Torso Locations Research Articles

Personalized Gym Trainer Using Mediapipe

Abstract: With the growing demand in personalised fitness experiences, this paper takes a fresh look at home workouts by creating a Personalised Gym Trainer with the Mediapipe library. This sophisticated device combines pose detection technology with voice assistance to provide users with real-time feedback and personalised instruction while exercising. The system identifies various exercises accurately and leverages torso point detection for greater precision by leveraging the capabilities of the Mediapipe library and OpenCV for camera tasks. Individual Python modules for certain workouts such as pull-ups and bench press are written to support a flexible and scalable solution. Python, Git, GitHub, and Jenkins are among the tools and technologies used in the process. Furthermore, the Findpose library is used to calculate angles between torso locations, resulting in Providing a quantitative assessment of proper posture. The suggested Personalised Gym Trainer is a promising leap in home fitness solutions, integrating cutting-edge pose detection with voice assistance for an interactive and personalised workout experience.

Vertical contact forces affect vibration perception in human hairy skin.

Skin is the largest organ of the human body and fulfills many important functions, like detecting mechanical stimuli. Skin can be divided into glabrous (non-hairy) and hairy skin. These two skin types differ with regard to their mechanical properties and in the distribution of mechanoreceptors. Although many investigations focus on glabrous skin, hairy skin still plays a fundamental role in various activities, e.g., with regard to the perception of pleasantness or for developing wearable vibrotactile devices for pattern recognition in persons with disabilities. Unfortunately, investigations on influencing factors, like vertical contactor force, are scarce for hairy skin. Similarly, it would also be interesting to investigate whether regional vibratory sensitivity differences are present across the human torso. Hence, this study investigated the effects of vertical contactor forces and different anatomical locations on vibration perception. Four anatomical torso regions were studied. Based on findings in glabrous skin, we generally hypothesized improved vibration perception with increasing contactor forces and regional sensitivity differences between the anatomical locations. Forty young and healthy individuals participated (23.0 ± 2.0 yrs), and vibration perception thresholds (VPTs) were determined at 30 Hz for three vertical force levels (0.6, 2.4, and 4.8 N) at four torso locations (sternum, deltoid/shoulder, lower back, middle lateral torso side). Higher contactor forces resulted in lower VPTs corresponding to improved vibration perception, regardless of anatomical location. In addition, the sternum region was more sensitive than the remaining three regions, regardless of force level. The reasons for these findings may be a varying number and activation pattern of afferents activated under the different conditions. The findings of this study complement the understanding of vibrotactile sensitivity in hairy skin and may offer implications when developing vibrotactile devices or clothing/textiles, for example.

The Effect of Human Tissue on Field Strength Measurements In Vivo Using a Resonant UHF Cavity-Backed Slot Antenna.

An inward-looking wearable antenna can be used for radio communications with internal transceivers in vivo. The radio transmissions are recorded using an array of electric field sensors on the skin. This paper reports the effect of living tissue on a small cavity-backed slot antenna pressed onto soft tissue of the human torso at 2.09 GHz. In-vivo measurements were made on the skin surface at 13 torso locations using eight participants (age range, 22-68 years old), with body mass indexranging between 20.3 and 31.6 kg/m2 . Ultrasound imaging was used to determine the skin and fat thickness at every measurement location. The variation in the antenna input impedance measurements demonstrated that the human tissues (fat and muscle) affect the antenna impedance but the mismatch creates field strength measurement errors of less than 2 dB. Fat thickness in the range of 3-30 mm can slightly degrade the performance of these wearable antennas. These effects can be partly mitigated by selective location and antenna retuning to improve transceiver communications. Bioelectromagnetics. 2021. © 2021 Bioelectromagnetics Society.

