Neural Networks With Gated Recurrent Units Reduce Glucose Forecasting Error due to Changes in Sensor Location.

BACKGROUND: Continuous glucose monitors (CGMs) have become important tools for providing estimates of glucose to patients with diabetes. Recently, neural networks (NNs) have become a common method for forecasting glucose values using data from CGMs. One method of forecasting glucose values is a time-delay feedforward (FF) NN, but a change in the CGM location on a participant can increase forecast error in a FF NN. METHODS: In response, we examined a NN with gated recurrent units (GRUs) as a method of reducing forecast error due to changes in sensor location. RESULTS: We observed that for 13 participants with type 2 diabetes wearing blinded CGMs on both arms for 12 weeks (FreeStyle Libre Pro-Abbott), GRU NNs did not produce significantly different errors in glucose prediction due to sensor location changes (P < .05). CONCLUSION: We observe that GRU NNs can mitigate error in glucose prediction due to differences in CGM location.

Examining Sensor Agreement in Neural Network Blood Glucose Prediction.

Successful measurements of interstitial glucose are a key component in providing effective care for patients with diabetes. Recently, there has been significant interest in using neural networks to forecast future glucose values from interstitial measurements collected by continuous glucose monitors (CGMs). While prediction accuracy continues to improve, in this work we investigated the effect of physiological sensor location on neural network blood glucose forecasting. We used clinical data from patients with Type 2 Diabetes who wore blinded FreeStyle Libre Pro CGMs (Abbott) on both their right and left arms continuously for 12 weeks. We trained patient-specific prediction algorithms to test the effect of sensor location on neural network forecasting (N = 13, Female = 6, Male = 7). In 10 of our 13 patients, we found at least one significant (P < .05) increase in forecasting error in algorithms which were tested with data taken from a different location than data which was used for training. These reported results were independent from other noticeable physiological differences between subjects (eg, height, age, weight, blood pressure) and independent from overall variance in the data. From these results we observe that CGM location can play a consequential role in neural network glucose prediction.

The association between janitor physical workload, mental workload, and stress: The SWEEP study.

BACKGROUND: Approximately 2.38 million janitors are employed in the U.S. While high physical workload may explain a lost-work days rate 2.7 times greater than other occupations, little is known about the association between janitors' physical workload, mental workload, and stress. OBJECTIVE: The objective of this study was to assess the associations between physical (ergonomic) and mental workload exposures and stress outcomes among janitors. METHODS: Questionnaire data, focused on ergonomic workload, mental workload and stress, were collected from Minnesota janitors for a one-year period. Physical workload was assessed with Borg Scales and Rapid Entire Body Assessments (REBA). Mental workload assessment utilized the NASA Task Load Index (TLX). Stress assessments utilized single-item ordinal stress scale (SISS) and Perceived Stress Scale-4 (PSS-4) measures. Descriptive and multivariable analyses, including bias adjustment, were conducted. RESULTS: Odds ratios (OR) and 95% confidence intervals (CI) for ergonomic workload (task frequency) effects on SISS were: REBA (1.18 OR, 1.02-1.37 CI); Borg (1.25 OR, 1.00-1.56 CI); combined REBA and Borg (1.10 OR, 1.01-1.20 CI). Mental workload was associated with higher PSS-4 levels (0.15 Mean Difference, 0.08-0.22 CI) and a 3% increased risk for each one-unit increase in the SISS scale (1.03 OR, 1.02-1.05 CI). CONCLUSIONS: This research demonstrated a moderate effect of physical and mental workloads on stress among janitors.

Cardiac patient-specific three-dimensional models as surgical planning tools.

BACKGROUND: Three-dimensional printing is an additive manufacturing method that builds objects from digitally generated computational models. Core technologies behind three-dimensional printing are evolving rapidly with major advances in materials, resolution, and speed that enable greater realism and higher accuracy. These improvements have led to novel applications of these processes in the medical field. METHODS: The process of going from a medical image data set (computed tomography, magnetic resonance imaging, ultrasound) to a physical three-dimensional print includes several steps that are described. Medical images originate from Digital Imaging and Communications in Medicine files or data sets, the current standard for storing and transmitting medical images. Via Digital Imaging and Communications in Medicine manipulation software packages, a segmentation process, and manual intervention by an expert user, three-dimensional digital and printed models can be constructed in great detail. RESULTS: Cardiovascular medicine is one of the fastest growing applications for medical three-dimensional printing. The technology is more frequently being used for patient and clinician education, preprocedural planning, and medical device design and prototyping. We report on three case studies, describing how our three-dimensional printing has contributed to the care of cardiac patients at the University of Minnesota. CONCLUSION: Medical applications of computational three-dimensional modeling and printing are already extensive and growing rapidly and are routinely used for visualizing complex anatomies from patient imaging files to plan surgeries and create surgical simulators. Studies are needed to determine whether three-dimensional printed models are cost effective and can consistently improve clinical outcomes before they become part of routine clinical practice.

