Search

Blood pressure (BP) is an important clinical vital sign that varies from beat-to-beat. Nevertheless, these variations cannot be captured by the conventional cuff-based BP monitors. This study proposes and evaluates novel cuffless frameworks to continuously estimate the 10-beat averaged systolic BP (SBP) and diastolic BP (DBP) during dynamic exercise by fusing information from multiple biosensors using five machine learning algorithms. Over 100 thousand beats of data were collected from 62 subjects (aged 59 ± 10 years), each underwent a maximal exercise stress test. The average length of recording for each subject was 35 minutes. The BP ranges were 75-280 mmHg for SBP and 36-157 mmHg for DBP respectively. Multiple physiological parameters were measured continuously and used as inputs to five machine learning algorithms for estimating the 10-beat SBP and DBP averages before, during and after the cycling exercise. The mean absolute error (MAE) of Gaussian process regression (GPR) model was 4.8 mmHg and 3.4 mmHg for SBP and DBP, respectively. The MAE of multiple linear regression (MLR), regression tree (RT), ensemble of trees (ETs), and support vector machine (SVM) models varied from 6.1 mmHg to 17.6 mmHg and from 4.0 mmHg to 9.7 mmHg for SBP and DBP, respectively. The GPR model outperformed the other four models and showed promising results in estimating the 10-beat averages of both SBP and DBP without a cuff in a general elderly population under dynamic conditions.

Electromagnetic Navigation Systems (eMNS) can be used to control a variety of multiscale devices within the human body for remote surgery. Accurate modeling of the magnetic fields generated by the electromagnets of an eMNS is crucial for the precise control of these devices. Existing methods assume a linear behavior of these systems, leading to significant modeling errors within nonlinear regions exhibited at higher magnetic fields, preventing these systems from operating at full capacity. In this paper, we use a random forest (RF) and an artificial neural network (ANN) to model the nonlinear behavior of the magnetic fields generated by an eMNS. Both machine learning methods outperformed the state-of-the-art linear multipole electromagnet model (MPEM). The RF and the ANN model reduced the root mean squared error (RMSE) of the MPEM when predicting the field magnitude by approximately 40% and 87%, respectively, over the entire current range of the eMNS. At high current regions, especially between 30 and 35 A, the field-magnitude RMSE improvement of the ANN model over the MPEM was 37 mT, equivalent to 90% error reduction. This study demonstrates the feasibility of using machine learning to model an eMNS for medical applications, and its ability to account for complex nonlinear behavior at high currents. The use of machine learning thus shows promise in developing accurate field predicting models, and ultimately improving surgical procedures that use magnetic navigation. © 2020 IEEE., 2020 IEEE International Conference on Robotics and Automation (ICRA), ISBN:978-1-7281-7395-5, ISBN:978-1-7281-7394-8, ISBN:978-1-7281-7396-2

Algorithms for localising colorectal polyps have been studied extensively; however, they were often trained and tested using the same database. In this study, we present a new application of a unified and real-time object detector based on You-Only-Look-Once (YOLO) convolutional neural network (CNN) for localizing polyps with bounding boxes in endoscopic images. The model was first pre-trained with non-medical images and then fine-tuned with colonoscopic images from three different databases, including an image set we collected from 106 patients using narrow-band (NB) imaging endoscopy. YOLO was tested on 196 white light (WL) images of an independent public database. YOLO achieved a precision of 79.3% and sensitivity of 68.3% with time efficiency of 0.06 sec/frame in the localization task when trained by augmented images from multiple WL databases. In conclusion, YOLO has great potential to be used to assist endoscopists in localising colorectal polyps during endoscopy. CNN features of WL and NB endoscopic images are different and should be considered separately. A large-scale database that covers different scenarios, imaging modalities and scales is lacking but crucial in order to bring this research into reality.

