Understanding the influence of context on real-world walking energetics.

Speeds that minimize energetic cost during steady-state walking have been observed during lab-based investigations of walking biomechanics and energetics. However, in real-world scenarios, humans walk in a variety of contexts that can elicit different walking strategies, and may not always prioritize minimizing energetic cost. To investigate whether individuals tend to select energetically optimal speeds in real-world situations and how contextual factors influence gait, we conducted a study combining data from lab and real-world experiments. Walking kinematics and context were measured during daily life over a week (N=17) using wearable sensors and a mobile phone. To determine context, we utilized self-reported activity logs, GPS data and follow-up exit interviews. Additionally, we estimated energetic cost using respirometry over a range of gait speeds in the lab. Gross and net cost of transport were calculated for each participant, and were used to identify energetically optimal walking speed ranges for each participant. The proportion of real-world steady-state stride speeds within these ranges (gross and net) were identified for all data and for each context. We found that energetically optimal speeds predicted by gross cost of transport were more predictive of walking speeds used during daily life than speeds that would minimize net cost of transport. On average, 82.2% of all steady-state stride speeds were energetically optimal for gross cost of transport for all contexts and participants, while only 45.6% were energetically optimal for net cost of transport. These results suggest that while energetic cost is a factor considered by humans when selecting gait speed in daily life, it is not the sole determining factor. Context contributes to the observed variability in movement parameters both within and between individuals.

Sensor Location Matters When Estimating Player Workload for Baseball Pitching.

Estimating external workload in baseball pitchers is important for training and rehabilitation. Since current methods of estimating workload through pitch counts and rest days have only been marginally successful, clubs are looking for more sophisticated methods to quantify the mechanical loads experienced by pitchers. Among these are the use of wearable systems. While wearables offer a promising solution, there remains a lack of standards or guidelines for how best to employ these devices. As a result, sensor location and workload calculation methods vary from system to system. This can influence workload estimates and blur their interpretation and utility when making decisions about training or returning to sport. The primary purpose of this study was to determine the extent to which sensor location influences workload estimate. A secondary purpose was to compare estimates using different workload calculations. Acceleration data from three sensor locations-trunk, throwing upper arm, and throwing forearm-were collected from ten collegiate pitchers as they threw a series of pitches during a single bullpen session. The effect of sensor location and pitch type was assessed in relation to four different workload estimates. Sensor location significantly influenced workload estimates. Workload estimates calculated from the forearm sensor were significantly different across pitch types. Whole-body workload measured from a trunk-mounted sensor may not adequately reflect the mechanical loads experienced at throwing arm segments. A sensor on the forearm was the most sensitive to differences in workloads across pitch types, regardless of the calculation method.

Using Sensors for Player Development: Assessing Biomechanical Factors Related to Pitch Command and Velocity.

Pitching biomechanical research is highly focused on injury prevention with little attention to how biomechanical data can facilitate skill development. The overall purpose of this study was to explore how sensor-derived segment kinematics and timing relate to command and ball velocity during baseball pitching. We used a cross-sectional design to analyze a series of pitches thrown from 10 collegiate baseball pitchers. We collected biomechanical data from six inertial sensors, subjective command from the pitchers, and ball velocity from a radar device. Stepwise regression analyses were used to explore biomechanical variables associated with command for all pitches and ball velocity for fastballs only. We found that only peak forearm linear acceleration was significantly associated with command, whereas several segment kinematic measures were significantly associated with ball velocity. Our results suggest that different biomechanical variables are linked to specific pithing skills. Our findings suggest that end-effector (forearm) movement is more important for pitch command, whereas proximal-to-distal (pelvis, trunk, upper arm, forearm) segmental movement is important for ball velocity.

Evaluation of Error-State Kalman Filter Method for Estimating Human Lower-Limb Kinematics during Various Walking Gaits.

