Proof-of-Concept of Microwave-Based Bladder State Detection Using Realistic Pelvic Models.

Goal: Urinary incontinence (UI) affects a significant proportion of the population and is associated with negative physical and psychological side-effects. Microwave-based technologies may have the potential to monitor bladder volume, providing a proactive, low-cost and non-invasive tool to support individuals with UI. Methods: Studies to date on microwave bladder monitoring have been limited to highly simplified computational and experimental scenarios. In this work, we study the most realistic models to date (both male and female), which incorporate dielectrically and anatomically representative tissues of the pelvic region. Results: We examine the ability of detecting bladder fullness through both reflection and transmission-based parameters and, for the first time, study the effect of urine permittivity. As a proof-of-concept of bladder state detection, we further investigate reconstructing differential radar images of the bladder with two different volumes of urine. Conclusions: The results indicate that there is strong potential for monitoring and detecting the bladder state using microwave measurements.

Applied machine learning for stroke differentiation by electrical impedance tomography with realistic numerical models.

Electrical impedance tomography (EIT) may have potential to overcome existing limitations in stroke differentiation, enabling low-cost, rapid, and mobile data collection. Combining bioimpedance measurement technologies such as EIT with machine learning classifiers to support decision-making can avoid commonly faced reconstruction challenges due to the nonlinear and ill-posed nature of EIT imaging. Therefore, in this work, we advance this field through a study integrating realistic head models with clinically relevant test scenarios, and a robust architecture consisting of nested cross-validation and principal component analysis. Specifically, realistic head models are designed which incorporate the highly conductive layers of cerebrospinal fluid in the subarachnoid space and ventricles. In total, 135 unique models are created to represent a large patient population, with normal, haemorrhagic, and ischemic brains. Simulated EIT voltage data generated from these models are used to assess the classification performance of support vector machines. Parameters explored include driving frequency, signal-to-noise ratio, kernel function, and composition of binary classes. Classifier accuracy at 60 dB signal-to-noise ratio, reported as mean and standard deviation, are (79.92% ± 10.82%) for lesion differentiation, (74.78% ± 3.79%) for lesion detection, (77.49% ± 15.90%) for bleed detection, and (60.31% ± 3.98%) for ischemia detection (after ruling out bleed). The results for each method were obtained with statistics from 3 independent runs with 17,280 observations, polynomial kernel functions, and feature reduction of 76% by PCA (from 208 to 50 features). While results of this study show promise for stroke differentiation using EIT data, our findings indicate that the achievable accuracy is highly dependent on the classification scenario and application-specific classifiers may be necessary to achieve acceptable accuracy.

All fiber-based illumination system for multi-exposure speckle imaging.

Monitoring blood flow is critical to treatment efficacy in many surgical settings. Laser speckle contrast imaging (LSCI) is a simple, real-time, label-free optical technique for monitoring blood flow that has emerged as a promising technique but lacks the ability to make repeatable quantitative measurements. Multi-exposure speckle imaging (MESI) is an extension of LSCI that requires increased complexity of instrumentation, which has limited its adoption. In this paper, we design and fabricate a compact, fiber-coupled MESI illumination system (FCMESI) that is substantially smaller and less complex than previous systems. Using microfluidics flow phantoms, we demonstrate that the FCMESI system measures flow with an accuracy and repeatability equivalent to traditional free space MESI illumination systems. With an in vivo stroke model, we also demonstrate the ability of FCMESI to monitor cerebral blood flow changes.

Using pressure-driven flow systems to evaluate laser speckle contrast imaging.

Significance: Microfluidic flow phantom studies are commonly used for characterizing the performance of laser speckle contrast imaging (LSCI) instruments. The selection of the flow control system is critical for the reliable generation of flow during testing. The majority of recent LSCI studies using microfluidics used syringe pumps for flow control. Aim: We quantified the uncertainty in flow generation for a syringe pump and a pressure-regulated flow system. We then assessed the performance of both LSCI and multi-exposure speckle imaging (MESI) using the pressure-regulated flow system across a range of flow speeds. Approach: The syringe pump and pressure-regulated flow systems were evaluated during stepped flow profile experiments in a microfluidic device using an inline flow sensor. The uncertainty associated with each flow system was calculated and used to determine the reliability for instrument testing. The pressure-regulated flow system was then used to characterize the relative performance of LSCI and MESI during stepped flow profile experiments while using the inline flow sensor as reference. Results: The pressure-regulated flow system produced much more stable and reproducible flow outputs compared to the syringe pump. The expanded uncertainty for the syringe pump was 8 to 20 × higher than that of the pressure-regulated flow system across the tested flow speeds. Using the pressure-regulated flow system, MESI outperformed single-exposure LSCI at all flow speeds and closely mirrored the flow sensor measurements, with average errors of 4.6 % ± 2.6 % and 15.7 % ± 4.6 % , respectively. Conclusions: Pressure-regulated flow systems should be used instead of syringe pumps when assessing the performance of flow measurement techniques with microfluidic studies. MESI offers more accurate relative flow measurements than traditional LSCI across a wide range of flow speeds.

