A team of researchers led by Iris Steineck and Nicolas Kurpiers of the University of Hildesheim, together with Uwe G. Kersting of the German Sport University Cologne, set out to test a suspicion that ski guides had long whispered about but never verified. Instructors in the German inclusive skiing program Snow & Eyes had repeatedly noticed that blind skiers seem to lean toward the acoustic signal of their guide. Skiers guided from the front appeared to lean forward; skiers guided from behind seemed to sit back on their skis. If true, this would matter enormously, because a backward-leaning position in alpine skiing is not merely a stylistic quirk. It is a known ingredient in some of the sport’s most dangerous knee injury mechanisms.

The two dominant guiding strategies in visually impaired alpine skiing are known as Front Guide and Rear Guide. In the Front Guide variant, the guide skis ahead and calls the line, with the visually impaired skier chasing the sound from a distance of roughly one to two and a half ski lengths. In the Rear Guide variant, the roles reverse: the guide skis behind, shouting commands toward the back of the skier’s head. Competitive Para alpine racing under the International Ski and Snowboard Federation mandates the Front Guide setup at World Championships and World Cup events, but in recreational skiing the choice varies by club, ski school, and personal preference, and the community has debated the merits of each variant for decades without hard data.

To settle the question, the researchers recruited 18 skiers, twelve of them fully blind and six severely visually impaired under German legal criteria, meaning a visual acuity of 0.05 or lower. The participants ranged from 15 to 60 years old, from absolute beginners to experts with up to 500 days of skiing experience, and some had learned to ski before losing their sight while others learned with their impairment already present. Each skier was tested in both guiding conditions on the same day, on the same groomed blue slope with a maximum inclination of 25 percent, in Norway, Austria, Germany, or Switzerland, with temperatures between minus 8 and plus 2 degrees Celsius and stable, calm weather throughout.

The measurement technology was a full-body inertial motion capture system: a suit embedding 17 inertial measurement units sampling at 240 hertz, worn beneath the skier’s clothing. Such wearable IMU systems have emerged in recent years as robust tools for field biomechanics, capable of reconstructing a skier’s center of mass and joint angles outside the laboratory. Calibration posed its own challenge, since standard procedures require standing still with straight knees, which is difficult for blind participants balancing in rigid ski boots. The team’s solution was to unbuckle the boots entirely during the reference pose, allowing participants to straighten their legs for the roughly ten seconds required, and only calibration files rated highly by the system’s internal quality classification were used.

From the motion data, the researchers calculated a single elegant metric they called Cpos: the horizontal distance between the skier’s center of mass and the midpoint between the two ankle joints, projected onto the direction of travel. A positive value means the center of mass sits ahead of the feet, a forward lean; a negative value means it trails behind. Six rhythmic, consistent turns per run were averaged for each guiding variant, and the data were fed into a linear mixed-effects model that accounted for repeated measurements within each participant while testing the effects of guiding variant, accustomed technique, skiing experience, blindness severity, and whether the impairment was congenital or acquired.

The headline result confirmed the guides’ intuition. The center of mass sat significantly more backward during Rear Guide skiing than during Front Guide skiing, with a mean difference of 4.8 centimeters between the conditions. That may sound modest, but it is the same order of magnitude as the six-centimeter forward-lean difference the same research group previously measured between skiing in soft versus stiff ski boot shafts, a difference large enough to change how a skier loads and controls the skis. Intriguingly, individual responses varied widely, from minus 3 to plus 18 centimeters, suggesting that some skiers are far more sensitive to the location of their guide’s voice than others. Neither the guiding variant a skier was accustomed to nor total skiing experience had a significant effect, meaning the posture shift appears even in veterans with decades of muscle memory.

The model also uncovered two subtler effects. Skiers with residual vision showed more forward center-of-mass positions than fully blind skiers, plausibly because they can still extract visual cues from the terrain and therefore depend less on orienting their bodies toward the guide’s acoustic signal. And skiers with congenital impairments leaned less forward than those whose impairment was acquired, which the authors interpret as a matter of trust: without a lifetime of visually acquired movement representations to draw on, adopting a committed forward lean over the ski tips may demand a greater leap of faith. Notably, the effect of guiding variant itself did not differ by severity or onset of impairment, and whether the guide used a speaker system or an unamplified voice made no measurable difference in the Rear Guide condition.

Why does the voice pull the body? The most plausible explanation is that skiers orient toward the source of the sound they depend on for survival on the slope. In the Front Guide setup, the acoustic signal arrives from ahead, from relative azimuth angles spanning roughly plus or minus 120 degrees around the skier’s heading, and the skier follows it, tilting the head and trunk forward. In the Rear Guide setup, commands arrive from behind, and the body seems to rotate backward toward that crucial information stream. Earlier work supports the idea: researchers have described upper-body lean changing in visually impaired long jumpers as they approach an acoustic signal, and ski instructors have observed blind skiers abandoning optimal posture when commands originate from a poorly chosen position. The new study turns these anecdotes into numbers.