Measurement of lower-limb asymmetry in professional rugby league: a technical note describing the use of inertial measurement units

BackgroundQuantifying lower-limb load and asymmetry during team sport match-play may be important for injury prevention and understanding performance. However, current analysis methods of lower-limb symmetry during match-play employ wearable microtechnology that may not be best suited to the task. A popular microtechnology is global positioning systems (GPS), which are torso worn. The torso location, and the summary workload measures calculated by GPS are not suited to the calculation of lower-limb load. Instead, research grade accelerometers placed directly on the lower-limb may provide better load information than GPS. This study proposes a new technique to quantify external mechanical load, and lower-limb asymmetry during on-field team sport play using inertial measurement units.MethodsFour professional rugby league players (Age: 23.4 ± 3.1 years; Height: 1.89 ± 0.05 m; Mass: 107.0 ± 12.9 kg) wore two accelerometers, one attached to each foot by the boot laces, during match simulations. Custom Matlab (R2017b, The Mathworks Inc, Natick, MA) code was used to calculate total time, area under the curve (AUC), and percentage of time (%Time) spent in seven acceleration categories (negative to very high, <0 g to >16 g), as well as minimum and maximum acceleration during match simulations. Lower-limb AUC and %Time asymmetry was calculated using the Symmetry Angle Equation, which does not require normalization to a reference leg.ResultsThe range of accelerations experienced across all participants on the left and right sides were 15.68–17.53 g, and 16.18–17.69 g, respectively. Clinically significant asymmetry in AUC and %Time was observed for all but one participant, and only in negative (<0 g) and very high accelerations (>16 g). Clinically significant AUC differences in very high accelerations ranged from 19.10%–26.71%. Clinically significant %Time differences in negative accelerations ranged from 12.65%–25.14%, and in very high accelerations from 18.59%–25.30%. All participants experienced the most AUC at very low accelerations (2–4 g), and the least AUC at very high accelerations (165.00–194.00 AU vs. 0.32–3.59 AU). The %Time results indicated that all participants spent the majority of match-play (73.82–92.06%) in extremely low (0–2 g) to low (4–6 g) acceleration intensities, and the least %Time in very high accelerations (0.01%–0.05%).DiscussionA wearable located on the footwear to measure lower-limb load and asymmetry is feasible to use during rugby league match-play. The location of the sensor on the boot is suited to minimize injury risk occurring from impact to the sensor. This technique is able to quantify external mechanical load and detect inter limb asymmetries during match-play at the source of impact and loading, and is therefore likely to be better than current torso based methods. The results of this study may assist in preparing athletes for match-play, and in preventing injury.

Gait Analysis of Pedestrians with the Aim of Detecting Disabled People

Gait classification is an effective and non-intrusive method for human identification and it has received significant attention in the recent years due to its applications in visual surveillance and monitoring systems. We analyse gait signatures using spatio-temporal motion characteristics of a person to answer the question ``is there a discriminating feature in the gait signal that can help to categorise a disabled person from healthy''. The procedure has three steps: detection of a pedestrian using YOLO followed by the silhouette extraction using the Gaussian Mixture Model (GMM). Finally, skeletonization from the silhouette image to estimate head and torso locations and their angles with the vertical axis. Furthermore, velocity and acceleration signals were recorded to look for accelerating behaviour of person walking with a limp. Manual segmentations shows that the gait signal has information about unusual walking patterns but existing pedestrian detectors lack accuracy in extracting an accurate gait signal due to localization errors.

RETRACTED: Gait Analysis of Pedestrians with the Aim of Detecting Disabled People

Gait classification is an effective and non-intrusive method for human identification and it has received significant attention in the recent years due to its applications in visual surveillance and monitoring systems. In this project, we analysed gait signatures using spatio-temporal motion characteristics of a person to answer the question ``is there a discriminating feature in gait signal that can help to categorise disable person from healthy''. The procedure has three steps. detection of a pedestrian using YOLO followed by the silhouette extraction using the Gaussian Mixture Model (GMM). Finally, skeletonization from the silhouette image to estimate head and torso locations and their angles with the vertical axis. Furthermore, velocity and acceleration signals were recorded to look for accelerating behaviour of person walking with a limp.