Ankle-Brachial Pressure Index (ABI), the ratio of ankle blood pressure to arm pressure, has been used to assess the severity of peripheral arterial disease (PAD) and to predict major cardiovascular events. This study proposes three indices which can be readily obtained from electrocardiogram and photoplethysmogram, and studies their correlation with ABI in supine and sitting positions. The proposed indices were derived from pulse arrival time (PAT). Specifically, the ratio of PAT measured from the fingertips to PAT measured from the toes of both sides (PAT ralio ), and the relative bilateral differences in PAT measured from the toes of the legs, normalised by average PAT from the two legs (PAT diff1 ) and by the minimal PAT from one side (PAT diff2 ) were calculated. The results showed that PAT ratlo is better correlated with ABI compared to PAT diffl and PAT dff2 . The coefficient of determination (R2) between PAT ratio and ABI is 0.57 when measured in sitting posture, indicating its potential role as an alternative index of ABI for PAD diagnosis. Compared with ABI, the advantages of PAT ratio are 1) it can be measured at ease with wearable sensors that are available at a relatively low cost; 2) the PAT-derived indices can be measured in a sitting posture which is more convenient to obtain; and 3) it can potentially provide longer term, continuous monitoring of the vascular function.

Body sensors are now commonly used as ingestible, wearable and implantable devices for clinical diagnosis and continuous physiological monitoring. Nevertheless, they usually have limited resources. Recent advancements in technologies provide a possible solution to mitigate the resource limitation of these devices by connecting them with mobile devices and cloud services. To evaluate the feasibility of the cloud-enabled body sensor networks, this paper presents simulation results on testing the feasibility of 24-hour operating time and concurrent user support for the cloud-enabled applications.

Wearable sensors have broad applications in many areas. For biomedical applications, they are being actively adopted for human-machine interactive control or health monitoring purposes. This study gives two examples of applying wearable sensors for these biomedical applications by utilizing information from sensor-human-machine interaction and multi-modal sensing. The first example demonstrated the intuitive control of the motions of surgical robots by human arms and hands movements in free space using electromyogram and acceleration measurements. The other one conducted a 20-hour continuous monitoring of multi-modal physiological signals in real life settings using multiple wearable sensors. After automatically removing the motion artifacts from the ambulatory data, median to strong dependence were found between some common cardiovascular parameters (heart rate and pulse amplitude) and the condition of the subject during measurements (psychological status and physical activity level). This study provides the insight that ambulatory monitoring of physiological variables in daily life settings should be interpreted with cautions, and contextual information are needed to standardize the measurement conditions and to get more accurate and reproducible results for better clinical decisions.

This paper presents a task specific robot for the procedure of endoscopic sub mucosal dissection (ESD). This robot has two arms with nine degrees of freedom (DOF), and the capability of tissue elevation and dissection. An optimal design of these robot arms requires the use of shape memory alloy (SMA) wire and steel wire actuators to develop an improved actuation mechanism, which enables force to be transmitted to the distal tip of the robot arms for an efficient tissue dissection. The details of the robot such as robot arm design and kinematics, driven unit mechanism and novel controller design are presented. Two endoscopists, who are also upper gastrointestinal surgeons, independently evaluated our robot in a bench porcine model ex vivo. Both of them were able to complete the procedure within 75-85% of the marked sub mucosal tissue and within 30 minutes.

Flexible and stretchable sensors have recently emerged as an active area of research for designing wearable, implantable and ingestible systems that allow pervasive and unobtrusive health monitoring in mobile health (m-Health) applications. Nevertheless, there lacks systematical study that compares the performances of these new types of sensors with their conventional counterparts. In this study, we reported the fabrication of serpentine-shape, flexible and stretchable electrodes that can be conformally attached to the skin via a printing-transfer process, and we introduced Area Density (AD) as a key parameter to guide the future design of these sensors. Sensors of AD 10, 20 and 40 percentages were designed and fabricated. The performances of these sensors for capturing one type of electrophysiological signals, i.e. electrocardiogram (ECG), were compared against ECG obtained by conventional gold and stainless steel metal clips. The results of this study found that flexible sensors designed with larger AD resulted in ECG of higher signal-to-noise ratio (SNR). In particular, ECG with SNR comparable to those measured by conventional metal clips (SNR>25dB) can be achieved when this new type of flexible sensors were designed with AD=40%. The results of this study shed light on the design of novel wearable and flexible electrodes, which can be used not only for on-body sensing but also in-body measurement inside the gastrointestinal tract.

