has been compared with two commercial systems, commonly used in rehabilitation laboratories.

Assessment of the enhancement potential of Halloysite Nanotubes for echographic imaging

Abstract: Halloysite Nanotubes (HNTs) are nanomaterials composed of double layered aluminosilicate minerals with a predominantly hollow tubular structure in submicron range. HNTs are characterized by a wide range of applications in anticancer therapy, sustained agent delivery, being particularly interesting because of their tunable release rates and fast adsorption rates. However systematic investigations of their acoustic properties are still poorly documented. This paper shows a quantitative assessment of the effectiveness of HNTs as scatterers at conventional ultrasonic frequencies (5.7 - 7 MHz) in low range of concentrations (1.5-5 mg/mL). Different samples of HNT (diameter: 40-50 nm; length: 0.5 to 2 microns, empty lumen diameter: 15-20 nm) containing agarose gel were imaged through a commercially available echographic system and acquired data were processed through a dedicated prototypal platform in order to extract the average ultrasonic signal amplitude associated to the considered sample. Relationships have been established among backscatter, HNT concentration and the employed echographic frequency. Our results demonstrated that improvement in image backscatter could be achieved incrementing HNT concentration, determining a non-linear signal enhancement due to the fact that they are poly-disperse in length. On the other hand the effect of different echographic frequencies used was almost constant at all concentrations, specifically using higher values of echographic frequency allows yielding a signal enhanced of a factor 1.75±0.26.

New sampling approach for wireless ECG systems with compressed sensing theory

Abstract: Wireless Body Area Networks (WBANs) consist of small intelligent biomedical wireless sensors attached on or implanted to the body to collect vital biomedical data such as electrocardiogram (ECG) signals to provide continuous health monitoring systems for diagnostic and therapeutic purposes. ECG signals are widely used in health care systems because they are noninvasive mechanisms to establish medical diagnosis of heart diseases. In order to fully exploit the benefits of WBANs to Electronic Health (EH), Mobile Health (MH), and Ambulatory Health Monitoring Systems (AHMS) the power consumption and sampling rate should be restricted to a minimum. With this in mind, Compressed Sensing (CS) procedure and the collaboration of Block Sparse Bayesian Learning (BSBL) framework is used to provide new sampling approach for wireless ECG systems with CS theory. Advanced wireless ECG systems based on our approach will be able to deliver healthcare not only to patients in hospital and medical centers; but also in their homes and workplaces thus offering cost saving, and improving the quality of life. Our simulation results illustrate 25% reduction of Percentage Root-mean-square Difference (PRD) and a good level of quality for Signal to Noise Ratio (SNR), sampling-rate, and power consumption.

A crowdsourcing web platform - hip joint segmentation by non-expert contributors

Abstract: In this paper a crowdsourcing web platform for the interactive segmentation of hip joint structures is introduced. The system collects information on how non-expert volunteers segment anatomical components from MR Images, thereby forming a knowledge base on the solution of this type of problems. The information collected permits to determine tuning parameters for automatic and semi-automatic segmentation approaches, and it provides data for training automatic segmentation algorithms. The findings on the human-computer interaction process can be applied in the design of user interfaces for manual and semi-automatic interactive segmentation tools.

Monitoring the relief of pressure points for pressure ulcer prevention: A subject dependent approach

Abstract: Pressure ulcers are of great cost to both the patient and the healthcare system. Devices have been developed with the goal of pressure ulcer prevention, but many available technically complex devices have been shown to be no more effective than low pressure overlays or mattresses. This paper proposes a subject dependent algorithm capable of automatically detecting when and where pressure points have been relieved from underneath a supine subject, without any user inputs or assumptions. Pressure sensitive mats, associated software, a laptop and a video camera were used to measure and collect pressure signals generated by a supine subject performing 3 movements: the subject rolling to one side of the body, then to the other side, and the subject attempting to roll without lifting any pressure points off the mattress. The data was zeroed, baseline values were found, differences in sensor score from baseline were calculated, and instances during which a valley on one side coincided with a peak on the other, were recorded. Examination of these results indicated that the algorithm was capable of determining when and where pressure points underneath the sacrum and foot regions were lifted off the bed, but not capable of determining if a scapula pressure point was relieved. These results suggest that the proposed algorithm is effective for some, but not all regions of the body. Future work will therefore focus on detection of all pressure points, and the adjustment of the algorithm for subject independence.

Automatic detection of high-frequency oscillations in invasive recordings

Abstract: High-frequency oscillations (HFOs) represent relatively new electrographic marker of epileptogenic tissue. It is starting to be used in presurgical examination to better plan surgical resection and to improve outcome of epilepsy surgery. Development of new techniques of unsupervised HFOs detection is required to further investigate the role of HFO in the pathophysiology of epilepsy and to increase the yield of presurgical examination. In this study we applied an envelope distribution modelling technique on experimental and human invasive data to detect HFOs. Application to experimental microelectrode recordings demonstrated satisfactory results with sensitivity 89.9% and false positive rate 2.1 per minute. Application of this algorithm to human invasive recordings achieved sensitivity 80%. High numbers of false positive detections required utilization of postprocessing steps to eliminate the majority of them. This study shows that envelope distribution modelling represents a promising approach to detect HFOs in intracranial recordings. Advantages of this approach are quick adjustments to changes in background activity and resistance to signal nonstationarities. However, successful application to clinical practice requires development of secondary processing steps that will decrease the rate of false positive detections.