Inertial measurement units (IMUs) offer an attractive way to study human lower-limb kinematics without traditional laboratory constraints. We present an error-state Kalman filter method to estimate 3D joint angles, joint angle ranges of motion, stride length, and step width using data from an array of seven body-worn IMUs. Importantly, this paper contributes a novel joint axis measurement correction that reduces joint angle drift errors without assumptions of strict hinge-like joint behaviors of the hip and knee. We evaluate the method compared to two optical motion capture methods on twenty human subjects performing six different types of walking gait consisting of forward walking (at three speeds), backward walking, and lateral walking (left and right). For all gaits, RMS differences in joint angle estimates generally remain below 5 degrees for all three ankle joint angles and for flexion/extension and abduction/adduction of the hips and knees when compared to estimates from reflective markers on the IMUs. Additionally, mean RMS differences in estimated stride length and step width remain below 0.13 m for all gait types, except stride length during slow walking. This study confirms the method's potential for non-laboratory based gait analysis, motivating further evaluation with IMU-only measurements and pathological gaits.

Quantifying Throw Counts and Intensities Throughout a Season in Youth Baseball Players: A Pilot Study.

Overuse injuries in youth baseball players due to throwing are at an all-time high. Traditional methods of tracking player throwing load only count in-game pitches and therefore leave many throws unaccounted for. Miniature wearable inertial sensors can be used to capture motion data outside of the lab in a field setting. The objective of this study was to develop a protocol and algorithms to detect throws and classify throw intensity in youth baseball athletes using a single, upper arm-mounted inertial sensor. Eleven participants from a youth baseball team were recruited to participate in the study. Each participant was given an inertial measurement unit (IMU) and was instructed to wear the sensor during any baseball activity for the duration of a summer season of baseball. A throw identification algorithm was developed using data from a controlled data collection trial. In this report, we present the throw identification algorithm used to identify over 17,000 throws during the 2-month duration of the study. Data from a second controlled experiment were used to build a support vector machine model to classify throw intensity. Using this classification algorithm, throws from all participants were classified as being "low," "medium," or "high" intensity. The results demonstrate that there is value in using sensors to count every throw an athlete makes when assessing throwing load, not just in-game pitches.

Workload Differentiates Breast Surgical Procedures: NSM Associated with Higher Workload Demand than SSM.

BACKGROUND: Breast surgery has evolved with more focus on improving cosmetic outcomes, which requires increased operative time and technical complexity. Implications of these technical advances in surgery for the surgeon are unclear, but they may increase intraoperative demands, both mentally and physically. We prospectively evaluated mental and physical demand across breast surgery procedures, and compared surgeon ergonomic risk between nipple-sparing (NSM) and skin-sparing mastectomy (SSM) using subjective and objective measures. METHODS: From May 2017 to July 2017, breast surgeons completed modified NASA-Task Load Index (TLX) workload surveys after cases. From January 2018 to July 2018, surgeons completed workload surveys and wore inertial measurement units to evaluate their postures during NSM and SSM cases. Mean angles of surgical postures, ergonomic risk, survey items, and patient factors were analyzed. RESULTS: Procedural duration was moderately related to surgeon frustration, mental and physical demand, and fatigue (p < 0.001). NSMs were rated 23% more physically demanding (M = 13.3, SD = 4.3) and demanded 28% more effort (M = 14.4, SD = 4.6) than SSMs (M = 10.8, SD = 4.7; M = 11.8, SD = 5.0). Incision type was a contributing factor in workload and procedural difficulty. Left arm mean angle was significantly greater for NSM (M = 30.1 degrees, SD = 6.6) than SSMs (M = 18.2 degrees, SD = 4.3). A higher musculoskeletal disorder risk score for the trunk was significantly associated with higher surgeon physical workload (p = 0.02). CONCLUSION: Nipple-sparing mastectomy required the highest surgeon-reported workload of all breast procedures, including physical demand and effort. Objective measures identified the surgeons' left upper arm as being at the greatest risk for a work-related musculoskeletal disorder, specifically from performing NSMs.

Non-linear statistical photocalibration of photodetectors without calibrated light sources.