The flexible endoscope has great potential to reach a diseased organ through a natural opening of the human body to perform surgeries without leaving a scar on the body surface. Despite such advances in reaching the diseased organ, using the conventional flexible endoscope to complete surgical procedure is still largely constrained by the limited degree-of-freedoms (DOFs) of endoscopic instruments. A 3-DOFs robotic gripper, which is designed to work through the channel of a flexible endoscope and driven by tendon-and-sheath system (TSS), has therefore been developed. In order to perform precise trajectory tracking for complex surgical procedures, this paper aims to analyze the friction loss of the flexible robotic gripper inherent to TSS and generate compensation methods for developing an intelligent motion controller for this robotic gripper. The experiment results showed improvement in the trajectory tracking performance after applying the designed compensation techniques.

A wearable cuff-less pulse transit time (PTT) based monitoring device is developed for ambulatory blood pressure (BP) monitoring. Ten healthy subjects (aged 27 ± 4 years old) underwent 24-hour ambulatory BP monitoring using 1) a standard brachial cuff-based oscillometric device as reference and 2) the proposed cuff-less PTT measuring system. Raw PTT and BP measurements were linearly interpolated and then smoothed by a low-pass filter to remove aliasing effect caused by the low sampling rate and synchronized. Resampled PTT and BP were assessed for correlation using correlation coefficients and Bland-Altman plots. Our study showed that PTT estimated systolic BP most accurately within 4.8 ± 4.3 mmHg on healthy young subjects during sleep time. We conclude from this study that the proposed cuff-less PTT-based BP monitoring system has potential to be a less intrusive alternative to standard oscillometric method for long-term overnight BP monitoring.

Baroreflex sensitivity (BRS) has emerged as a prognostic factor in cardiology. Spontaneous sequence technique is one of the most commonly used methods in measuring BRS. The technique requires continuous beat-to-beat blood pressure (BP) which, however, cannot be obtained conveniently. Considering that pulse transit time (PTT) is inversely related to BP and can be easily and continuously acquired from electrocardiogram (ECG) and photoplethysmogram (PPG), we proposed in this study a novel method that estimates BRS from PTT based on the sequence technique with low-pass filtering. The experiment was conducted on 25 healthy subjects from whom continuous BP, PTT and R-R interval (RRI) of ECG were obtained before, during and after drinking 400ml of water. SBP-based BRS and PTT-based BRS were separately calculated. The results showed high correlations between these two BRS estimates in all situations (correlation coefficient γ=0.90, 0.70 and 0.81 before, during and after drinking respectively). Furthermore, results from both methods indicate that BRS increased significantly (p

Nowadays, non-invasive blood pressure (BP) is commonly measured by cuff-based devices, which cause discomfort during measurement and can only provide snapshots of BP. In order to measure beat-to-beat BP continuously and cufflessly, models based on pulse arrival time (PAT) and RR interval (RRI), which can both be conveniently acquired by wearable devices, have been previously proposed. Nevertheless, the existing models often estimate BP based on PAT and RRI of the same beat and ignore the possible time shifts between these variables. In this paper, the cross correlation between RRI, systolic BP (SBP), diastolic BP (DBP) and PAT were studied on 25 young healthy subjects. In total, 1015 cross correlation coefficients, each from a 1-minute signal recording, were calculated for any two cardiovascular variables. The results of the study showed that on average, maximum correlation is found between RRI and DBP when RRI led by 1.3 seconds and between RRI and PAT or SBP when RRI led them by 2.5 seconds. Based on the results of this study, it is therefore concluded that future BP estimation models should take into account the time shifts between these cardiovascular variables when using RRI for the estimation of SBP and DBP as well as when using PAT for the estimation of DBP.

HEALTH informatics refers to the processing, storage, transmission, acquisition, and retrieval (p-STAR) of information from biological systems for the early detection, early prediction, early prevention, early diagnosis and early treatment of diseases [1]–[2]. Since one of the ultimate goals for the research in this area is to detect symptoms or risk factors at an early stage, multi-parameter sensors with higher sensitivity and multi-modal imaging devices with higher resolution are demanded.