Multi-exposure Speckle Imaging for Quantitative Evaluation of Cortical Blood Flow.

Laser speckle contrast imaging (LSCI) is a label-free optical imaging technique that can quantify flow dynamics across an entire image. Multi-exposure speckle imaging (MESI) is an extension of LSCI that allows for reproducible and quantifiable measurements of flow. MESI has the potential to provide quantitative cerebral blood flow information in both preclinical and clinical applications; in fact, MESI can be extended to resolve the flow dynamics in any exposed tissue. A MESI system can be divided into three primary components: (i) the illumination optics, consisting of the optical source and a method of modulating and gating the illumination intensity; (ii) the collection optics, consisting of a high-speed camera that can be triggered and gated to match the pulsed illumination; and finally (iii) post-processing hardware and software to extract the flow information from the recorded raw intensity images. In the following protocol, we offer a guide to design, operate, and test a MESI system.

Dielectric Properties of Ovine Heart at Microwave Frequencies.

Accurate knowledge of the dielectric properties of biological tissues is important in dosimetry studies and for medical diagnostic, monitoring and therapeutic technologies. In particular, the dielectric properties of the heart are used in numerical simulations of radiofrequency and microwave heart ablation. In one recent study, it was demonstrated that the dielectric properties of different components of the heart can vary considerably, contrary to previous literature that treated the heart as a homogeneous organ with measurements that ignored the anatomical location. Therefore, in this study, we record and report the dielectric properties of the heart as a heterogeneous organ. We measured the dielectric properties at different locations inside and outside of the heart over the 500 MHz to 20 GHz frequency range. Different parts of the heart were identified based on the anatomy of the heart and their function; they include the epicardium, endocardium, myocardium, exterior and interior surfaces of atrial appendage, and the luminal surface of the great vessels. The measured dielectric properties for each part of the heart are reported at both a single frequency (2.4 GHz), which is of interest in microwave medical applications, and as parameters of a broadband Debye model. The results show that in terms of dielectric properties, different parts of the heart should not be considered the same, with more than 25% difference in dielectric properties between some parts. The specific Debye models and single frequency dielectric properties from this study can be used to develop more detailed models of the heart to be used in electromagnetic modeling.

The Perceptions of People with Dementia and Key Stakeholders Regarding the Use and Impact of the Social Robot MARIO.

People with dementia often experience loneliness and social isolation. This can result in increased cognitive decline which, in turn, has a negative impact on quality of life. This paper explores the use of the social robot, MARIO, with older people living with dementia as a way of addressing these issues. A descriptive qualitative study was conducted to explore the perceptions and experiences of the use and impact of MARIO. The research took place in the UK, Italy and Ireland. Semi-structured interviews were held in each location with people with dementia (n = 38), relatives/carers (n = 28), formal carers (n = 28) and managers (n = 13). The data was analyzed using qualitative content analysis. The findings revealed that despite challenges in relation to voice recognition and the practicalities of conducting research involving robots in real-life settings, most participants were positive about MARIO. Through the robot's user-led design and personalized applications, MARIO provided a point of interest, social activities, and cognitive engagement increased. However, some formal carers and managers voiced concern that robots might replace care staff.

Dielectric profile of blood clots to inform ischemic stroke treatments.