Calibration of CCD arrays is commonly conducted using dark frames. Non-absolute calibration techniques only measure the relative response of the detectors. For absolute calibration to be achieved, a second calibration is sometimes utilized by looking at sources with known radiances. A process like this can be used to calibrate photodetectors if a calibration source is available and sensor time can be spared to perform the operation. A previous attempt at creating a procedure for calibrating a photodetector using the underlying Poisson nature of the photodetection statistics relied on a linear model. This effort produced the statistically applied non-uniformity calibration algorithm, which demonstrated an ability to relate the measured signal with the true radiance of the source. Reliance on a completely linear model does not allow for non-linear behaviors to be described, thus potentially producing poor photocalibration over large dynamic ranges. In this paper, a photocalibration procedure is defined that requires only first and second moments of the measurements and allows the response to be modeled using a non-linear function over the dynamic range of the detector. The technique is applied to image data containing a light source measured with different integration times showing that the non-linear technique achieves significant improvement over the linear model over a large dynamic range.

Medication Adherence and Liquid Level Tracking System for Healthcare Provider Feedback.

A common problem for healthcare providers is accurately tracking patients' adherence to medication and providing real-time feedback on the management of their medication regimen. This is a particular problem for eye drop medications, as the current commercially available monitors focus on measuring adherence to pills, and not to eye drops. This work presents an intelligent bottle sleeve that slides onto a prescription eye drop medication bottle. The intelligent sleeve is capable of detecting eye drop use, measuring fluid level, and sending use information to a healthcare team to facilitate intervention. The electronics embedded into the sleeve measure fluid level, dropper orientation, the state of the dropper top (on/off), and rates of angular motion during an application. The sleeve was tested with ten patients (age ≥65) and successfully identified and timestamped 94% of use events. On-board processing enabled event detection and the measurement of fluid levels at a 0.4 mL resolution. These data were communicated to the healthcare team using Bluetooth and Wi-Fi in real-time, enabling rapid feedback to the subject. The healthcare team can therefore monitor a log of medication use behavior to make informed decisions on treatment or support for the patient.

Near earth space object detection using parallax as multi-hypothesis test criterion.

The US Strategic Command (USSTRATCOM) operated Space Surveillance Network (SSN) is tasked with Space Situational Awareness (SSA) for the U.S. military. This system is made up of Electro-Optic sensors, such as the Ground-based Electro-Optical Deep Space Surveillance (GEODSS) and RADAR based sensors, such as the Space Fence Gaps. They remain in the tracking of Resident Space Objects (RSO's) in Geosynchronous Orbits (GEO), due to limitations of SST and GEODSS system implementation. This research explores a reliable, ground-based technique used to quickly determine an RSO's altitude from a single or limited set of observations. Implementation of such sensors into the SSN would mitigate GEO SSA performance gaps. The research entails a method used to distinguish between the point spread function (PSF) observed by a star and the PSF observed from an RSO by using Multi-Hypothesis Testing with parallax as a test criterion. Parallax is the effect that an observed object will appear to shift when viewed from different positions. This effect is explored by generating PSFs from telescope observations of space objects at different baselines. The research has shown the PSF of an RSO can be distinguished from that of a star using single, simultaneous observations from reference and parallax sensing telescopes. This report validates these techniques with both simulations and experimental data from the SST and Naval Observatory sensors.

Effect of IMU Design on IMU-Derived Stride Metrics for Running.

Researchers employ foot-mounted inertial measurement units (IMUs) to estimate the three-dimensional trajectory of the feet as well as a rich array of gait parameters. However, the accuracy of those estimates depends critically on the limitations of the accelerometers and angular velocity gyros embedded in the IMU design. In this study, we reveal the effects of accelerometer range, gyro range, and sampling frequency on gait parameters (e.g., distance traveled, stride length, and stride angle) estimated using the zero-velocity update (ZUPT) method. The novelty and contribution of this work are that it: (1) quantifies these effects at mean speeds commensurate with competitive distance running (up to 6.4 m/s); (2) identifies the root causes of inaccurate foot trajectory estimates obtained from the ZUPT method; and (3) offers important engineering recommendations for selecting accurate IMUs for studying human running. The results demonstrate that the accuracy of the estimated gait parameters generally degrades with increased mean running speed and with decreased accelerometer range, gyro range, and sampling frequency. In particular, the saturation of the accelerometer and/or gyro induced during running for some IMU designs may render those designs highly inaccurate for estimating gait parameters.