Anytime sport activity risk level calculation using HOSVD based hierarchical fuzzy models

Abstract: In this paper a fuzzy logic-based risk calculation model is introduced, which is used to assess the risk level of sport activity in real-time. In these kinds of systems the computational complexity is a key factor, because the sufficiently accurate results should be available in time. The aim is to find the balance between the computational complexity and the accuracy. Anytime techniques are well-suited for these types of problems, because the combination of the soft computing and anytime algorithms can cope with the dynamically changing and possible insufficient amount of resources and reaction time and it is able to adaptively work with the available information which is usually imperfect or even missing. In this study the Singular Value Decomposition (SVD)-based algorithm is used to reduce the basic fuzzy model complexity.

Wearable biopotential measurement using the TI ADS1198 analog front-end and textile electrodes

Abstract: The development of mobile systems for monitoring bioelectric signals outside a hospital environment involves many challenges that do not arise when it is in a controlled environment, like a hospital. The dimensions of these systems are an important factor to consider in order to facilitate their use without interfering with the daily activities of individuals. The purpose of this work is the implementation of a single-supply battery-powered, low power ECG/EMG signal monitoring system based on the ADS1198 Analog Front-End from Texas Instruments. The system was designed to acquire ECG signals from three electrodes using the integrated Right-Leg-Drive (RLD) circuit from the ADS1198. The developed analog front-end was connected for testing purposes through the SPI interface to a NI-USB 8451 board and signals were acquired using LabVIEW. The circuit was tested in several situations and proved to provide high quality signals using textile integrated electrodes and conventional disposable gel electrodes.

Impact of motion artifacts on Heart Rate Variability measurements and classification performance

Abstract: Analysis of Heart Rate Variability (HRV) is an active area of research in the engineering and the medical communities. Current studies use medical-grade ECG signals, from a limited number of available databases. On the other hand, the trend for physiological measurements is towards less obtrusive, wearable devices. It is therefore of interest to apply HRV analysis to signals from such wearable sensors, whose outputs could exhibit varying levels of noise caused by motion artifacts. The main contribution of this paper is the quantification of the impact of imperfect HRV measurements due to motion artifacts on classifiers which use standard time domain and spectral HRV features. The analysis could potentially lead to development of more robust features and classifiers for use in wearable and non-controlled environments.

Individual identification using electrocardiogram morphology

Abstract: The use of electrocardiogram as biometric has raised attention in the last decade and a wide variety of ECG features were explored to verify the feasibility of such a signal. In this work the authors aim to describe a simple template based approach to the electrocardiographic biometric identification using the morphology of individual's heartbeat. The developed algorithm was tested on different recordings made available in the Physionet public database Fantasia: two different sets of heartbeats were extracted from individual recordings one was used for the template building while the second for the tests. The performances of the algorithm are encouraging with a true acceptance rate of 99.4%, however, the procedure needs to be tested on different recordings of the same individual, or during the course of a whole day or physical activity.

Concentration detection in water-glucose mixtures for medical applications using ultrasonic velocity measurements

Abstract: This paper presents the possibility to measure the blood sugar content detecting different ultrasonic velocities in water-glucose mixtures by using two ultrasonic transducers. With the transmission-trough technique and a fixed transducer distance the required wave propagation time can be measured precisely. The ultrasonic actor transducer, excited at the same frequency of f = 2 MHz during different mixtures, uses a burst signal to separate reflections from the direct path. With this technique a very sensitive glucose-water concentration of less than 30 mg/dl can be separated.

The MouthPad: A tongue-computer interface

Abstract: We have developed a tongue-computer interface that displays the tongue position and movement onto a computer screen, and translates them into mouse cursor movements. The system employs an intraoral electrode array, and an embedded controller that provides communication with the target computer. The position and the movement of the tongue are detected by measuring the contact impedance between the tip of the tongue and the electrode array. The system is implemented such that it can easily be expanded to support the translation of tongue gestures into mouse cursor movements, and clicks, for the purpose of controlling a computer, in a similar fashion to a laptop touchpad.

Towards accurate NIBP simulators: Manufacturers' and researchers' contributions

Abstract: Hypertension, the most commonly managed problem in general practice, in its early stages can be diagnosed only by measurement of blood pressure. The prevalent automated method for this purpose is based on the oscillometric principle. The evaluation of the measurement accuracy for oscillometric automated devices involves costly and time consuming population studies, often providing contradictory results. The NIBP simulators, that regenerate oscillometric waveforms, promise an alternative to clinical trials. In the paper a comparison among the most used commercial NIBP simulators is reported, as well as the last research proposals useful to the development of IEEE P1721 project for a new Standard on Objective Measurement of Systemic Arterial Blood Pressure in Humans are analyzed.