The safety implications are what make the finding potentially viral within the skiing community. Backward lean is implicated in the classic skiing knee injury mechanisms, including the so-called phantom foot pattern, in which a backward position combined with internal rotation of the tibia tears the knee, and the boot-induced anterior drawer, in which a flexed knee and backward position lever the tibia forward and threaten the anterior cruciate ligament. Because blind skiers cannot anticipate changes in slope or snow conditions the way sighted skiers can, they must react in real time, and a forward-ready position is considered central to that reactive capacity. The authors are careful to note that actual injury outcomes were not assessed and that a residual confounding effect of the audio equipment cannot be fully excluded, since all Front Guide trials used speaker amplification. Still, the alignment with established biomechanical reasoning is striking, and the team argues that the Front Guide variant may help skiers develop and maintain an adequate position on the skis, particularly during the learning stages.

The study is the first scientific investigation of its kind, and its authors are candid about the hurdles: recruiting blind skiers willing to ski with sensors and strangers took three years, four countries, and meticulous planning, including indoor rehearsal sessions and special training for helpers. But the payoff is a first quantitative foundation for what has until now been a matter of tradition and anecdote. For the roughly small population of blind and severely visually impaired skiers, their guides, their families, and the organizations that train them, the message is concrete: where the guide skis is not neutral. The voice that leads the skier down the mountain also, quite literally, leans their body toward safety or away from it, and choosing that position may be one of the simplest, cheapest interventions available for making the slopes more accessible to everyone.

Subject of Research: Effects of front versus rear acoustic guiding methods on the ski-specific body posture of blind and severely visually impaired skiers

Article Title: The effects of different guiding methods on ski-specific body posture of severely visually impaired and blind skiers

Article References: Steineck, I., Kersting, U. G., & Kurpiers, N. (2026). The effects of different guiding methods on ski-specific body posture of severely visually impaired and blind skiers. Sports Engineering, 29(2), Article 34. https://doi.org/10.1007/s12283-026-00566-3

Image Credits: AI Generated

DOI: 10.1007/s12283-026-00566-3

Keywords: visually impaired skiing, blind skiers, alpine skiing, guiding methods, body posture, center of mass, inertial measurement units, biomechanics, Paralympic sport, knee injury prevention, acoustic orientation, adaptive sports

A new review published in BioMedical Engineering OnLine by Syed Riyas Ahamed and Sandip Saha of the School of Advanced Sciences at Vellore Institute of Technology in Chennai, together with Awani Bhushan of the university’s School of Mechanical Engineering, takes stock of whether that barrier is finally falling. The team sifted through approximately 180 recent studies of low-cost gait analysis systems, comparing them against commercial systems that serve as performance benchmarks. Their verdict is nuanced and, in places, sobering: affordable sensors have matured dramatically, but price tells you almost nothing about how well a system performs, and laboratory success remains a poor promise of real-world reliability.

The review organizes the crowded landscape of low-cost gait technology into five modality families, each with its own physics and its own trade-offs. Vision-based systems use cameras and computer vision algorithms to track body landmarks and reconstruct movement in space. Inertial measurement units, or IMUs, are small wearable devices containing accelerometers and gyroscopes that measure acceleration and rotation as the body moves. Electromyography, or EMG, systems record the electrical signals that muscles generate when they contract. Pressure-sensing systems, including smart insoles and flexible sensors, map how force is distributed across the plantar surface of the foot during contact with the ground. Finally, hybrid multimodal systems combine two or more of these approaches, betting that fusing data streams will compensate for the weaknesses of each.

Each modality occupies a distinct niche in the measurement space. IMUs deliver high temporal resolution, which makes them excellent for capturing the fine timing of kinematic parameters such as stride duration and gait cadence. Vision-based systems offer rich spatial and kinematic information, reconstructing not just when the leg moves but where it moves in three-dimensional space. Pressure-sensing insoles capture plantar contact dynamics, the hidden story of how load shifts from heel to toe with every step. EMG-based systems contribute something the others cannot: complementary neuromuscular information, revealing whether the muscles driving the movement are firing in a healthy pattern. Hybrid systems, the review finds, do improve overall performance by integrating these streams, but they pay for that gain with added complexity in hardware, calibration, and data processing.

To judge these systems fairly, the authors evaluated them along three dimensions: quantitative performance, system cost, and clinical validation. Performance was assessed through error rates, reliability, and classification accuracy — how closely a cheap sensor’s output matches the gold standard, how repeatable its measurements are, and how well it can distinguish, say, a pathological gait pattern from a normal one. Costs were reported in the currencies used by the original studies, with US dollar values referenced where available, an approach the authors adopted deliberately to acknowledge regional variation in purchasing power and component pricing. That variation matters: a system assembled from off-the-shelf parts in one country may cost several times more in another, even before differences in labor and import duties are considered.

One of the review’s most striking findings is what the data refuses to show. Costs across the field span several orders of magnitude, from ultra-low-cost prototypes built for a few dozen dollars to commercial systems costing thousands, yet price remains a poor predictor of performance. A modestly priced IMU can, in the right application, rival far more expensive equipment, while an expensive system can still stumble on the metrics that matter. The authors also found that direct comparison across systems is genuinely difficult, because studies use heterogeneous evaluation metrics and report costs inconsistently. Two papers may both claim high accuracy while measuring entirely different things against entirely different reference standards, leaving clinicians and researchers without a common yardstick.