Platelet and fibrin-rich blood clots can respond differently to red blood cell rich clots during ischemic stroke treatment, which includes thrombolysis and mechanical thrombectomy. Currently, there is no accurate way to identify the type of clot in advance of treatment. If the type of blood clot can be identified, the optimum clot removal process can be chosen and patient outcomes can be improved. In this paper we fabricate physiologically relevant blood clot analogues from human blood, that cover a range of red blood cell, fibrin, and platelet concentrations. We characterize the dielectric profile of these formed clots using an open-ended coaxial probe method across a wide frequency range. After the dielectric measurements are completed, histology on each blood clot is performed to determine the concentration of red blood cells present. In total, 32 unique blood clots were measured.With this completed analysis, we investigate the correlation between the dielectric properties across this frequency range and the red blood cell count of the formed blood clots. Furthermore, we develop a model to predict whether an unknown blood clot can be categorized as red blood cell rich or platelet and fibrin-rich based solely on the measured dielectric properties.Clinical Relevance-Using the dielectric profile of a clot we can predict whether a clot is platelet and fibrin-rich or red blood cell rich allowing clinicians to more easily determine treatment methods during an intervention for ischemic stroke.

Multi-frequency symmetry difference electrical impedance tomography with machine learning for human stroke diagnosis.

OBJECTIVE: Multi-frequency symmetry difference electrical impedance tomography (MFSD-EIT) can robustly detect and identify unilateral perturbations in symmetric scenes. Here, an investigation is performed to assess if the algorithm can be successfully applied to identify the aetiology of stroke with the aid of machine learning. METHODS: Anatomically realistic four-layer finite element method models of the head based on stroke patient images are developed and used to generate EIT data over a 5 Hz-100 Hz frequency range with and without bleed and clot lesions present. Reconstruction generates conductivity maps of each head at each frequency. Application of a quantitative metric assessing changes in symmetry across the sagittal plane of the reconstructed image and over the frequency range allows lesion detection and identification. The algorithm is applied to both simulated and human (n = 34 subjects) data. A classification algorithm is applied to the metric value in order to differentiate between normal, haemorrhage and clot values. MAIN RESULTS: An average accuracy of 85% is achieved when MFSD-EIT with support vector machines (SVM) classification is used to identify and differentiate bleed from clot in human data, with 77% accuracy when differentiating normal from stroke in human data. CONCLUSION: Applying a classification algorithm to metrics derived from MFSD-EIT images is a novel and promising technique for detection and identification of perturbations in static scenes. SIGNIFICANCE: The MFSD-EIT algorithm used with machine learning gives promising results of lesion detection and identification in challenging conditions like stroke. The results imply feasible translation to human patients.

Evaluation of a Companion Robot for Individuals With Dementia: Quantitative Findings of the MARIO Project in an Irish Residential Care Setting.

The current study focuses on the short-term effect of MARIO, a social robot, on quality of life, depression, and perceived social support in persons with dementia (PWD) and evaluates their acceptability of MARIO. Ten PWD in one nursing home took part in a 4-week pilot study, where each participant had up to 12 sessions with MARIO. Sessions comprised engagement in music, news, reminiscence, games, and calendar applications. Standardized questionnaires were administered before and after the 4-week period. Participants had a sustained interest in MARIO during their interactions and an acceptance of MARIO's appearance, sound, and applications. Consequently, participants spent more time socially engaged. No statistically significant differences were found in quality of life, depression, and perceived social support. PWD can engage with a social robot in a real-world nursing home. Future research should incorporate a larger sample and longer intervention period. [Journal of Gerontological Nursing, 45(7), 36-45.].

Sensitivity and Specificity Estimation Using Patient-Specific Microwave Imaging in Diverse Experimental Breast Phantoms.

Many new clinical investigations of microwave breast imaging have been published in recent years. Trials with over one hundred participants have indicated the potential of microwave imaging to detect breast cancer, with particularly encouraging sensitivity results reported from women with dense breasts. The next phase of clinical trials will involve larger and more diverse populations, including women with no breast abnormalities or benign breast diseases. These trials will need to address clinical efficacy in terms of sensitivity and specificity. A number of challenges exist when using microwave imaging with broad populations: 1) addressing the substantial variance in breast composition observed in the population and 2) achieving high specificity given differences between individuals. This paper analyses these challenges using a diverse phantom set which models the variance in breast composition and tumor shape and size seen in the population. The data show that the sensitivity of microwave breast imaging in breasts of differing density can suffer if patient-specific beamforming is not used. Moreover, the results suggest that achieving high specificity in dense breasts may be difficult, but that patient-specific beamforming does not adversely affect the expected specificity. In summary, this paper finds that patient-specific beamforming has a tangible impact on expected sensitivity in experimental cases and that achieving high specificity in dense breasts may be challenging.