Improved space object detection using short-exposure image data with daylight background.

Space object detection is of great importance in the highly dependent yet competitive and congested space domain. The detection algorithms employed play a crucial role in fulfilling the detection component in the space situational awareness mission to detect, track, characterize, and catalog unknown space objects. Many current space detection algorithms use a matched filter or a spatial correlator on long-exposure data to make a detection decision at a single pixel point of a spatial image based on the assumption that the data follow a Gaussian distribution. Long-exposure imaging is critical to detection performance in these algorithms; however, for imaging under daylight conditions, it becomes necessary to create a long-exposure image as the sum of many short-exposure images. This paper explores the potential for increasing detection capabilities for small and dim space objects in a stack of short-exposure images dominated by a bright background. The algorithm proposed in this paper improves the traditional stack and average method of forming a long-exposure image by selectively removing short-exposure frames of data that do not positively contribute to the overall signal-to-noise ratio of the averaged image. The performance of the algorithm is compared to a traditional matched filter detector using data generated in MATLAB as well as laboratory-collected data. The results are illustrated on a receiver operating characteristic curve to highlight the increased probability of detection associated with the proposed algorithm.

Method for Estimating Three-Dimensional Knee Rotations Using Two Inertial Measurement Units: Validation with a Coordinate Measurement Machine.

Three-dimensional rotations across the human knee serve as important markers of knee health and performance in multiple contexts including human mobility, worker safety and health, athletic performance, and warfighter performance. While knee rotations can be estimated using optical motion capture, that method is largely limited to the laboratory and small capture volumes. These limitations may be overcome by deploying wearable inertial measurement units (IMUs). The objective of this study is to present a new IMU-based method for estimating 3D knee rotations and to benchmark the accuracy of the results using an instrumented mechanical linkage. The method employs data from shank- and thigh-mounted IMUs and a vector constraint for the medial-lateral axis of the knee during periods when the knee joint functions predominantly as a hinge. The method is carefully validated using data from high precision optical encoders in a mechanism that replicates 3D knee rotations spanning (1) pure flexion/extension, (2) pure internal/external rotation, (3) pure abduction/adduction, and (4) combinations of all three rotations. Regardless of the movement type, the IMU-derived estimates of 3D knee rotations replicate the truth data with high confidence (RMS error < 4 ° and correlation coefficient r ≥ 0.94 ).

Estimating Stair Running Performance Using Inertial Sensors.

Stair running, both ascending and descending, is a challenging aerobic exercise that many athletes, recreational runners, and soldiers perform during training. Studying biomechanics of stair running over multiple steps has been limited by the practical challenges presented while using optical-based motion tracking systems. We propose using foot-mounted inertial measurement units (IMUs) as a solution as they enable unrestricted motion capture in any environment and without need for external references. In particular, this paper presents methods for estimating foot velocity and trajectory during stair running using foot-mounted IMUs. Computational methods leverage the stationary periods occurring during the stance phase and known stair geometry to estimate foot orientation and trajectory, ultimately used to calculate stride metrics. These calculations, applied to human participant stair running data, reveal performance trends through timing, trajectory, energy, and force stride metrics. We present the results of our analysis of experimental data collected on eleven subjects. Overall, we determine that for either ascending or descending, the stance time is the strongest predictor of speed as shown by its high correlation with stride time.

Unequal a priori probability multiple hypothesis testing in space domain awareness with the space surveillance telescope.