The clinical validation picture is equally revealing. Most validation studies of low-cost gait systems have been conducted on small, controlled samples — often healthy young volunteers or small cohorts of patients in tightly supervised settings. Geographically, the evidence base is heavily concentrated in North America, Europe, and East Asia, while South Asia, Sub-Saharan Africa, Latin America, and the Middle East are substantially underrepresented. That gap is more than a cartographic curiosity. Gait is shaped by body size, footwear habits, terrain, culture, and daily environment, and a system validated exclusively on one population and one continent cannot be assumed to transfer cleanly to a clinic in Lagos or Lima. For technologies whose central promise is accessibility, the irony is sharp: the regions that stand to benefit most from affordable gait analysis are the ones where the evidence is thinnest.

The review’s most cautionary conclusion concerns the chasm between the laboratory and the world outside it. Laboratory performance, the authors stress, does not guarantee real-world reliability. A camera-based system that tracks a subject flawlessly in a controlled room with even lighting may fail when sunlight flickers across a hallway or a walker passes through the frame. An insole whose sensors drift after a week of daily wear may produce beautiful data on day one and unreliable data by day seven. Real-world deployment demands robustness to varied footwear, uneven ground, uncontrolled lighting, sensor placement errors by non-expert users, and the sheer messiness of daily life — conditions that most validation studies, by design, never test.

Why does this matter beyond the biomechanics lab? Because gait analysis sits at the intersection of some of medicine’s most pressing challenges. Populations are aging, and falls among older adults are a leading cause of injury and loss of independence. Neurodegenerative conditions such as Parkinson’s disease alter gait patterns years before more obvious symptoms appear, making walking a potential early-warning signal. Rehabilitation after stroke or surgery depends on tracking whether a patient’s walking is actually improving, not just whether the patient feels it is. If low-cost, validated gait systems could move into homes, community clinics, and physiotherapy practices, continuous monitoring could replace the occasional snapshot of a single laboratory visit, capturing how a person walks on an ordinary Tuesday rather than on the artificial stage of a gait lab.

The path forward, according to the review, runs through three priorities: standardized evaluation, large-scale clinical validation, and cost transparency. Standardized evaluation would let researchers compare systems on common metrics instead of incomparable ones. Large-scale clinical validation would test these technologies on the diverse populations and uncontrolled environments where they are actually meant to work, filling the geographic and demographic gaps in the current evidence. Cost transparency would require honest, complete accounting of what systems truly cost to build, buy, and maintain. No single modality, the authors conclude, offers everything a clinician needs — IMUs excel at timing, cameras at space, insoles at pressure, EMG at muscle, and hybrids at integration. But if the field can agree on how to measure its own progress, the dream of gait analysis that costs hundreds rather than tens of thousands of dollars, available in any clinic on Earth, moves from aspiration toward engineering reality.

Subject of Research: Cross-modal evaluation of low-cost gait analysis systems for performance, cost, and clinical validation

Article Title: Toward accessible gait analysis: a cross-modal evaluation of performance, cost and clinical validation

Article References: Ahamed, S. R., Saha, S., & Bhushan, A. (2026). Toward accessible gait analysis: a cross-modal evaluation of performance, cost and clinical validation. BioMedical Engineering OnLine. https://doi.org/10.1186/s12938-026-01629-z

Image Credits: AI Generated

DOI: 10.1186/s12938-026-01629-z

Keywords: gait analysis, low-cost sensors, inertial measurement units, wearables, smart insoles, electromyography, vision-based motion capture, biomedical engineering, rehabilitation, Parkinson's disease, clinical validation, real-time monitoring

The review, conducted by Edwin Ho Yin Lui and Ralph Jasper Mobbs of the University of New South Wales and the NeuroSpine Surgery Research Group in Sydney, followed the PRISMA-ScR reporting guidelines and searched PubMed, Embase, and Scopus for peer-reviewed studies that used a smartphone’s internal inertial measurement unit, or IMU, to quantify mobility, gait, or postural transitions. From an initial yield of 1,063 records, 84 studies made the final cut. The researchers then charted each study across four domains: the clinical target and purpose, the methodological implementation, the technological infrastructure, and the extracted metrics and validation standards. The goal was not to pool results into a meta-analysis — the field is far too heterogeneous for that — but to map where the science actually stands and where it stalls on the road to the clinic.

The headline finding is that smartphone-based mobility analysis has spread far beyond the engineering lab. Neurology dominated the clinical landscape, accounting for 36.9 percent of studies, with Parkinson’s disease alone featured in 19 percent and multiple sclerosis in 6 percent. Geriatrics followed at 14.3 percent and orthopedics at 13.1 percent, with smaller contributions from rheumatology and oncology. Perhaps more striking is what researchers are trying to do with the data: only 10.7 percent of studies were primarily validation exercises, while the overwhelming majority — 89.3 percent — pursued clinically directed aims such as diagnosis (36.9 percent), prognostication (38.1 percent), or longitudinal monitoring (14.3 percent). In other words, the field has largely moved past asking whether phones can measure movement and is now asking whether those measurements mean something for patients.