P-Health, a future health model that can be described as a 6-P's paradigm, aims to provide low cost and high quality health care via redesigning care practice and networking information systems at different levels. To realise p-Health, a multi-level health information system has to be developed for the processing, storage, transmission, acquisition and retrieval (P-STAR) of health information that spans multiple temporal and spatial scales and consists of multi-modality. This paper uses wearable devices, which have to be miniaturised, integrated, networked, digitalised, smart and standardised (MINDSS), as examples to illustrate how two or more P-STAR technologies are integrated together to implement a specific health care application under p-Health. In particular, standardisation of MINDSS devices that required a new measurement principle, such as the calibration procedure of cuff-less blood pressure measurement devices, is discussed.

Body sensor networks (BSNs) have emerged as a new technology for healthcare applications, but the security of communication in BSNs remains a formidable challenge yet to be resolved. The paper discusses the typical attacks faced by BSNs and proposes a fast biometric based approach to generate keys for ensuing confidentiality and authentication in BSN communications. The approach was tested on 900 segments of electrocardiogram. Each segment was 4 seconds long and used to generate a 128-bit key. The results of the study found that entropy of 96% of the keys were above 0.95 and 99% of the hamming distances calculated from any two keys were above 50 bits. Based on the randomness and distinctiveness of these keys, it is concluded that the fast biometric based approach has great potential to be used to secure communication in BSNs for health applications.

Spectral components of cardiovascular oscillations are related to complex autonomic control mechanisms. Since these oscillations are non-stationary, time-frequency (T-F) analysis has been used in studying cardiovascular oscillations such as variabilities of heart rate and blood pressure; however, few studies have extended the T-F analysis to other cardiovascular signals such as pulse transit time (PTT). In this preliminary study, a T-F analysis method based on a time-variant autoregressive model was used to study the beat to beat spectra of R-R interval (RRI) of electrocardiogram, systolic blood pressure (SBP) and PTT on 8 normotensive subjects after exercise. It is observed from the results that spectra of RRI, SBP and PTT comprised of two distinguished components in the 0.03–0.15 Hz and 0.18–0.4 Hz respectively and that the spectra of SBP and PTT were relatively similar compared to spectra of RRI. However, the beat to beat power spectral densities (PSD) of the signals were not the same all the time. Furthermore, a physiological mechanism-based model was proposed for the generation of ECG, BP and PPG signals, and to interpret the phenomena observed in experiments. Further investigation is needed to quantify this similarity and to study the inherent correlations and mechanisms mediating these cardiovascular oscillations.

Security and privacy are among the most crucial issues for data transmission in tele-healthcare applications. The paper proposes a biometrics based solution, combining encryption and authentication for wireless communication within a body sensor network (BSN), as well as between a BSN and a remote server (RS) of a tele-healthcare system. The method aims to use static and dynamic biometric traits to generate authentication and encryption keys respectively. Keys of 64 and 128 bits were generated from electrocardiogram and photoplethysmogram of 9 subjects and fingerprint images of 20 subjects. The entropy of the keys ranged from 0.662 to 1 and the hamming distances between them were all non-zero. The results of this study found that random and distinctive keys can be generated by a biometric approach for encrypting and authenticating data in tele-healthcare systems.

Nighttime blood pressure (BP) is found to best predict the 5-year risk of cardiovascular death in comparison to daytime BP, BP measured over a 24-hour period and clinical BP. In view of this, a novel contactless system has been developed on a sleeping bed for the cuffless and continuous estimation of BP at night. Experiments were conducted on 11 subjects to evaluate the contactless system, particularly its performance compared to a contact system. The results of this study showed that the accuracy of the contactless system to estimate BP by a cuffless approach is comparable to that of the contact system when measured at the same posture. More studies have to be conducted in order to understand the difference of the cuffless BP estimation approach when measuring at supine and sitting postures.

