Kathleen H. Sienko

Appropriate exercise dosing is essential for optimizing functional outcomes in rehabilitation for individuals with neurological disorders. While guidelines for aerobic and resistance training intensity are well established, comparable standards for balance exercise remain limited, despite its central role in reducing falls and promoting safe mobility. Clinically feasible approaches to quantifying balance exercise difficulty are needed to support safe progression and individualized prescription. This exploratory study explored the relationship between participants perceived balance difficulty and clinician-rated assistance during progressive balance training. Sixteen adults (68.1 ± 8.0 years old; 69% female) with balance disorders (peripheral vestibular hypofunction, peripheral neuropathy, or age-related balance impairment) completed an 18-session balance rehabilitation program over 6-weeks. Participants and clinicians rated each exercise using standardized scales, yielding 1,658 rating comparisons. Agreement between participant-perceived difficulty and clinician-rated assistance was moderate [Cohen's kappa ( k ) = 0.42, 59% exact agreement , p < 0.001], with higher agreement observed using weighted analysis (linear weighted k = 0.58; quadratic weighted k = 0.74, both p < 0.001). Most rating discrepancies (85%) differed by only one category, suggesting clinically minor disagreements. These preliminary findings indicate that participant-perceived difficulty may provide a clinically accessible indicator of balance exercise challenge. However, validation in larger samples with independent rating procedures and examination of relationships between difficulty ratings and functional outcomes are needed before clinical implementation. This work provides initial evidence for incorporating participant feedback into balance exercise prescription as part of individualized, patient-centered rehabilitation approaches.

Background Balance is a fundamental component of daily activities and plays a critical role in preventing falls. Balance can be influenced by a variety of factors, including age-related physiological changes, making it important to consider age when assessing balance performance. However, an empirical basis for estimating the difficulty of balance exercises has yet to be developed. The primary aim of this study was to determine the effect of age and different balance exercise conditions on difficulty of exercises as determined by self-reported perceived difficulty, and to show that Rating of Perceived Difficulty (RPD) can serve as a practical measure of difficulty for guiding balance exercise prescription and progression. Methods Sixty-two healthy adults between the ages of 20 and 85 years with a mean age of 55 ± 20 years (50% female) participated in this cross-sectional study. Subjects performed four 30 s trials of 24 static standing balance exercises, and the average of these four trials was used for analysis. For each exercise, subjects’ ratings of perceived difficulty (RPD) were recorded using a 0–10 scale. Results A significant increase in RPD across all balance exercises occurred as age increased ( p < 0.02). From the youngest age group to the oldest, RPD increased by more than 100%. Ratings of perceived difficulty increased on a foam surface (110%), eyes closed (43%), semi-tandem stance (150%), and head movement (50%) compared with a firm surface, eyes open, feet apart, and head held still, respectively ( p < 0.01). Conclusion RPD measurements across a range of standing balance exercises can be used as a measure of difficulty and a practical tool for prescribing and progressing balance exercises in rehabilitation programs.

Abstract Engineering design applications that emphasize positive societal impacts are growing in popularity, yet often overlook the critical importance of engineering designers’ and stakeholders’ positionalities – their unique identities, experiences and resulting perspectives and social positions relative to others – in shaping design decisions. Insufficient attention to positionality can limit designers’ abilities to navigate complex problem contexts, engage diverse perspectives and address power dynamics, ultimately constraining the effectiveness and equity of design outcomes. However, little is known about how designers conceptualize and account for positionality in practice, particularly in the early stages of design when problem framing decisions are made. Therefore, this study explored how 10 engineering students and 10 practitioners conceptualized positionality in the initial stages of design for “social good,” where its impacts are especially pronounced. Each participant engaged in a written reflection and semistructured interview. Key findings include limitations in participants’ available language and strategies for accounting for positionality in design processes, particularly in the early stages, and that participants’ learning about positionality was largely driven by exposure to diverse identities and contexts. These insights highlight the limitations of engineering training and skillsets in design-for-social-good and emphasize the need for strategic, intentional consideration of positionality in design practice and education.