The clinical logic is compelling. Walking speed and gait characteristics have been repeatedly linked to cognitive decline, mortality, neurodegenerative trajectories, and disease progression — leading some researchers to call gait the sixth vital sign. Yet routine assessment still relies heavily on patient-reported questionnaires and episodic clinic visits, which capture only a snapshot of function under artificial conditions. Laboratory gait analysis, with its three-dimensional optoelectronic motion capture and force plates, offers exquisite precision but is expensive, immobile, and utterly unsuited to continuous monitoring. Dedicated research wearables fare better but impose costs and sustained-use burdens of their own. Smartphones, by contrast, are already carried by billions of people every day, making them arguably the most scalable mobility-sensing platform ever deployed.

But the review’s detailed breakdown reveals a deep methodological disconnect between the promise and the practice. More than half of the studies — 54.8 percent — confined their assessments to controlled clinical or laboratory environments, and 67.9 percent required a clinician or researcher to supervise the test. A striking 73.8 percent depended on active testing protocols, in which participants deliberately performed structured tasks such as short walk tests, the timed up-and-go, or the six-minute walk test. Only 23.8 percent of studies captured passive, free-living mobility — the spontaneous, unscripted walking that happens in kitchens, sidewalks, and shopping malls. Similarly, 63.1 percent collected a single cross-sectional snapshot rather than continuous longitudinal data. The very feature that makes smartphones revolutionary — their unobtrusive presence in daily life — remains largely untapped.

The technological infrastructure tells a similar story of fragmentation. In 71.4 percent of studies, participants had to secure the phone to a fixed anatomical location, typically the waist or lower back, using belts, straps, or pouches — a constraint that partly recreates the burden of dedicated wearables and may be impractical for older adults or people with cognitive impairment. Software is even more fragmented: 82.1 percent of studies relied on custom-built applications or proprietary research platforms such as mPower, GaitTrack, and GaitMate, while only 7.1 percent integrated with native operating-system health platforms like Apple HealthKit or Google Fit. This patchwork of bespoke pipelines makes it difficult to compare results across studies, reproduce findings, or transport algorithms from one device model to another. Differences in accelerometer resolution, sampling frequency, timestamp accuracy, calibration, and sensor-fusion pipelines — not to mention operating-system updates and aggressive power-management policies that can silently interrupt background data collection — all conspire to undermine reproducibility.

What are researchers actually extracting from these sensors? The most common measures are the macro-mobility staples: gait speed appeared in 38.1 percent of studies, step count in 33.3 percent, cadence in 32.1 percent, and step length in 28.6 percent. Gait-quality metrics — stance-to-swing ratios, asymmetry indices, temporal variability, and nonlinear measures such as approximate entropy — were reported less often, despite their potential to reveal neurological and musculoskeletal dysfunction that simple step counts miss. Notably, nearly a third of studies skipped interpretable kinematic parameters altogether, feeding raw triaxial acceleration into machine-learning classifiers. Validation, where performed, typically involved comparison against three-dimensional motion capture, pressure mats, dedicated wearables, or clinical rating scales. The review’s authors are careful to distinguish what such comparisons prove: technical validity, meaning the measure is accurate and reliable; clinical validity, meaning it associates with or predicts a meaningful health state; and clinical utility, meaning it actually changes clinical decisions or improves patient outcomes. The mapped literature demonstrates the first two in progress — but routine clinical utility has not been established.

The ecological question looms largest. Free-living monitoring can capture fatigue, diurnal fluctuations, and long-term variability that a brief clinic visit cannot, and the review argues that passive and structured assessment should be treated as complementary rather than interchangeable. Yet real-world data come with real-world noise: walking surface, gradient, footwear, assistive devices, crowding, and whether the phone is in a pocket, hand, or bag can all shift the measured signal independently of any clinical change. Crucially, the authors argue, this contextual variation is not always noise to be eliminated — it may reveal how environmental demands shape functional performance. Emerging analytical strategies, including placement-robust models, automated detection and segmentation of walking bouts, steady-state filtering, and within-person longitudinal comparison, aim to separate context-related from health-related variation. The experience of the mPower study, which enrolled more than 12,000 participants remotely for Parkinson’s research but suffered substantial attrition, underscores that scalability does not guarantee sustained engagement — a burden that passive background architectures may help relieve.

Privacy and equity complete the translation checklist. Continuous passive sensing, even without GPS, can expose daily routines, periods of inactivity, and changes in health status, demanding transparent consent, data minimization, on-device processing where feasible, encryption, and predefined retention and deletion policies. Governance must also confront data ownership, linkage with health records, third-party access, and accountability for missed deterioration — plus equitable alternatives for patients who do not own or consistently carry compatible smartphones. The review’s bottom line is measured but optimistic: smartphones are genuinely promising candidate platforms for multidimensional mobility sensing, and the field’s shift from validation toward diagnosis and prognostication shows real momentum. What stands between promising digital measures and routine clinical practice is disciplined translation — standardized acquisition and reporting, cross-device and cross-platform validation, context-aware interpretation, privacy-preserving governance, and prospective trials proving that these pocket-sized sensors actually improve clinical decisions and patients’ lives.