This paper investigates the ability to improve Space Domain Awareness (SDA) by increasing the number of detectable Resident Space Objects (RSOs) from space surveillance sensors. With matched filter based techniques, the expected impulse response, or Point Spread Function (PSF), is compared against the received data. In the situation where the images are spatially undersampled, the modeled PSF may not match the received data if the RSO does not fall in the center of the pixel. This aliasing can be accounted for with a Multiple Hypothesis Test (MHT). Previously, proposed MHTs have implemented a test with an equal a priori prior probability assumption. This paper investigates using an unequal a priori probability MHT. To determine accurate a priori probabilities, three metrics are computed; they are correlation, physical distance, and empirical. Using the calculated a priori probabilities, a new algorithm is developed, and images from the Space Surveillance Telescope (SST) are analyzed. The number of detected objects by both an equal and unequal prior probabilities are compared while keeping the false alarm rate constant. Any additional number of detected objects will help improve SDA capabilities.

Improving space domain awareness through unequal-cost multiple hypothesis testing in the space surveillance telescope.

This paper investigates algorithms to improve the detection of space objects with the space surveillance telescope (SST) system. These space objects include natural objects such as asteroids and artificial satellites in Earth orbit. Using a proposed multiple hypothesis test (MHT), the detection performance is compared to the currently used algorithm as well as a matched filter and an equal-cost MHT algorithm. To compare these algorithms, a data set collected by the SST of a geosynchronous Earth orbit satellite, ANIK-F1 entering the Earth's eclipse, is utilized. It is found that an unequal-cost MHT gives increased performance over a point detector, a matched filter, and equal-cost MHT over a large range of potential intensities. Results are presented as probability of detection and receiver operating characteristic curves. In addition, the performance of the algorithm as a function of number of hypotheses used is investigated.

Classification of human walking context using a single-point accelerometer.

Real-world walking data offers rich insights into a person's mobility. Yet, daily life variations can alter these patterns, making the data challenging to interpret. As such, it is essential to integrate context for the extraction of meaningful information from real-world movement data. In this work, we leveraged the relationship between the characteristics of a walking bout and context to build a classification algorithm to distinguish between indoor and outdoor walks. We used data from 20 participants wearing an accelerometer on the thigh over a week. Their walking bouts were isolated and labeled using GPS and self-reporting data. We trained and validated two machine learning models, random forest and ensemble Support Vector Machine, using a leave-one-participant-out validation scheme on 15 subjects. The 5 remaining subjects were used as a testing set to choose a final model. The chosen model achieved an accuracy of 0.941, an F1-score of 0.963, and an AUROC of 0.931. This validated model was then used to label the walks from a different dataset with 15 participants wearing the same accelerometer. Finally, we characterized the differences between indoor and outdoor walks using the ensemble of the data. We found that participants walked significantly faster, longer, and more continuously when walking outdoors compared to indoors. These results demonstrate how movement data alone can be used to obtain accurate information on important contextual factors. These factors can then be leveraged to enhance our understanding and interpretation of real-world movement data, providing deeper insights into a person's health.

Gait kinematics differ by bout duration and setting.

BACKGROUND: Gait kinematics differ between settings and among young and older adults with and without knee osteoarthritis. Out-of-lab data has a variety of walking bout characteristics compared to controlled in-lab settings. The effect of walking bout duration on gait analysis results is unclear, and there is no standardized procedure for segmenting or selecting out-of-lab data for analysis. RESEARCH QUESTION: Do gait kinematics differ by bout duration or setting in young and older adults with and without knee osteoarthritis? METHODS: Ten young (28.1±3.5 yrs), ten older adults (60.8±3.3 yrs), and ten older adults with knee osteoarthritis (64.1±3.6 yrs) performed a standard in-lab gait analysis followed by a prescribed walking route outside the lab at a comfortable speed with four IMUs. Walking speed, stride length, and sagittal hip, knee, and ankle angular excursion (ROM) were calculated for each identified stride. Out-of-lab strides included straight-line, level walking divided into strides that occurred during long (>60 s) or short (≤60 s) bouts. Gait kinematics were compared between in-lab and both out-of-lab bout durations among groups. RESULTS: Significant main effects of setting or duration were found for walking speed and stride length, but there were no significant differences in hip, knee, or ankle joint ROM. Walking speed and stride length were greater in-lab followed by long and short bout out-of-lab. No significant interaction was observed between group and setting or bout duration for any spatiotemporal variables or joint ROMs. SIGNIFICANCE: Out-of-lab gait data can be beneficial in identifying gait characteristics that individuals may not encounter in the traditional lab setting. Setting has an impact on walking kinematics, so comparisons of in-lab and free-living gait may be impacted by the duration of walking bouts. A standardized approach for to analyzing out-of-lab gait data is important for comparing studies and populations.