Directional postural responses induced by vibrotactile stimulations applied to the torso

It has been shown that torso-based vibrotactile feedback significantly reduces postural sway in balance-compromised adults during quiet standing and in response to perturbations. This study aimed to determine whether vibrotactile stimulations applied to different torso locations induced directional postural responses and whether torso cutaneous information contributes to body representation. Eleven healthy young adults equipped with an inertial measurement unit (IMU) placed on the torso were asked to maintain an upright posture with closed eyes. Six vibrators (tactors) were placed on the torso in contact with the skin over the left and right external oblique, internal oblique, and erector spinae muscles at the L4/L5 level. Each tactor was randomly activated four times per location at a frequency of 250Hz for a period of 5s. The IMU results indicated that vibration applied individually over the internal oblique and erector spinae muscles induced a postural shift of about one degree oriented in the direction of the stimulation, while simultaneous activation of all tactors and activation of tactors over external oblique muscles produced insignificant postural effects. The root mean square of the sway signal was significantly higher during vibration than before or after. However, the center of pressure displacement, measured by a force plate, was uninfluenced by any vibration. These results suggest a multi-joint postural response including a torso inclination associated with vibration-induced changes in cutaneous information. The directional aspect of vibration-induced postural shifts suggests that cutaneous information from the stimulated areas contributes to proprioception and upper body spatial representation.

Multiview human pose estimation with unconstrained motions

We present a system for human pose estimation by using a single frame and without making assumptions on temporal coherence. The system uses 3D voxel data reconstructed from multiple synchronized video streams as input, and computes, for each frame, a skeleton model which best fits the body pose. This system adopts a hierarchical approach where the head and torso locations are found first based on template fitting with their specific shapes and dimensions. It is followed by a limb detection procedure that estimates the pose parameters of four limbs. However, a problem generally faced with skeleton models is the means to find adequate measurements to fit the model. In this paper, voxel data, together with two novel local shape features, are used for this purpose. Experiments show that this system is robust to several perturbations associated with the input data, such as voxel reconstruction errors and complex poses with self-contact, and also allows unconstrained motions, such as fast or unpredictable movements.

Reconstruction of the pelvic floor and perineum with human acellular dermal matrix and thigh flaps following pelvic exenteration

Background Patients who undergo pelvic floor resection as treatment for recurrent cancer following radiation therapy have increased rates of complications, particularly if permanent prosthetic mesh is used for reconstruction. Human acellular dermal matrix (HADM), commonly used for reconstruction in other torso locations, is associated with lower rates of complications (including infection, adhesions and cutaneous exposure) than synthetic mesh. We describe an effective technique to reconstruct the pelvic floor and perineum with HADM and thigh-based flaps following pelvic exenteration and radical vulvectomy. Case A 75-year-old woman underwent radical resection of the pelvic floor and perineum to treat recurrent vulvar squamous cell carcinoma and osteoradionecrosis. The pelvic floor and perineal soft tissue defect were reconstructed with HADM (AlloDerm; LifeCell Corporation, Branchburg, NJ) and bilateral, thigh-based tissue flaps, respectively. Despite a large resection, previous irradiation therapy and bacterial contamination the wounds healed without complications. Conclusion Reconstruction of pelvic floor defects using HADM is an option when wound conditions are unfavorable for the use of permanent prosthetic meshes.

The Relation between Visceral Fat Measurement and Torso Level—Is One Level Better Than Another?

Intraabdominal fat (IAF) area is often measured indirectly in epidemiologic studies. The authors recruited 147 participants from the second examination (1990-1992) of the Atherosclerosis Risk in Communities Study to examine IAF area and determine whether there were differences in IAF area and distribution by location. Magnetic resonance imaging was used to image four 10-mm slices between the second and fourth lumbar vertebrae by an inverse recovery method, and IAF was calculated from each image. The authors constructed gender-specific mixed models with IAF area as the outcome and the location of imaging along the torso as the independent variable, using random intercepts to account for between-person variation in IAF area. The torso location of IAF measurement was a significant predictor of IAF area in both men (p = 0.02) and women (p < 0.0001) after adjustment for body mass index. A significant positive interaction between age and location was seen in men, with increasing IAF area moving down the torso with older ages. Using magnetic resonance imaging, location along the torso yields different IAF areas and distributions independently of body mass index in both genders, with measurement at the second lumbar vertebra (slightly above the umbilicus) capturing the largest amount of IAF. Studies that attempt to link IAF with cardiovascular disease risk factors should consider measurement location to accurately capture the association.