Subject of Research: Smartphone-based mobility and gait analysis for clinical assessment

Article Title: Clinical utility and methodological implementation of smartphone-based mobility analysis: a scoping review

Article References: Lui, E. H. Y., & Mobbs, R. J. (2026). Clinical utility and methodological implementation of smartphone-based mobility analysis: a scoping review. Journal of Neurology, 273(10), Article 597. https://doi.org/10.1007/s00415-026-14131-2

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14131-2

Keywords: smartphones, gait analysis, mobility, digital biomarkers, neurology, Parkinson's disease, wearable technology, inertial measurement units, mHealth, scoping review, clinical utility, passive monitoring

Restraint, as the researchers define it, is any practice that limits an infant’s ability to move or change posture freely. That includes caregiver restraint, such as being held or carried, and device restraint, such as sitting in a high chair, stroller, swing, or car seat. The definition deliberately avoids a fixed list of equipment. Instead, it focuses on the infant’s real-time opportunities for movement, which means an infant strapped into a high chair counts as restrained, while an infant placed in a large crib where they can crawl and reposition freely does not. This broad framing matters because restraint is not simply a safety measure or a convenience for caregivers. It fundamentally shapes what infants can do, see, hear, and touch. Previous work has linked restraint to reduced leg movement and diminished spinal and hip muscle activity, and to fewer opportunities for object interaction and reduced exposure to adult speech. At the same time, restraint can be beneficial: supported sitting frees a baby’s hands for exploring objects, and being carried places infants higher off the ground, giving them richer views of faces and distant surroundings.

Measuring this everyday phenomenon has always been the bottleneck. Parent surveys, whether retrospective questionnaires or ecological momentary assessments delivered by phone several times a day, depend on memory or on sparse sampling, and neither can produce a continuous record of the day. Video recording solves the continuity problem in principle, but in practice it is cumbersome. Stationary cameras lose sight of infants behind furniture, handheld cameras can alter behavior, and the annotation process is so labor-intensive that prior studies typically captured only half an hour to an hour of footage, often biased toward periods when infants were awake and active. The periods most relevant to restraint research, such as a baby seated in front of a screen or parked in a seat during mealtime, were precisely the ones most likely to be skipped. What was needed was a method that could run all day, unobtrusively, without asking anything of the caregiver.

The solution the team built rests on inertial measurement units, miniature sensors that record linear acceleration from accelerometers and angular velocity from gyroscopes. In this study, four lightweight MC10 Biostamp sensors were secured in a customized infant garment at the right thigh, left thigh, right ankle, and left ankle. Each sensor sampled motion at 62.5 hertz, a rate fine enough to capture the subtle signatures of infant movement, and the battery lasted more than a full day. The dataset drew on 146 home-visit sessions from 66 infants, split into a younger cohort of 30 babies aged 4 to 7 months and an older cohort of 36 babies aged 11 to 14 months. During each visit, a GoPro camera recorded the first 1.5 hours of the session, and trained coders labeled every stretch of that footage as restrained, unrestrained, or missing when the infant left the frame. Two independent coders achieved 98.79 percent agreement, providing a trustworthy ground truth against which the machine learning model could be judged.

The modeling pipeline is a careful exercise in feature engineering. Raw motion data were aggregated within sliding windows, and within each window the researchers computed a rich set of statistics for every signal from every sensor: means, medians, minima, maxima, percentiles, skewness, kurtosis, standard deviations, and sums, along with cross-sensor and cross-orientation correlations and pairwise differences. In total, 436 features described each window. A random forest classifier, trained with leave-one-session-out cross-validation, then mapped those features onto the binary restrained or unrestrained label. Crucially, the team varied the window length, testing windows of 4, 16, and 30 seconds, because the choice of temporal scale involves a genuine trade-off. Short windows offer fine temporal resolution but may lack enough context to distinguish, say, a baby kicking freely on the floor from a baby kicking while strapped into a seat. Longer windows capture more context but risk blurring abrupt transitions.

All three window lengths performed well, with accuracies between 87 and 89 percent and kappa values between 0.70 and 0.73, but the 30-second window showed the most favorable pattern. The model carried a slight bias toward overestimating unrestrained time, likely because unrestrained periods were more common in the training data, and the longer window shrank that bias: the difference between predicted and human-annotated restraint prevalence fell from 4.70 percent with 4-second windows to 1.70 percent with 30-second windows. The authors offer an intuitive explanation for why longer windows help. Infants may look motionally similar whether restrained or not while they sit still, but the telltale differences emerge when they move: a caregiver’s lap or a seat absorbs and damps motion in ways that a bare floor does not. Those damping signatures appear only in movement episodes, and a 30-second window is far more likely to contain one. Longer windows also generalize better across the bewildering variety of restraint forms, from stationary seats to swings to being carried around the house. The practical lesson for the field is that window length should be matched to the natural duration of the behavior being detected, and restrained bouts, which typically last minutes rather than seconds, call for longer windows than the 2-to-4-second windows used in earlier body-posture classifiers.

When the final 30-second model was applied to full-day recordings, it produced estimates that converged with prior survey-based findings. Younger infants aged 4 to 7 months spent an average of 55.38 percent of their awake time restrained, while older infants aged 11 to 14 months spent 33.42 percent, a statistically significant age-related decline that mirrors earlier reports. Individual variability was striking, with some infants restrained far more than others even within the same age group. The average awake wearing time in the 133 analyzable full-day sessions was about 6.18 hours, after excluding naps and periods when the garment was removed, which caregivers logged on paper forms. The authors acknowledge that reliance on parent logs is a limitation, since caregivers may forget or misrecord these intervals, and they suggest future work could cross-validate wearing time using automated sleep detection from audio recordings or sensor-based movement detection.