BACKGROUND: Effective balance rehabilitation requires training at an appropriate level of exercise intensity given an individual's needs and abilities. Typically balance intensity is assessed through in-clinic visual observation by physical therapists (PTs), which limits the ability to monitor and progress intensity during home-based components of training programs. The goal of this study was to train and evaluate machine learning models for estimating physical therapists' perceived balance exercise intensity using data from full-body wearable sensors to support the development of home-based training exercise dosage monitoring. METHODS: Balance exercise participants (n = 47) participated in a single-day balance training session where they were filmed performing static standing exercises at various levels of intensity. Kinematic data from 13 full-body wearable inertial measurement units (IMUs) and self-ratings of balance intensity were also collected. An additional cohort of PT participants (n = 42) was recruited to watch the videos of the balance exercise participants and provide ratings of balance intensity. The mean PT rating for each video was used as a ground truth (GT) label of balance intensity. We trained and evaluated Convolutional Neural Networks (CNN)-based models to predict balance intensity based on performance as captured through the IMUs. Model performance was evaluated by calculating the root-mean-square error (RMSE) of predications. A sensitivity analysis was also performed to assess the effect of the number of IMUs used on model performance. RESULTS: Models trained on orientation derived from all 13 IMUs achieved good predictive performance as indicated by a RMSE of 0.66 [0.62, 0.69], which was within the threshold defined by typical inter-rater variabilities between PTs (RMSE of 0.74 [0.72, 0.76]). Sensitivity analysis indicated that model performance stabilized at four sensors with the best performance corresponding to sensors placed on both thighs and the lower and upper back. CONCLUSIONS: Findings from this study indicated that balance intensity assessment can be achieved through wearable sensors and a CNN model, which could support the supervision and effectiveness of home-based balance rehabilitation.

To develop and evaluate an external cephalic version (ECV) simulation-based educational program. We used Kern’s 6-Step Framework to develop a program composed of six learning objectives with six distinct components specifically for low-resource settings. We conducted a three-round modified Delphi panel to modify the program. We evaluated the program with a cluster randomized study of obstetrics and gynecology house officers, junior residents, senior residents/fellows, and consultants/attendings in January 2025 at Korle Bu Teaching Hospital in Ghana. Self-reported comfort and confidence, knowledge, skill, and program feasibility and acceptability were analyzed with t-test, Fisher’s exact, and Spearman Rho. The Delphi panel consisted of 14 experts from five countries. At least 80% consensus on all ECV program components was reached prior to implementation. The learning group’s self-reported comfort and confidence (p<0.01), and knowledge measured via assessment (p<0.01) improved post- compared to pre-program. The learning group’s skill measured via procedural checklist increased compared to the control group (p<0.01). Within the learning group, self-reported comfort and confidence (p<0.01), and skill completing key steps on the procedural checklist (p=0.02) was higher among those with a higher compared to a lower level of training. The participants were satisfied with the program and found it acceptable, appropriate, and feasible (5-point rating, mean 4.42, standard deviation 0.54). Our ECV simulation-based program improves comfort and confidence, knowledge, and skill among Ghanaian clinicians. Given ECV success depends on the skill of the performing clinician, our program may help increase ECV rates and in result, decrease avoidable cesarean delivery rates and its complications.

Reduced fetal movement (RFM) can indicate that a fetus is at risk, but current monitoring methods provide only a "snapshot in time" of fetal health and require trained clinicians in clinical settings. To improve antenatal care, there is a need for continuous, objective fetal movement monitoring systems. Wearable sensors, like inertial measurement units (IMUs), offer a promising data-driven solution, but distinguishing fetal movements from maternal movements remains challenging. The potential benefits of using linear acceleration and angular rate data for fetal movement detection have not been fully explored. In this study, machine learning models were developed using linear acceleration and angular rate data from twenty-three participants who wore four abdominal IMUs and one chest reference while indicating perceived fetal movements with a handheld button. Random forest (RF), bi-directional long short-term memory (BiLSTM), and convolutional neural network (CNN) models were trained using hand-engineered features, time series data, and time-frequency spectrograms, respectively. The results showed that combining accelerometer and gyroscope data improved detection performance across all models compared to either one alone. CNN consistently outperformed other models but required larger datasets. RF and BiLSTM, while more sensitive to signal noise, offered reasonable performance with smaller datasets and greater interpretability.