The Wall Apposition Evaluation for a Mechanical Embolus Retrieval Device.

A computational evaluation approach to the wall apposition of a cerebral mechanical emboli retrieval device (MERD) is presented. The typical enclosed multilattice structure, manufactured from the thin-walled Nitinol tube, consists of repeated "V"-shaped unit cells. During interventional thrombectomy, the MERD system is delivered inside an artery stenosis segment to capture emboli and restore cerebral blood flow. The wall apposition, which deteriorates during embolus capture, occurs during system migration along the tortuous intracranial vessel. The commercial finite element analysis (FEA) solver ABAQUS 6.10 Standard and user subroutine (UMAT/Nitinol) are utilized to study the ability to remain in close contact with the curved vessel wall during migration. In this numerical analysis, the influence of the contacting interference loadings on structure deformation and strain field distribution is obtained and analyzed. The results indicate that the middle segment of the MERD seriously contracts or collapses inside the curved vessel. In addition, the peak strain is in the apex flow-prone region and maintains at the safe range.

Beneficial Effects of Spatial Remapping for Reading With Simulated Central Field Loss.

Purpose: People with central field loss (CFL) lose information in the scotomatous region. Remapping is a method to modify images to present the missing information outside the scotoma. This study tested the hypothesis that remapping improves reading performance for subjects with simulated CFL. Methods: Circular central scotomas, with diameters ranging from 4° to 16°, were simulated in normally sighted subjects using an eye tracker on either a head-mounted display (HMD) (experiments 1, 2) or a traditional monitor (experiment 3). In the three experiments, reading speed was measured for groups of 7, 11, and 13 subjects with and without remapping of text. Results: Remapping increased reading speed in all three experiments. On the traditional monitor, it increased reading speed by 34% (8°), 38% (12°), and 35% (16°). In the two HMD experiments, remapping increased reading speed only for the largest scotoma size, possibly due to latency of updating of the simulated scotoma. Conclusions: Remapping significantly increased reading speed in simulated CFL subjects. Additional testing should examine the efficacy of remapping for reading and other visual tasks for patients with advanced CFL.

Pelvic floor dynamics during high-impact athletic activities: A computational modeling study.

BACKGROUND: Stress urinary incontinence is a significant problem in young female athletes, but the pathophysiology remains unclear because of the limited knowledge of the pelvic floor support function and limited capability of currently available assessment tools. The aim of our study is to develop an advanced computer modeling tool to better understand the dynamics of the internal pelvic floor during highly transient athletic activities. METHODS: Apelvic model was developed based on high-resolution MRI scans of a healthy nulliparous young female. A jump-landing process was simulated using realistic boundary conditions captured from jumping experiments. Hypothesized alterations of the function of pelvic floor muscles were simulated by weakening or strengthening the levator ani muscle stiffness at different levels. Intra-abdominal pressures and corresponding deformations of pelvic floor structures were monitored at different levels of weakness or enhancement. FINDINGS: Results show that pelvic floor deformations generated during a jump-landing process differed greatly from those seen in a Valsalva maneuver which is commonly used for diagnosis in clinic. The urethral mobility was only slightly influenced by the alterations of the levator ani muscle stiffness. Implications for risk factors and treatment strategies were also discussed. INTERPRETATION: Results suggest that clinical diagnosis should make allowances for observed differences in pelvic floor deformations between a Valsalva maneuver and a jump-landing process to ensure accuracy. Urethral hypermobility may be a less contributing factor than the intrinsic sphincteric closure system to the incontinence of young female athletes.

A Head and Neck Support Device for Inducing Local Hypothermia.

The present work describes the design of a device/system intended to induce local mild hypothermia by simultaneously cooling a patient's head and neck. The therapeutic goal is to lower the head and neck temperatures to 33-35 °C, while leaving the core body temperature unchanged. The device works by circulating a cold fluid around the exterior of the head and neck. The head surface area is separated into five different cooling zones. Each zone has a cooling coil and can be independently controlled. The cooling coils are tightly wrapped concentric circles of tubing. This design allows for a dense packing of tubes in a limited space, while preventing crimping of the tubing and minimizing the fluid pressure head loss. The design in the neck region also has multiple tubes wrapping around the circumference of the patient's neck in a helix. Preliminary testing indicates that this approach is capable of achieving the design goal of cooling the brain tissue (at a depth of 2.5 cm from the scalp) to 35 °C within 30- 40 min, without any pharmacologic or circulatory manipulation. In a comparison with examples of current technology, the device has shown the potential for improved cooling capability.