The study is candid about the boundaries of its model. Machine learning classifiers are constrained by their training data, so strong performance on this dataset does not guarantee comparable accuracy on new populations, new sensor brands, or unfamiliar forms of restraint. The authors point to culturally distinctive practices such as the Tajik gahvora cradle, Tseltal slings, and the use of sandbags or heavy blankets elsewhere, all of which could produce hip and ankle movement patterns absent from the current training set. Applying the model in such contexts will require fresh ground-truth coding and possible re-tuning. There is also the deeper issue that detecting when restraint happens does not reveal what the experience is like. An infant in a high chair during animated face-to-face feeding lives a very different moment from an infant in the same chair while a caregiver washes dishes, and interpreting restraint’s developmental consequences will require pairing sensor data with audio, proximity tracking, and other contextual streams.

Even with those caveats, the implications are considerable. A validated, automated system for quantifying restraint opens questions that were previously unanswerable: what immediately precedes a caregiver’s decision to pick up or put down a baby, how restraint bouts are distributed across the day, whether prolonged individual bouts matter differently from the same total time fragmented into short ones, and how speech input, object interaction, and visual experience differ between restrained and unrestrained periods. The authors note that Tajik and US infants can show comparable overall restraint time yet different motor outcomes, suggesting that the temporal texture of restraint, not just its total, may be developmentally decisive. The team also envisions extending the approach to locomotion, steps, and falls, and to the underexplored outdoor environments where strollers, carriers, and car seats dominate. Sensor data, video annotations, and analysis code are publicly shared, lowering the barrier for other labs to build on the work. What began as a simple parental intuition, that babies spend much of their day constrained, can now be measured with second-by-second precision, and that measurement may reshape how developmental science understands the architecture of a baby’s everyday learning environment.

Subject of Research: Machine learning detection of infants' daily restraint from wearable inertial sensors

Article Title: Quantifying infants’ everyday restrained experiences in the home using wearable inertial sensors

Article References: Quantifying infants’ everyday restrained experiences in the home using wearable inertial sensors. (n.d.). https://doi.org/10.3758/s13428-026-03185-9

Image Credits: AI Generated

DOI: 10.3758/s13428-026-03185-9

Keywords: infant development, wearable sensors, machine learning, physical restraint, motor development, inertial measurement units, home observation, random forest, caregiving, Behavior Research Methods, developmental psychology, motion classification

The review followed a rigorous scoping methodology aligned with the PRISMA Extension for Scoping Reviews and the Joanna Briggs Institute Population–Concept–Context framework. The team searched three engineering and computer science databases, Scopus, the ACM Digital Library and IEEE Xplore, for publications between January 2000 and February 2025. In a striking methodological decision, the researchers deliberately excluded the word rhythm itself, along with synonyms such as timing, tempo, flow and cadence, from their search strings. Because rhythm is conceptually complex and used differently across disciplines, searching for it directly would have flooded the results with irrelevant work. Instead, they captured the broad landscape of augmentation and feedback technologies in sport, then examined whether the holistic, expert-understood concept of rhythm was actually being addressed. After screening 442 unique records from an initial pool of 529, the team distilled the literature down to 129 technology-focused studies.

The first headline finding is the era of the wearable sensor. Wearable devices were the primary technology in 48.8 percent of the included studies, and inertial measurement units, the tiny accelerometer and gyroscope packages that detect motion, appeared in 44.2 percent of the entire corpus. Their dominance reflects maturity, low cost and high ecological validity: unlike laboratory motion capture rigs, IMU-based wearables can record rhythm-relevant features such as step frequency, stance duration and inter-movement intervals directly in the field, on the track, in the pool or on the slopes. Examples in the corpus range from flexible sensor networks for sprint and jump biomechanics to smart dumbbells with embedded inertial units that evaluate strength training form in real time. For anyone hoping to quantify the temporal texture of movement, the sensor hardware problem is, in essence, already solved.

The second headline is a turn toward seeing rather than feeling the data. Computer vision systems accounted for 20.2 percent of primary technologies and extended reality systems for 22.5 percent, together representing more than 40 percent of the corpus. Visual capture technologies spanning ordinary RGB cameras, LiDAR and full motion capture appeared in one third of all hardware implementations. This suggests a field diverging into two camps: one that attaches discrete sensors to the body, and another that leverages optics and immersive environments where users quite literally inhabit their own performance data. The rise of extended reality, in particular, signals a shift beyond passive data logging toward experiential interfaces, exemplified by virtual reality training systems evaluated with national-level sitting volleyball players that produced measurable skill improvement.

Underpinning both camps is artificial intelligence. Explicit use of AI or machine learning to generate or adapt feedback appeared in 42 percent of the included studies. Rather than being the primary technology, machine learning frequently serves as the translation layer that converts raw sensor streams into meaningful, actionable cues. Examples range from wearable systems using deep learning to provide adaptive auditory feedback for weight training to computer vision pipelines that assess landing form in basketball. In principle, this is exactly the kind of computational machinery needed to move from simple real-time alerts toward the sophisticated, personalized, adaptive feedback systems that the authors argue elite sport actually requires.