Physical therapists (PTs) play an important role in balance rehabilitation as they make observations of patients to aid their decision making. Eye tracking can provide a record of these observations. Prior work has used eye tracking to compare PTs' visual behaviors across different experience levels; however, these studies have not considered how PTs' visual behaviors change with respect to patients' performances. This study aimed to identify the regions of the body PT participants focused on while evaluating standing balance and whether these regions of interest change with different observed levels of balance performance. Eleven pairs of older adults and PTs participated. The PT participants wore eye tracking glasses while watching the older adult participants perform standing balance exercises and provided performance ratings. Areas of interest, including the head, torso, upper extremities (UE), and lower extremities (LE), were defined to quantify the number and average duration of visits. PTs had the most and longest visits to the LE and the fewest and shortest visits to the UE across all ratings. As balance performance worsened, PTs increased the number of visits to the head, while decreasing the average visit duration to the torso and LE. These results suggest lower body balance strategies are important visual characteristics PTs consider while evaluating balance performance and, as exercises become more challenging, PTs demonstrate increasingly rapid visual scans of the whole-body to continually update their understanding of performance. This understanding of PT visual behaviors has implications for the future development of PT-informed balance assessment models.

Disorders affecting the vestibular system affect the ability to maintain balance and increase the risk of falls. Identifying individuals with vestibular deficits is an important step to inform referral to specialized testing, vestibular rehabilitation, and fall-prevention interventions. Instrumented gait assessments using wearable inertial measurement units (IMUs) and machine learning (ML) algorithms could support the accurate and automated identification of individuals with vestibular deficits. While prior work has demonstrated the feasibility of the automatic classification of vestibular gait, it relied on manual feature-engineering whereby discriminative features are identified and calculated prior to model training. The goal of this study was to develop and validate ML models that automatically learn from minimally pre-processed IMU data to classify gait kinematics from individuals with vestibular deficits and age-matched controls. Thirty study participants (15 with vestibular deficits and 15 age-matched controls) walked with their eyes closed on a 6-meter walkway with an IMU placed on the left arm. Two Bi-directional LSTM (BiLSTM) models were trained on the minimally pre-processed timeseries data alone as well as fusing the timeseries data with engineered features used in prior work. Classification performance was reported and compared to performance from feature-based approaches in terms of area under the receiver operating characteristic curve (AUROC) scores. Results showed that the BiLSTM models trained on minimally pre-processed time-series data achieved excellent classification performance (AUROC = 0.86), and their performance was comparable (p-value > 0.05) to previously published Random Forest models trained on engineered gait features extracted from the same dataset (AUROC = 0.89). These findings highlight that BiLSTM models were able to learn discriminative patterns from the minimally pre-processed IMU data in vestibular gait classification tasks.

Fetal movement is a commonly monitored indicator of fetal wellbeing with reductions in fetal movement being associated with poor perinatal outcomes. However, more informative datasets of fetal movement are required for improved clinical decision making. Wearable sensors coupled with machine learning (ML) methods could support accurate detection of fetal movement. While prior work has demonstrated the feasibility of accelerometer-based detection, using angular rate data to train ML models has not been fully explored. The goal of this study was to train and validate ML models using acceleration and angular rate features for detection of fetal movement. Ten pregnant participants wore an array of four abdominal inertial measurement units (IMUs) and one chest reference, while holding a toggle for maternal perception of fetal movement. Three random forest classifiers were trained on acceleration features, angular rate features, and a combination of both feature sets, respectively. Classification performance was reported in terms of area under the receiver operating characteristic curve (AUROC) and standard performance metrics. Results showed that all three models achieved good performance (AUROC = 0.70-0.77). The model combining acceleration and angular rate features achieved a notably higher positive predictive value (PPV) compared to the other models developed, indicating discriminative power over either feature set alone.