Design by dragging: an interface for creative forward and inverse design with simulation ensembles.

We present an interface for exploring large design spaces as encountered in simulation-based engineering, design of visual effects, and other tasks that require tuning parameters of computationally-intensive simulations and visually evaluating results. The goal is to enable a style of design with simulations that feels as-direct-as-possible so users can concentrate on creative design tasks. The approach integrates forward design via direct manipulation of simulation inputs (e.g., geometric properties, applied forces) in the same visual space with inverse design via 'tugging' and reshaping simulation outputs (e.g., scalar fields from finite element analysis (FEA) or computational fluid dynamics (CFD)). The interface includes algorithms for interpreting the intent of users' drag operations relative to parameterized models, morphing arbitrary scalar fields output from FEA and CFD simulations, and in-place interactive ensemble visualization. The inverse design strategy can be extended to use multi-touch input in combination with an as-rigid-as-possible shape manipulation to support rich visual queries. The potential of this new design approach is confirmed via two applications: medical device engineering of a vacuum-assisted biopsy device and visual effects design using a physically based flame simulation.

Grand challenge: applying regulatory science and big data to improve medical device innovation.

Understanding how proposed medical devices will interface with humans is a major challenge that impacts both the design of innovative new devices and approval and regulation of existing devices. Today, designing and manufacturing medical devices requires extensive and expensive product cycles. Bench tests and other preliminary analyses are used to understand the range of anatomical conditions, and animal and clinical trials are used to understand the impact of design decisions upon actual device success. Unfortunately, some scenarios are impossible to replicate on the bench, and competitive pressures often accelerate initiation of animal trials without sufficient understanding of parameter selections. We believe that these limitations can be overcome through advancements in data-driven and simulation-based medical device design and manufacturing, a research topic that draws upon and combines emerging work in the areas of Regulatory Science and Big Data. We propose a cross-disciplinary grand challenge to develop and holistically apply new thinking and techniques in these areas to medical devices in order to improve and accelerate medical device innovation.

The big picture on nanomedicine: the state of investigational and approved nanomedicine products.

Developments in nanomedicine are expected to provide solutions to many of modern medicine's unsolved problems, so it is no surprise that the literature contains many articles discussing the subject. However, existing reviews tend to focus on specific sectors of nanomedicine or to take a very forward-looking stance and fail to provide a complete perspective on the current landscape. This article provides a more comprehensive and contemporary inventory of nanomedicine products. A keyword search of literature, clinical trial registries, and the Web yielded 247 nanomedicine products that are approved or in various stages of clinical study. Specific information on each was gathered, so the overall field could be described based on various dimensions, including FDA classification, approval status, nanoscale size, treated condition, nanostructure, and others. In addition to documenting the many nanomedicine products already in use in humans, this study identifies several interesting trends forecasting the future of nanomedicine. FROM THE CLINICAL EDITOR: In this one of a kind review, the state of nanomedicine commercialization is discussed, concentrating only on nanomedicine-based developments and products that are either in clinical trials or have already been approved for use.

Recommendations for nanomedicine human subjects research oversight: an evolutionary approach for an emerging field.

The nanomedicine field is fast evolving toward complex, "active," and interactive formulations. Like many emerging technologies, nanomedicine raises questions of how human subjects research (HSR) should be conducted and the adequacy of current oversight, as well as how to integrate concerns over occupational, bystander, and environmental exposures. The history of oversight for HSR investigating emerging technologies is a patchwork quilt without systematic justification of when ordinary oversight for HSR is enough versus when added oversight is warranted. Nanomedicine HSR provides an occasion to think systematically about appropriate oversight, especially early in the evolution of a technology, when hazard and risk information may remain incomplete. This paper presents the consensus recommendations of a multidisciplinary, NIH-funded project group, to ensure a science-based and ethically informed approach to HSR issues in nanomedicine, and to integrate HSR analysis with analysis of occupational, bystander, and environmental concerns. We recommend creating two bodies, an interagency Human Subjects Research in Nanomedicine (HSR/N) Working Group and a Secretary's Advisory Committee on Nanomedicine (SAC/N). HSR/N and SAC/N should perform 3 primary functions: (1) analysis of the attributes and subsets of nanomedicine interventions that raise HSR challenges and current gaps in oversight; (2) providing advice to relevant agencies and institutional bodies on the HSR issues, as well as federal and federal-institutional coordination; and (3) gathering and analyzing information on HSR issues as they emerge in nanomedicine. HSR/N and SAC/N will create a home for HSR analysis and coordination in DHHS (the key agency for relevant HSR oversight), optimize federal and institutional approaches, and allow HSR review to evolve with greater knowledge about nanomedicine interventions and greater clarity about attributes of concern.