Yet the feedback side of the equation reveals a striking monoculture. Visual feedback dominated the literature, appearing in 81 percent of studies, most commonly as on-screen graphs and numerical displays delivered via smartphone or tablet in 61 percent of the corpus. A representative system for hammer throw training gives coaches a real-time plot of the wire tension curve, elegant for a coach watching but demanding that an athlete look away from their own movement. The authors point out that visual attention during performance is a limited and critical resource, and studies on augmented reality workout systems have found that on-screen cues can impose additional cognitive load during dynamic movement. For an embodied skill like rhythm, which is often felt as much as seen, defaulting to screens may be actively counterproductive.

The untapped opportunity, the review argues, lies in the senses sport has largely ignored. Auditory feedback appeared in only 25 percent of studies and haptic feedback in a mere 11 percent. Their temporal affordances make them natural candidates for rhythm cueing: the ear is a timing organ, and the skin can feel a pulse without any attention being paid. The few studies that do explore these channels hint at real promise. Interactive sonification, in which movement is converted into continuous sound, has provided effective low-latency feedback on body roll in swimming, and electrical muscle stimulation has been used to directly correct running foot-strike patterns. Crucially, the auditory channel often remains unengaged during high-concentration physical exercise, leaving bandwidth free for guidance that refines rather than disrupts an athlete’s flow.

The review also documents a troubling gap between who builds these systems and who is supposed to benefit from them. Elite and professional athletes were the target population in only 26 percent of studies, and coaches in just 18.5 percent, while novices and amateurs, often university students, are heavily over-represented. The most common application domain was general fitness training, and 70 percent of studies targeted posture or technique form, compared with only 35 percent targeting timing or tempo, 20 percent targeting consistency or flow, and 12 percent targeting stride parameters. In other words, the literature measures discrete kinematic snapshots when rhythm is an integrated temporal quality. Elite performers possess refined, idiosyncratic motor programs; their challenge is micro-adjustment of an already optimized system, requiring higher-resolution feedback than novice learning studies typically provide.

The proof problem compounds the population problem. Although 81 percent of systems were empirically evaluated, the primary goal of most evaluations was system usability or feasibility, in 45 percent of studies, while only 22.5 percent measured performance improvement and 8.5 percent measured skill acquisition. Nearly one in five studies, 19.4 percent, reported no formal evaluation at all. Perhaps most tellingly, only 26.5 percent of studies explicitly measured rhythm or a direct synonym as an outcome variable. Systems for tennis swing classification, for example, celebrated machine learning model accuracy rather than any change in player skill. The result is a literature full of working prototypes with largely unevaluated training impact.

The authors distill these observations into four challenges, concerning what is sensed, who is studied, how feedback is delivered and what proof exists, and offer them as exploratory, hypothesis-generating directions rather than evidence-based prescriptions. Their near-term proposal is compelling in its simplicity: couple the already mature wearable inertial sensing platforms with continuous, low-latency, non-visual feedback such as movement sonification or complementary haptic cues, and design evaluations that separate technical validation, user experience and genuine performance outcomes as distinct levels of evidence, a trajectory they map onto the Technology Readiness Level framework from proof-of-concept to proof-of-impact. The review has limitations, including its restriction to English-language, technology-indexed literature across three engineering databases, so field-level conclusions about rhythm training await a cross-disciplinary synthesis. But the message to sports engineers is clear: the sensors are ready, the audio and touch channels are wide open, and the athletes at the top are still waiting for technology that speaks their temporal language.

Subject of Research: Human augmentation technologies and feedback modalities for training rhythm in elite sport

Article Title: Augmenting rhythm: a scoping review of technologies and feedback modalities in elite sport

Article References: Hossain, M. T., Lindeman, R. W., Ingram, M., & Lukosch, S. (2026). Augmenting rhythm: a scoping review of technologies and feedback modalities in elite sport. Sports Engineering, 29(2), Article 32. https://doi.org/10.1007/s12283-026-00562-7

Image Credits: AI Generated

DOI: 10.1007/s12283-026-00562-7

Keywords: rhythm, elite sport, wearable sensors, inertial measurement units, real-time feedback, sonification, haptic feedback, extended reality, computer vision, artificial intelligence, sports engineering, scoping review

The research, led by Giulia Casu and Massimiliano Pau of the University of Cagliari together with colleagues at the Regional Multiple Sclerosis Center of Sardinia, set out to answer a deceptively simple question: what actually happens to the quality of walking as fatigue accumulates over six minutes of continuous walking? While clinicians routinely use the six-minute walk test to gauge exercise capacity, the fine-grained temporal and spatial coordination of gait during this endurance challenge had been explored only through a single metric, leaving much of the picture unexamined.

To capture that picture, the team recruited 65 people with MS who could walk unaided for at least 200 meters, along with 43 age- and sex-matched unaffected individuals. Each participant wore eight miniaturized inertial measurement units sampling at 200 hertz, positioned on the lower spine, sacrum, thighs, tibiae, and forefeet. These wearable sensors tracked accelerations and angular velocities throughout the walk, allowing the researchers to reconstruct heel-strike events and joint-angle trajectories for hip and knee flexion-extension as well as ankle dorsi-plantarflexion. The 180-degree turns between consecutive 30-meter walking segments were carefully identified from pelvic yaw signals and excluded, ensuring that only steady-state straight-line walking entered the analysis.