Patient-Specific Debye Parameters for Human Blood.

This paper develops a patient-specific model for the Debye parameters of human blood based on hemoglobin content. Blood samples were collected from 176 patients visiting the University Hospital, with both permittivity measurements and standard hematological analysis performed on each blood draw. The complete blood count of each sample provided information on the hemoglobin concentration of each sample; in total there were 73 distinct hemoglobin concentrations reported. An iterative process was used to find patient-specific, based on hemoglobin content, Debye parameters. First, a two-stage genetic algorithm was used to solve for the parameters of a two-pole Debye model based on the mean-blood properties. Then, a modified two-pole Debye model incorporating hemoglobin information was developed, and those parameters were solved for using the same two-stage genetic algorithm. This paper presents the parameters for both the mean-blood model and the patient-specific model. The patient-specific model has a mean-fractional error across all 73 samples of 3.41% compared to 7.64% when using the mean-blood model to represent the entire population. This work demonstrates the range in the dielectric properties of human blood samples and highlights the need for incorporating patient-specific information when using the Debye parameters to model the dielectric properties of human blood.

Image-based classification of bladder state using electrical impedance tomography.

OBJECTIVE: In this study, we examine the potential of using machine learning classification to determine the bladder state ('not full', 'full') with electrical impedance tomography (EIT) images of the pelvic region. Accurate classification of these states would enable urinary incontinence (UI) monitoring to alert the patient, before involuntary voiding occurs, in a low-cost and discrete manner. APPROACH: Using both numerical and experimental data, we form datasets that contain diverse observations with varying clinical parameters such as bladder volume, urine conductivity, and the reference used for time-difference imaging. We then classify the bladder state using both pixel-wise and feature extraction-based classification techniques. We employ principal component analysis, wavelets, and image segmentation to help create features. MAIN RESULTS: The performance was compared across several classifier algorithms. The minimum accuracy was 77.50%. The highest accuracy observed was 100%, and was found by combining principal component analysis and the Gaussian radial based function kernel support vector machine. This combination also offered the best trade-off between classification performance and the costs of training time and memory space. The biggest challenge in bladder state classification is classifying volumes near the separation volume of not full and full, in which choosing the most suitable classifier combination can minimize this error. SIGNIFICANCE: We performed the first machine learning classification of bladder EIT images, achieving high classification accuracies with both numerical and experimental data. This work highlights the potential of using image-based machine learning with an EIT device to support bladder monitoring for those suffering from UI.

Brain haemorrhage detection using a SVM classifier with electrical impedance tomography measurement frames.

Brain haemorrhages often require urgent treatment with a consequent need for quick and accurate diagnosis. Therefore, in this study, we investigate Support Vector Machine (SVM) classifiers for detecting brain haemorrhages using Electrical Impedance Tomography (EIT) measurement frames. A 2-layer model of the head, along with a series of haemorrhages, is designed as both numerical models and physical phantoms. EIT measurement frames, taken from an electrode array placed on the head surface, are used to train and test linear SVM classifiers. Various scenarios are implemented on both platforms to examine the impact of variables such as noise level, lesion location, lesion size, variation in electrode positioning, and variation in anatomy, on the classifier performance. The classifier performed well in numerical models (sensitivity and specificity of 90%+) with signal-to-noise ratios of 60 dB+, was independent of lesion location, and could detect lesions reliably down to the tested minimum volume of 5 ml. Slight variations in electrode layout did not affect performance. Performance was affected by variations in anatomy however, emphasising the need for large training sets covering different anatomies. The phantom models proved more challenging, with maximal sensitivity and specificity of 75% when used with the linear SVM. Finally, the performance of two more complex classifiers is briefly examined and compared to the linear SVM classifier. These results demonstrate that a radial basis function (RBF) SVM classifier and a neural network classifier can improve detection accuracy. Classifiers applied to EIT measurement frames is a novel approach for lesion detection and may offer an effective diagnostic tool clinically. A challenge is to translate the strong results from numerical models into real world phantoms and ultimately human patients, as well as the selection and development of optimal classifiers for this application.

Supervised Learning Classifiers for Electrical Impedance-based Bladder State Detection.