Development and evaluation of the ARM algorithm: A novel approach to quantify musculoskeletal disorder risk factors in manual wheelchair users in the real world.

This study aimed to develop and evaluate the ARM (arm repetitive movement) algorithm using inertial measurement unit (IMU) data to assess repetitive arm motion in manual wheelchair (MWC) users in real-world settings. The algorithm was tested on community data from four MWC users with spinal cord injury and compared with video-based analysis. Additionally, the algorithm was applied to in-home and free-living environment data from two and sixteen MWC users, respectively, to assess its utility in quantifying differences across activities of daily living and between dominant and non-dominant arms. The ARM algorithm accurately estimated active and resting times (>98%) in the community and confirmed asymmetries between dominant and non-dominant arm usage in in-home and free-living environment data. Analysis of free-living environment data revealed that the total resting bout time was significantly longer (P = 0.049) and total active bout time was significantly shorter (P = 0.011) for the non-dominant arm. Analysis of active bouts longer than 10 seconds showed higher total time (P = 0.015), average duration (P = 0.026), and number of movement cycles per bout (P = 0.020) for the dominant side. These findings support the feasibility of using the IMU-based ARM algorithm to assess repetitive arm motion and monitor shoulder disorder risk factors in MWC users during daily activities.

Phase retrieval and Zernike decomposition using measured intensity data and the estimated electric field.

A new direct search phase retrieval technique for determining the optical prescription of an imaging system in terms of Zernike coefficients is described. The technique provides coefficient estimates without the need to defocus point source images to generate phase diversity by using electric field (E-field) estimates in addition to intensity data. Numerical analysis shows that E-field patterns in the image plane produced by the Zernike polynomials are less correlated with each other than the intensity patterns. Therefore, the E-field pattern provides more information for Zernike coefficient estimation than the intensity pattern alone. The phase retrieval is accomplished through an iterative process that uses the measured point source data to estimate the E-field pattern in the image plane with the Gerchberg-Saxton (GS) algorithm. The estimated E-field is correlated with a modeled E-field to produce estimates of the Zernike coefficients. Then the coefficients that minimize the error between measured data and the intensity model are selected. By using E-field estimates rather than phase estimates from the GS algorithm, the limitations of phase unwrapping for Zernike decomposition are avoided. Simulated point source data shows the new phase retrieval algorithm avoids getting trapped in local minima over a wide range of random aberrations. Experimental point source data are used to demonstrate the phase retrieval effectiveness.

Challenges and advances in the use of wearable sensors for lower extremity biomechanics.

The use of wearable sensors for the collection of lower extremity biomechanical data is increasing in popularity, in part due to the ease of collecting data and the ability to capture movement outside of traditional biomechanics laboratories. Consequently, an increasing number of researchers are facing the challenges that come with utilizing the data captured by wearable sensors. These challenges include identifying/calculating meaningful measures from unfamiliar data types (measures of acceleration and angular velocity instead of positions and joint angles), defining sensor-to-segment alignments for calculating traditional biomechanics metrics, using reduced sensor sets and machine learning to predict unmeasured signals, making decisions about when and how to make algorithms freely available, and developing or replicating methods to perform basic processing tasks such as recognizing activities of interest or identifying gait events. In this perspective article, we present our own approaches to common challenges in lower extremity biomechanics research using wearable sensors and share our perspectives on approaching several of these challenges. We present these perspectives with examples that come mostly from gait research, but many of the concepts also apply to other contexts where researchers may use wearable sensors. Our goal is to introduce common challenges to new users of wearable sensors, and to promote dialogue amongst experienced users towards best practices.