Visual observations provide an important foundation for physical therapist (PT) decision making in balance rehabilitation. This study used eye tracking to identify areas of the body focused on by PTs while evaluating standing balance across different exercise difficulties. Eye tracking data were obtained from five PTs who observed and rated older adults performing standing balance exercises, using a five-point scale. Excluding ratings of five, PTs had an average of approximately 215% more mean visits and an average of approximately 195% longer mean average visit durations to the lower extremities (LE) compared to the other regions of the body. However, visits to the LE shortened with increasing balance challenge, with approximately 75% shorter mean average visit durations to the LE for exercises rated as five compared to exercises rated as one. The number of visits to the head increased with balance challenge, with approximately 230% more mean number of visits to the head for ratings of five compared to ratings of one. Shortened visit durations and an increased number of visits to the upper body suggest an increase in whole body scan patterns with increasing challenge level.

Direct observation of design contexts allows engineers to collect detailed data in ways that are not possible with other methods, and is therefore a key method in sociotechnical engineering design, especially during the front-end of design processes. The development of design observation skills for engineering students presents challenges, however, including the effort required to reach representative observation sites and the uncertainty involved in real-world design environments. Students have often struggled to demonstrate recommended practices in sociotechnical design activities such as observation, yet may need observation skills during design project opportunities including curricular and co-curricular design projects. In addition, skills development may be especially challenging and critical when design environments are difficult to access, such as those in sensitive or remote locations. Therefore, this study explored the efficacy of a prototype VR-based design observation training tool with four undergraduate students engaged in a co-curricular global health technology design program. Participants were first given classroom-and VR-based design observation training, then interviewed before and after real-world design observation practice to elicit perceptions of the advantages, limitations, and overall effectiveness of the VR training experience. Across approximately six hours of collected interview data, participants reported positive general perceptions of the VR tool, which was described as more engaging and realistic than classroom-based training. Participants also discussed the limits of VR in preparing them for real-world observation, and technical and usability limitations of the VR system; they also identified variables to consider for the design of future design observation tools. Overall, participants suggested that VR may be most valuable as a complementary tool to other training formats.

Abstract Incorporating relevant contextual factors, e.g., socio-cultural, environmental, and industrial considerations, during design processes is required to develop solutions that function appropriately in their intended context of use, particularly in global health settings. Prior work has determined that "lacking the contextual knowledge needed" is a common reason for the failure of engineering projects intended for use in Low- and Middle-Income Countries (LMICs). Our prior work has investigated which contextual factors engineering designers consider and how they incorporate contextual factors into their global health design processes. In this study, we extended this prior research to compare the design behavior of novice and experienced global health engineering designers. As part of this research, we conducted semi-structured interviews with fifteen experienced design engineers who work on health-related technologies in LMICs. We also conducted semi-structured interviews and reviewed final reports from six mechanical engineering capstone teams working on global health-themed projects. While novices tended to aggregate many different "low-resource" contexts together, experienced global health designers exhibited a much more nuanced view of differences across unique LMIC contexts. We also identified that experienced designers regularly reframed their design problems and accounted for implementation decisions throughout their design processes, while novices viewed problem framing and implementation as largely outside the scope of their projects. In this study, we describe the preliminary conceptions of a framework that could support engineering design students during both curricular and co-curricular design activities. The framework guides students through multiple categories of contextual factors and provides examples and prompts for methods of incorporating contextual factors into decisions iteratively throughout their design processes in a curricular engineering design project. The findings from this work have implications for engineering design pedagogy and, ultimately, the potential to improve engineering graduates' abilities to develop contextually suitable solutions.