Urinary Incontinence affects over 200 million people worldwide, severely impacting the quality of life of individuals. Bladder state detection technology has the potential to improve the lives of people with urinary incontinence by alerting the user before voiding occurs. To this end, the objective of this study is to investigate the feasibility of using supervised machine learning classifiers to determine the bladder state of 'full' or 'not full' from electrical impedance measurements. Electrical impedance data was obtained from computational models and a realistic experimental pelvic phantom. Multiple datasets with increasing complexity were formed for varying noise levels in simulation. 10-Fold testing was performed on each dataset to classify 'full' and 'not full' bladder states, including phantom measurement data. Support vector machines and k-Nearest-Neighbours classifiers were compared in terms of accuracy, sensitivity, and specificity. The minimum and maximum accuracies across all datasets were 73.16% and 100%, respectively. Factors that contributed most to misclassification were the noise level and bladder volumes near the threshold of 'full' or 'not full'. This paper represents the first study to use machine learning for bladder state detection with electrical impedance measurements. The results show promise for impedance-based bladder state detection to support those living with urinary incontinence.

Interacting with Dementia: The MARIO Approach.

MARIO is a companion robot that aims to help people with dementia (PWD) to battle isolation and loneliness by enabling them to stay socially active by providing a number of applications focused on hobbies (music, movies, etc), staying engaged with communities (reading headlines, reading local twitter feeds etc.) and staying connected with family and friends (telephoning them, reading their news from twitter, etc.). This paper presents the results from the initial trials of MARIO interacting with PWD involving a limited set of applications. It confirms some of the challenges hypothesized at the outset of the study and provides guidelines for future development work.

Investigating the Effect of Social Robot Embodiment.

The experiment described in this paper is an early assessment to identify if the embodiment of a verbal and visual user interaction system in a robot is more effective in people with dementia than when using the same system in a simple laptop. This study provides input for the robot's design.

A Wearable Microwave Antenna Array for Time-Domain Breast Tumor Screening.

In this work, we present a clinical prototype with a wearable patient interface for microwave breast cancer detection. The long-term aim of the prototype is a breast health monitoring application. The system operates using multistatic time-domain pulsed radar, with 16 flexible antennas embedded into a bra. Unlike the previously reported, table-based prototype with a rigid cup-like holder, the wearable one requires no immersion medium and enables simple localization of breast surface. In comparison with the table-based prototype, the wearable one is also significantly more cost-effective and has a smaller footprint. To demonstrate the improved functionality of the wearable prototype, we here report the outcome of daily testing of the new, wearable prototype on a healthy volunteer over a 28-day period. The resulting data (both signals and reconstructed images) is compared to that obtained with our table-based prototype. We show that the use of the wearable prototype has improved the quality of collected volunteer data by every investigated measure. This work demonstrates the proof-of-concept for a wearable breast health monitoring array, which can be further optimized in the future for use with patients with various breast sizes and tissue densities.

Flexible 16 Antenna Array for Microwave Breast Cancer Detection.

Radar-based microwave imaging has been widely studied for breast cancer detection in recent times. Sensing dielectric property differences of tissues has been studied over a wide frequency band for this application. We design single- and dual-polarization antennas for wireless ultrawideband breast cancer detection systems using an inhomogeneous multilayer model of the human breast. Antennas made from flexible materials are more easily adapted to wearable applications. Miniaturized flexible monopole and spiral antennas on a 50-µm Kapton polyimide are designed, using a high-frequency structure simulator, to be in contact with biological breast tissues. The proposed antennas are designed to operate in a frequency range of 2-4 GHz (with reflection coefficient (S11) below -10 dB). Measurements show that the flexible antennas have good impedance matching when in different positions with different curvature around the breast. Our miniaturized flexible antennas are 20 mm × 20 mm. Furthermore, two flexible conformal 4 × 4 ultrawideband antenna arrays (single and dual polarization), in a format similar to that of a bra, were developed for a radar-based breast cancer detection system. By using a reflector for the arrays, the penetration of the propagated electromagnetic waves from the antennas into the breast can be improved by factors of 3.3 and 2.6, respectively.

Breast monitoring via time-domain microwave radar: early clinical trial study.

This work describes early results from our firststage clinical trial involving the monitoring of healthy volunteers with our time-domain microwave breast screening system. The system is composed of a 16-sensor multistatic array that records the electromagnetic energy scattered off of the breast tissue. All measurements are performed in the timedomain. We present here the system setup, patient-interface considerations, volunteer criteria and initial results from breast monitoring.