On average, people with MS covered 477.95 meters over six minutes, significantly less than the 626.94 meters managed by unaffected participants, and both groups slowed progressively. But the most revealing findings came from the coordination metrics. The researchers computed the Phase Coordination Index, or PCI, a temporal measure that captures how accurately and consistently the left and right legs maintain ideal anti-phased stepping at 180 degrees. Healthy individuals typically score between 2 and 4 percent; lower values indicate better bilateral coordination. In this study, the MS group showed significantly higher PCI values than controls across the entire test, with a significant positive linear trend across the six minutes for both groups.

Strikingly, by the end of the test PCI had risen by about 53 percent in unaffected individuals but only 32 percent in people with MS. This counterintuitive pattern — the group with the disease deteriorating less — suggests a ceiling effect. People with MS started from a baseline of already-impaired coordination, constrained by reduced interhemispheric communication between the primary motor cortices, microstructural spinal cord damage, and degraded integrity of corticospinal, cerebellar, and brainstem pathways. With less adaptive capacity remaining, there was simply less room for further fatigue-driven decline. Their consistently slower walking speed may also have limited the range of available coordination strategies, restricting how much further measurable change was possible.

The study’s key methodological innovation was pairing this temporal metric with spatial measures derived from cyclograms — angle-angle diagrams that plot one joint’s motion against another across the gait cycle. From these closed curves, the team extracted geometric features including area, orientation, perimeter, and a dimensionless shape ratio. A perfectly symmetrical gait would collapse the inter-limb cyclogram onto a 45-degree line with zero area; larger areas and orientations deviating from that line signal asymmetry. For inter-joint analysis, hip-knee and knee-ankle cyclograms revealed how conjointly the adjacent joints modulate their movements through each stride.

The spatial data confirmed significant group differences at nearly every point measured. People with MS showed larger inter-limb cyclogram areas, orientations, and trend symmetry values at the hip and knee, indicating reduced symmetry, while their inter-joint cyclogram areas and perimeters for the hip-knee and knee-ankle couples were consistently reduced, reflecting a diminished dynamic range of motion and a reduced ability to modulate joint interactions across the phases of the gait cycle. Correlation analysis added clinical weight: within the MS group, PCI at minute one showed a moderate positive correlation with the MSWS-12 patient-reported walking scale, while hip-knee and knee-ankle cyclogram areas showed moderate-to-large negative correlations, linking laboratory measures of coordination directly to patients’ own perception of walking disability.

The temporal evolution of inter-joint coordination proved particularly instructive at the knee-ankle couple. In unaffected individuals, cyclogram area dropped significantly from the second minute onward; in people with MS, the reduction began a minute later. By the end of the test, however, the overall magnitude of change was similar in both groups. The authors interpret this as a possible convergence toward a shared, fatigue-driven motor strategy that prioritizes stability over flexibility at the distal joints. The ankle, central to propulsion and foot clearance, appears especially vulnerable to fatigue-related impairments such as reduced plantarflexion at push-off, which tightens its coupling with the knee. In contrast, the knee joint alone showed significant fatigue-related worsening only in unaffected individuals, possibly because compensatory strategies common in MS — increased reliance on proximal joints and muscular co-contraction to enhance stability — mask further deterioration at that level.

The authors acknowledge limitations, including a group size too small for stratified analyses of disability level, sex, or fatigability, and the fact that gait speed was not experimentally controlled, which means coordination changes over time cannot be fully disentangled from concurrent speed changes. The repeated turns demanded by the six-minute protocol may also have added sensorimotor load not directly quantified. Even so, the consistency of the baseline coordination deficits across both temporal and spatial domains, and their agreement quantitatively with values reported in earlier independent studies, underscores the robustness of the approach.

The clinical implications reach well beyond the laboratory. Because baseline coordination deficits and fatigue-related changes followed dissociable patterns, rehabilitation strategies for MS may need to target coordination quality and adaptability explicitly, not merely walking endurance. The sensitivity of cyclogram-based measures to knee-ankle coupling suggests that distal coordination is a promising target for interventions aiming to improve propulsion efficiency and stability during prolonged walking. And because temporal and spatial metrics captured joint- and domain-specific behaviors that a single index missed, the multimodal framework developed here offers clinicians a more complete toolkit for tailoring rehabilitation, monitoring therapeutic response, and identifying patients who rely on compensation rather than genuine adaptive motor control — a step toward better walking performance, reduced fatigability, and greater independence in daily life for people living with MS.

Subject of Research: Temporal and spatial gait coordination changes during the six-minute walk test in people with multiple sclerosis, measured with wearable inertial sensors.

Article Title: Changes in temporal and spatial gait coordination during the six‑minute walk test in people with multiple sclerosis

Article References: Casu, G., Lostia di Santa Sofia, E., Leban, B., Inglese, C., Coghe, G., Frau, J., Cocco, E., & Pau, M. (2026). Changes in temporal and spatial gait coordination during the six‑minute walk test in people with multiple sclerosis. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-026-03678-x

Image Credits: AI Generated

DOI: 10.1007/s11517-026-03678-x

Keywords: multiple sclerosis, gait coordination, six-minute walk test, phase coordination index, cyclograms, inertial measurement units, gait analysis, fatigue, rehabilitation, biomechanics, mobility, neurology