Arterial stiffness is an important index for cardiovascular events. The objective of this study is to examine possible parameters related to arterial stiffness that can be estimated during simple arm movements. An experiment was conducted on 32 subjects divided into two groups, one with an age of 26+/-4 years old, and the other 61+/-9. The pulse transit time measured from electrocardiogram to finger photoplethysmogram (PPG) and the amplitude of PPG were calculated beat-to-beat for the subjects while they had their arms lowered. The results of the study showed that the ratio between percentage changes in PTT and finger height are significantly different for the two groups of subjects with different age and health conditions, indicating that parameters can be potentially extracted from this procedure to represent the difference in arterial stiffness of the two groups of subjects.

In this paper, a body sensor network (BSN) system for measuring systolic blood pressure (SBP) by a cuffless approach based on h-Shirt has been developed. Two experiments were conducted on a total of 22 subjects to evaluate the cuffless calibration approach and BSN system respectively. The results showed that using the estimated time and distance travelled by a pulse and a parameter derived from the secondary derivative of PPG, the BSN system can measure SBP completely without using a cuff within 1.2±6.0 mmHg of the reference on 10 healthy subjects aged 21-35 yrs.

This paper proposes a new calibration parameter from the dicrotic notch of photoplethysmography (PPG) waveform for systolic blood pressure (SBP) estimation using pulse transit time (PTT). An experiment including exercise was conducted on twelve subjects. The results show that estimation with PTT and a new parameter, i.e. the Relative Amplitude of Secondary peak (RAS), can predict SBP within 1.7 plusmn 6.8 mmHg of the reference for all the subjects before and after exercise (including within 5 minutes and 20-40 minutes after exercise). The underestimated bias after exercise if PTT was used alone can be largely reduced when RAS is introduced. The preliminary analysis indicates that RAS is a promising parameter for calibrating the PTT-based approach for cuffless BP estimation.

The evaluation of blood pressure (BP) measuring devices typically involves more than one parameter. Because the comparison of devices using two or more parameters is complex, a single parameter evaluation system is desirable. Four widely used difference measures are chosen for the study: mean absolute difference (MAD), root mean square difference (RMSD), mean absolute percentage difference (MAPD), and root mean square percentage difference (RMSPD). Theoretical and experimental results indicated that MAD and MAPD are potentially better than the other two parameters to establish a single parameter evaluation system.

Continuous and long-term measurement of arterial blood pressure (BP) and its variability is an important but unsolved topic in physiological monitoring. A possible solution is to use an approach that estimate BP based on pulse transit time. Since the method requires information collected from sensors placed at various body locations, a network could be setup to connect the different sensors effectively and securely. In this paper, we propose using multiple modes of communication, including communication via wired, wireless and/or biological channels (bio-channels) of the human body, to network the various sensors, i.e. the idea of hybrid body sensor network (h-BSN). Preliminary results show that h-BSN is potentially useful for monitoring BP over extensive periods of time and the idea can be extended to other applications that employ multiple on-body or in-body sensors.

Pulse transit time (PTT) is important in the study of arterial viscoelastic properties, wave reflections and sympathetic activities. PTT is usually defined as the time interval between the R-wave in the electrocardiogram (ECG) and a characteristic point in the photoplethysmogram. However, it was found that photoplethysmographic signals are affected by the contacting force between the photoplethysmographic sensor and the measurement site. Therefore, the aim of this study is to examine the effect of contacting force on PTT. Four characteristic points in photoplethysmogram were selected to define PTT: (1) the point at 90% of the pulse amplitude (PTT1), (2) the point at 10% of the pulse amplitude (PTT2), (3) the peak of the first derivative of photoplethysmogram (PTT3), and (4) the peak of the second derivative of photoplethysmogram (PTT4). Fifteen healthy subjects participated in the experiment and ECG and photoplethysmographic signals were recorded under different contacting forces ranging from 0.1 N to 0.5 N. It was found that the PTT defined by the point near the foot (PTT2 and PTT4) or on the rising limb (PTT3) of photoplethysmogram increased with the contacting force till near the zero transmural pressure, after that PTT kept at an almost constant level. Whereas such changing trend was not found in PTT defined by the point near the peak of photoplethysmogram