Interactive slice WIM: navigating and interrogating volume data sets using a multisurface, multitouch VR interface.

We present Interactive Slice World-in-Miniature (WIM), a framework for navigating and interrogating volumetric data sets using an interface enabled by a virtual reality environment made of two display surfaces: an interactive multitouch table, and a stereoscopic display wall. The framework addresses two current challenges in immersive visualization: 1) providing an appropriate overview+detail style of visualization while navigating through volume data, and 2) supporting interactive querying and data exploration, i.e., interrogating volume data. The approach extends the WIM metaphor, simultaneously displaying a large-scale detailed data visualization and an interactive miniature. Leveraging the table+wall hardware, horizontal slices are projected (like a shadow) down onto the table surface, providing a useful 2D data overview to complement the 3D views as well as a data context for interpreting 2D multitouch gestures made on the table. In addition to enabling effective navigation through complex geometries, extensions to the core Slice WIM technique support interacting with a set of multiple slices that persist on the table even as the user navigates around a scene and annotating and measuring data via points, paths, and volumes specified using interactive slices. Applications of the interface to two volume data sets are presented, and design decisions, limitations, and user feedback are discussed.

The Minnesota pelvic trainer: a hybrid VR/physical pelvis for providing virtual mentorship.

Obtaining accurate understanding of three dimensional structures and their relationships is important in learning human anatomy. To leverage the learning advantages of using both physical and virtual models, we built a hybrid platform consisting of virtual and mannequin pelvis, motion tracking interface, anatomy and pathology knowledge base. The virtual mentorship concept is to allow learners to conveniently manipulate and explore the virtual pelvic structures through the mannequin model and VR interface, and practice on anatomy identification tasks and pathology quizzes more intuitively and interactively than in a traditional self-study classroom, and to reduce the demands of access to dissection lab or wet lab.

The biomechanics of erections: two- versus one-compartment pressurized vessel modeling of the penis.

Previous biomechanical models of the penis simulated penile erections utilizing 2D geometry, simplified 3D geometry or made inaccurate assumptions altogether. These models designed the shaft of the penis as a one-compartment pressurized vessel fixed at one end when in reality it is a two-compartment pressurized vessel in which the compartments diverge as they enter the body and are fixed at two separate anatomic sites. This study utilizes the more anatomically correct two-compartment penile model to investigate erectile function. Simplified 2D and 3D models of the erect penis were developed using the finite element method with varying anatomical considerations for analyzing structural stresses, axial buckling, and lateral deformation. This study then validated the results by building and testing corresponding physical models. Finally, a more complex and anatomically accurate model of the penis was designed and analyzed. When subject to a lateral force of 0.5 N, the peak equivalent von Mises (EVM) stress in the two-compartment model increased by about 31.62%, while in the one-compartment model, the peak EVM stress increased by as high as 70.11%. The peak EVM stress was 149 kPa in the more complex and anatomically accurate penile model. When the perforated septum was removed, the peak EVM stress increased to 455 kPa. This study verified that there is significant difference between modeling the penis as a two- versus a one-compartment pressurized vessel. When subjected to external forces, a significant advantage was exhibited by two corporal based cavernosal bodies separated by a perforated septum as opposed to one corporal body. This is due to better structural integrity of the tunica albuginea when subjected to external forces.

Acute effects of elastic bands during the free-weight barbell back squat exercise on velocity, power, and force production.