Abstract Incorporating contextual factors into engineering design processes is recommended to develop solutions that function appropriately in their intended use contexts. In global health settings, failing to tailor solutions to their broader context has led to many product failures. Since prior work has thus far not investigated the use of contextual factors in global health design practice, we conducted semi-structured interviews with 15 experienced global health design practitioners. Our analysis identified 351 instances of participants incorporating contextual factors in their previous design experiences, which we categorized into a taxonomy of contextual factors, including 9 primary and 32 secondary classifications. We summarized and synthesized key patterns within all the identified contextual factor categories. Next, this study presents a descriptive model for incorporating contextual factors developed from our findings, which identifies that participants actively sought contextual information and made conscious decisions to adjust their solutions, target markets and implementation plans to accommodate contextual factors iteratively throughout their design processes. Our findings highlight how participants sometimes conducted formal evaluations while other times they relied on their own experience, the experience of a team member or other stakeholder engagement strategies. The research findings can ultimately inform design practice and engineering pedagogy for global health applications.

Abstract Many scholars and practitioners claim that engineered solutions must be contextually appropriate, particularly when designing for global health applications. Although neglecting to understand context has been cited as a common cause of project failure in global health work, engineering designers and researchers have a limited number of available tools, methods, and frameworks to support the collection and application of contextual information to design decisions. This paper presents findings from multiple studies and proposes an advancement to the definition of context and a preliminary framework for using contextual factors in engineering design. The framework includes definitions for two primary concepts: (1) classifications, i.e., defined categories of context called “contextual factors”, and (2) applications, i.e., specific strategies and methods for incorporating contextual factors into various design stages. We provide examples of the framework’s potential use in design practice and research, such as improving reporting in design literature and assessing the transferability of solutions from one context to another. Ultimately this work advances our understanding and definition of “context” in engineering design processes for global health applications. We aim for these findings to advance the fields of mechanical engineering and design science and seed subsequent scholarship aimed at evolving the theoretical constructs associated with gathering, synthesizing, and applying contextual information within engineering design work.

Introduction: In order for balance therapy to be successful, the training must occur at the appropriate dosage. However, physical therapist (PT) visual evaluation, the current standard of care for intensity assessment, is not always effective during telerehabilitation. Alternative balance exercise intensity assessment methods have not previously been compared to expert PT evaluations. The aim of this study was therefore to assess the relationship between PT participant ratings of standing balance exercise intensity and balance participant self-ratings or quantitative posturographic measures. Methods: Ten balance participants with age or vestibular disorder-related balance concerns completed a total of 450 standing balance exercises (three trials each of 150 exercises) while wearing an inertial measurement unit on their lower back. They provided per-trial and per-exercise self-ratings of balance intensity on a scale from 1 (steady) to 5 (loss of balance). Eight PT participants reviewed video recordings and provided a total of 1,935 per-trial and 645 per-exercise balance intensity expert ratings. Results: PT ratings were of good inter-rater reliability and significantly correlated with exercise difficulty, supporting the use of this intensity scale. Per-trial and per-exercise PT ratings were significantly correlated with both self-ratings (r = 0.77-0.79) and kinematic data (r = 0.35-0.74). However, the self-ratings were significantly lower than the PT ratings (difference of 0.314-0.385). Resulting predictions from self-ratings or kinematic data agreed with PT ratings approximately 43.0-52.4% of the time, and agreement was highest for ratings of a 5. Discussion: These preliminary findings suggested that self-ratings best indicated two intensity levels (i.e., higher/lower) and sway kinematics were most reliable at intensity extremes.