The pulse transit time (PTT)-based approach has been proposed as a noninvasive and cuffless alternative to the conventional occluding-cuff approaches to be used in wearable medical devices for long-term and continuous monitoring of blood pressure (BP). Nevertheless, the PTT-based technique requires an individualized calibration procedure. In order to derive a simple procedure for estimating individualized coefficients for this approach, this paper aims to model PTT under the effects of hydrostatic pressure due to hand elevation. In this study, 11 volunteers were recruited. BP and PTT were measured while the subjects elevated their hands to specific heights above heart level (h). The results of the study show that PTT changes significantly with h (p

The two widely used standards for validating the conventional cuff-based BP devices are those set up by the American Association for the Advancement of Medical Instrumentation (AAMI) and the British Hypertension Society (BHS) respectively. It was reported that although both standards work well for most BP devices, they do not agree with each other in all circumstances. Previous attempts to model the measurement errors as Gaussian could not explain the phenomena. Results of this study indicate that the general t distribution fits better to the BP measurement errors reported in previous literatures and collected from our clinical experiments. Furthermore, the theoretical mapping of the AAMI accuracy criteria and BHS grading system derived from the t distribution model can better reflect the real situations, where some devices can pass the BHS criteria but fail that of the AAMI. Based on this mapping relationship, it is found that current standards work well for devices with error distribution close to Gaussian but they may not be suitable for the newly developed wearable BP measuring devices used in the experiment, where the errors followed a t distribution with a smaller degree of freedom. It is suggested that a statistical measure that takes into account both the probability and the absolute value of the extreme errors could be introduced into the standard in order to better evaluate this kind of wearable, cuffless BP measuring devices.

It is suggested using wearable medical devices and sensors as the nodes of body sensor networks (BSN) could allow better long-term monitoring of health condition. Nonetheless, the protocol and criteria for validating these nodes in clinical settings is often overlooked. By using the validation of blood pressure (BP) measurement devices as an example, this paper reveals the need for new standards for nodes of BSN and proposes essential considerations to it. It is observed from a previous clinical survey and our theoretical analysis that there are disagreements in existing standards on the validation of conventional BP devices. Moreover, an experiment carried out by our group on 85 subjects demonstrates the inappropriateness of using the existing protocols that are setup for validating cuff-based BP devices to evaluate new cuff-less measurement techniques. The results suggested that a third objective measure could be introduced to relate the variant standards. It is also proposed that different validation criteria should be used for nodes of BSN that are developed based on new measurement principles.

This study examines the recoverability trend of blood pressure (BP) and pulse transit time (PTT) during the 60 minutes immediately following a treadmill exercise. Experiments were conducted on 12 healthy and normotensive subjects, including 8 males and 4 females. All subjects were directed to run on a treadmill at 10 kph for 4 minutes. BP was measured at the finger by the Finometer and also at brachial arteries by a sphygmomanometer. PTT was determined as the time interval between the peak of the R wave of electrocardiogram and the foot of photoplethysmogram. Immediate post exercise PTT is significantly lower than its pre-exercise baseline level (p < 0.001). During the recovery period, PTT initially increases to the original level at 9 minutes after exercise and then increases continuously, peaks at 20 minutes after exercise, and returns to the baseline at 60 minutes. Although PTT is temporarily higher than the baseline during the recovery period, the changes are not significant (p = 0.37). Immediately after exercise, both finger BP and brachial BP are significantly different from pre-exercise baseline. There is no significant difference in brachial BP by 5 minutes after exercise and in finger BP, by 9 minutes

Software quality assurance (SQA) is an essential process to ensure high quality products. With the maturing of the agent technology, SQA can take advantage of this new development. This paper presents a framework of a multi-agent environment (AUTOQ) that can aid software quality assurance. SQA activities that can be performed by agents have been identified. A scenario of how multi-site software development can use AUTOQ is presented. It is expected that the use of AUTOQ will improve the monitoring and controlling of SQA activities and eliminates most manual effort.