The use of elastic bands in resistance training has been reported to be effective in increasing performance-related parameters such as power, rate of force development (RFD), and velocity. The purpose of this study was to assess the following measures during the free-weight back squat exercise with and without elastic bands: peak and mean velocity in the eccentric and concentric phases (PV-E, PV-C, MV-E, MV-C), peak force (PF), peak power in the concentric phase, and RFD immediately before and after the zero-velocity point and in the concentric phase (RFDC). Twenty trained male volunteers (age = 26.0 ± 4.4 years) performed 3 sets of 3 repetitions of squats (at 55% one repetition maximum [1RM]) on 2 separate days: 1 day without bands and the other with bands in a randomized order. The added band force equaled 20% of the subjects' 55% 1RM. Two independent force platforms collected ground reaction force data, and a 9-camera motion capture system was used for displacement measurements. The results showed that PV-E and RFDC were significantly (p < 0.05) greater with the use of bands, whereas PV-C and MV-C were greater without bands. There were no differences in any other variables. These results indicate that there may be benefits to performing squats with elastic bands in terms of RFD. Practitioners concerned with improving RFD may want to consider incorporating this easily implemented training variation.

Ambulatory device for urinary incontinence detection in females.

This paper describes the design and use of an ambulatory monitoring device for recording urological response to intense physical activities of women. The system integrates a tri-axial accelerometer, a 360 degree biaxial inclinometer and a specially designed urine leakage detector(ULD) for sensing body motion and urine discharge. The device is small, lightweight and battery powered, and can be worn comfortably. All recordings are taken non-invasively and transmitted wirelessly to a receiver for real-time data logging. The experimental results show that the proposed system can record acceleration, inclination efficiently and detect urine leakage of amounts from 0.5ml to 10ml accurately. The unique design of the ULD sensor exempts it from sweating interference during vigorous activities.

IBCOM (intra-brain communication) microsystem: wireless transmission of neural signals within the brain.

We report our preliminary work to explore a new method of signal transmission for bio-implantable microsystems. Intra-brain communication or IBCOM is a wireless signal transmission method that uses the brain itself as a conductive medium to transmit the data and commands between neural implants and data processing systems outside the brain. Two miniaturized IBCOM (micro-IBCOM) CMOS chips were designed and fabricated for an in vivo test bed to transmit two prerecorded neural signals at different binary frequency shift keying (BFSK) carrier frequencies to validate the feasibility of IBCOM concept. The chips were packaged for full implantation in a rat brain except for external power delivery. The original neural signal waveforms were successfully recovered after being transmitted between two platinum electrodes separated by 15 mm with transmission power less than 650 pJ/bit for the CMOS implementation.

Feasibility of using a computer modeling approach to study SUI Induced by landing a jump.

Stress urinary incontinence (SUI) occurs due to anatomic and/or neurologic factors involving connective tissues, muscles and nerves. Although SUI is more common in post-menopausal and multiparous women, studies have also shown a high prevalence of SUI in young, physically fit female athletes. With a goal toward dynamic subject-specific mechanical characterization of the interaction between anatomical structures during physical activities that elicit SUI in females during physical or daily activities, a computer aided design (CAD)-based computer model of the female pelvis has been developed to test the feasibility of the computer modeling approach in understanding the measurable differences between stress-continent and stress-incontinent women. In the present study, a fluid-structure interaction analysis was conducted by using the finite element (FE) analysis technique based on the CAD-based computer model of the female pelvis to investigate the urine leakage in females during jumping. To the best of our knowledge, this is the first application of a fluid-structure interaction FE analysis approach in understanding the mechanisms of SUI in females. Through a series of computer simulations, the effects of varying impact forces determined by jumping height and bladder volume were investigated. The dynamic computer simulation results revealed that jumping heights have a significant influence on the volume of urine leakage caused by the landing impact of jumping. Bladder volume did not have a significant influence on leakage when the jumping heights were smaller than 1 ft, which indicates that normal walking (corresponds to a jumping height smaller than 0.1 ft) is not the primary cause of urine leakage for healthy females. The computer simulation results also showed that the deformation difference between the anterior and posterior portion of the female pelvis causes opening of the urethra and resultant urine leakage. The present study demonstrates the feasibility of using a computer modeling approach to study female SUI during physical and daily activities.

Advanced finite element mesh model of female SUI research during physical and daily activities.

With a goal towards dynamic subject-specific mechanical characterization of the etiology of stress urinary incontinence (SUI) in females during physical or daily activities, a finite element (FE) mesh model generation procedure has been developed to build the subject-specific FE model of the female pelvis from multiple high resolution magnetic resonance (MR) acquisitions with varying contrasts. The advanced female pelvis FE model was developed by using this procedure which consisted of over 35 anatomical parts including: 10 pelvic muscles, 10 pelvic ligaments, 6 pelvic bones, skin, fat tissue, bladder, urethra, uterus, vagina, colon, rectum, anus etc. basically all the major parts of the female pelvis. This comprehensive pelvis model is ready to be used to characterize relative relationships and structures during the physical activities that elicit SUI during activities of daily living.