Presenting directions with a vibrotactile torso display

Vibrotactile displays covering the torso present spatial information in an intuitive way since the stimuli are directly mapped to the body coordinates; left is left, front is front, etc. The present study investigated the direction in the horizontal plane to which a specific torso location is mapped using a 15 tactor linear display. Participants indicated the observed external direction of a localized vibration by positioning a remotely controlled cursor. The results show that the observed direction is toward the midsagittal plane (i.e. the navel or the spine) compared to the tactor direction. This bias is consistent over the observers tested, and up to 10° for the oblique directions. Inspection of the response patterns revealed that observers did not use the body midaxis as the origin for the observed direction, but used two spatially separated internal reference points, one for each body half. Therefore, the use of a torso-related transfer function in applications that require a high level of precision is suggested. The variability of the responses also depends on the tactor direction. It is higher for the left – right direction and lower for the fore – aft direction, which may be caused by the spine and navel acting as anchor points. The present results are relevant for tactile navigation displays for visually handicapped individuals or for jet pilots.

16-Lead ECG changes with coronary angioplasty

Percutaneous transluminal coronary angioplasty (PTCA) occlusion in five individual coronary artery distributions produced significant ST elevation ("current of injury") in 48/50 PTCAs in 46 patients. Four patients had PTCA of two separate coronary arteries. Two patients had no significant ischemic ST changes in the 16 simultaneous lead ECG and no chest pain with PTCA. The six limb leads were recorded from Mason-Likar locations modified by moving them centrally on the anterior torso; the V leads were recorded in standard locations, except V1 was moved to V3R. Four extra leads were placed as follows: (1) left axilla, (2) left subcostal margin, (3) V8, and (4) midback at the level of V4-V8. The left axillary and back leads discriminated diagonal and left circumflex (LCX) PTCAs from the others and from each other. V6 showed ST elevation in all LCX PTCAs and in only 10% of left anterior descending occlusions. V3R showed ST elevation in 82% of right coronary PTCAs. In 48/50 (96%) of PTCA occlusions the ST elevation was localized to the torso locations defined in Forward Model Simulations as specific for the arterial perfusion bed involved. These data strongly support the hypothesis that additional resolution and sensitivity to ischemic change is to be expected with a broader array of ECG leads.

Electrocardiographic effects of myocardial ischemia induced by atrial pacing in dogs with coronary stenosis. II. Comparison of S-T segment and QRS changes in dogs with left circumflex arterial narrowing

The effects of tachycardia on the QRS complex in the absence and in the presence of coronary arterial constriction are of possible clinical significance. To study the spatial and temporal distributions of these effects, a canine model simulating progressive coronary occlusion was studied. Tachycardia was induced by atrial pacing, coronary narrowing was produced by implantation of an ameroid constrictor about the left circumflex coronary artery and electrocardiographic effects were studied by isopotential mapping techniques at rates of up to 250 beats per minute, one to three weeks after constrictor implant. In five dogs. pacing without coronary constriction demonstrated the spatial and temporal dependence of QRS changes; tachycardia-induced changes varied from one torso site to another at any instant, and from one moment to another at any one torso location. Patterns in ten dogs with constrictors were dependent upon time after ameroid placement. Resting QRS and S-T segment distributions remained unchanged. One week after surgery, QRS and S-T segment patterns during pacing were as in controls. Two weeks after surgery, QRS patterns were as in controls but flat S-T segment depression resulted at high rates. After three weeks, QRS patterns during pacing were abnormal and S-T segment depression was observed. At each of the later times, the QRS response was similar at rates which did and which did not cause S-T depression. Thus, the effects of tachycardia on excitation and recovery forces in the presence of restricted coronary flow are independent of one another. This suggests that the two responses to ischemia may be due to different basic mechanisms and, hence, that their detection may reflect different myocardial abnormalities.