High quality user requirements are positively correlated with successful design outcomes, but engaging stakeholders within low-income contexts can present financial and time-related challenges to product developers from non-local industrial and academic institutions with limited knowledge of the context. Existing literature provides guidance for engaging stakeholders during the early stages of product design in high-income country contexts, but few studies have examined the effectiveness of these methods in low-income country contexts. This study evaluated three user requirements elicitation and prioritization methods including open-ended, clustering, and discrete choice. Ghanaian healthcare delivery stakeholders with varying types of expertise, years of experience, and from various types of healthcare facilities were recruited to allow for diversity of responses. Participants included physicians (n = 10), nurses/midwives (n = 16), biomedical technicians (n = 14), and public health officers (n = 7). A hypothetical mechanical device for managing and treating postpartum hemorrhage was chosen to characterize each method's ability to elicit and prioritize user requirements. The open-ended method captured general requirements of a design concept, yet resulted in predominantly generic requirements. The results from the open-ended method were used to inform the clustering and discrete choice methods. The clustering and discrete choice methods were useful for inferring in-depth user requirements and eliciting stakeholder priorities. The clustering method revealed that usability and affordability were high-priority requirements among all four stakeholder groups. An individual difference scaling analysis was performed using the clustering method outcomes, which indirectly identified ease-of-use, availability, and effectiveness as the priority user requirements categories. Stakeholders ranked ease-of-use as the highest-priority user requirement, followed by performance, cost, and place-of-origin requirements, using the discrete choice method. Given the significance of the ease-of-use requirement, an analytical framework based on sub-requirements was developed for quantifying stakeholder needs. Lastly, the relative merits of the three elicitation approaches and their implications for use with different stakeholder groups were examined.

A shift in the traditional technocentric view of medical device design to a human-centered one is needed to bridge existing translational gaps and improve health equity. To ensure the successful and equitable adoption of health technology innovations, engineers must think beyond the device and the direct end user and must seek a more holistic understanding of broader stakeholder needs and the intended context of use early in a design process. The objectives of this review article are ( a) to provide rationale for the need to incorporate meaningful stakeholder analysis and contextual investigation in health technology development and biomedical engineering pedagogy, ( b) to review existing frameworks and human- and equity-centered approaches to stakeholder engagement and contextual investigation for improved adoption of innovative technologies, and ( c) to present case studyexamples of medical device design that apply these approaches to bridge the gaps between biomedical engineers and the contexts for which they are designing.

Abstract Using prototypes to engage stakeholders during front-end design activities is crucial for successful design outcomes. Compared to prototyping that is used for iterative refinement during back-end engineering design activities, prototyping that informs problem definition, requirements and specifications development, concept generation, and other front-end design activities is understudied. To identify patterns in prototyping strategies for engaging stakeholders during the design front end, we conducted semi-structured interviews with 26 design practitioners across three product design domains: automotive, consumer products, and medical devices. Seventeen strategies evident across the collection of practitioners were used in generally consistent ways, with some variation based on context, e.g., project scope, stakeholders engaged, and the stakeholder interaction situation. Twelve of those 17 strategies were used by industry practitioners across the three domains, and five of those 17 strategies were used by practitioners from the medical device domain and either the automotive or consumer products domain. The descriptions and in-context examples of prototyping strategies used to engage stakeholders during front-end design can guide the design strategies of both experienced and novice designers.

Abstract Engineers must engage project stakeholders effectively if stakeholder needs are to be met, and prototypes are key tools for communicating design form and function. Quality stakeholder engagement in the front end of design processes, in particular, is critical in the success or failure of design projects. As remote stakeholder engagement has become increasingly common as industry trends toward distributed design, there is a need to develop the theory and practices behind effective remote design processes, which have not yet been as well-studied as in-person design. This study explored the prototyping strategies for remote stakeholder engagement during front-end design used by 10 engineering practitioners and 10 senior engineering students through semi-structured interviews. Prototyping strategies were found to overlap with many of the strategies described by prior literature that are not specific to remote engagement modes, though several of these strategies were adapted to the remote context, and three emergent strategies for prototyping in remote engagements were identified. Designers’ perceptions of remote versus in-person prototyping strategies for stakeholder engagement in front-end design, including perceived advantages and limitations, were also explored, and recommendations for educators to better prepare engineering students for hybrid and remote work are provided.