Clustering and classification of regional peak plantar pressures of diabetic feet

Machine learning for lumbar and pelvis kinematics clustering

Clustering algorithms such as k-means and agglomerative hierarchical clustering (HCA) may provide a unique opportunity to analyze time-series kinematic data. Here we present an approach for determining number of clusters and which clustering algorithm to use on time-series lumbar and pelvis kinematic data. Cluster evaluation measures such as silhouette coefficient, elbow method, Dunn Index, and gap statistic were used to evaluate the quality of decision making. The result show that multiple clustering evaluation methods should be used to determine the ideal number of clusters and algorithm suitable for clustering time-series data for each dataset being analyzed.

In-shoe plantar pressure measurement technologies for the diabetic foot: A systematic review

Introduction

Loss of cutaneous protective sensation and high plantar pressures increase the risk for diabetic foot patients. Trauma and ulceration are imminent threats, making assessment and monitoring essential. This systematic review aims to identify systems and technologies for measuring in-shoe plantar pressures, focusing on the at-risk diabetic foot population.

Methods

A systematic search was conducted across four electronic databases (Scopus, Web of Science, PubMed, Oxford Journals) using PRISMA methodology, covering articles published in English from 1979 to 2024. Only studies addressing systems or sensors exclusively measuring plantar pressures inside the shoe were included.

Results

A total of 87 studies using commercially available devices and 45 articles proposing new systems or sensors were reviewed. The prevailing market offerings consist mainly of instrumented insoles. Emerging technologies under development often feature configurations with four, six or eight resistive sensors strategically placed within removable insoles. Despite some variability due to the inherent heterogeneity of human gait, these devices assess plantar pressure, although they present significant differences between them in measurement results. Individuals with diabetic foot conditions appears exhibit elevated plantar pressures, with reported peak pressures reaching approximately 1000 kPa. The results also showed significant differences between the diabetic and non-diabetic groups.

Conclusion

Instrumented insoles, particularly those incorporating resistive sensor technology, dominate the field. Systems employing eight sensors at critical locations represent a pragmatic approach, although market options extend to systems with up to 960 sensors. Differences between devices can be a critical factor in measurement and highlights the importance of individualized patient assessment using consistent measurement devices.

Identification of movement phenotypes from occupational gesture kinematics: Advancing individual ergonomic exposure classification and personalized training

The identification of personalized preventive strategies plays a major role in contrasting the occurrence of work-related musculoskeletal disorders. This requires the identification of distinct movement patterns within large samples and the attribution of a proper risk level to each identified movement phenotype. We assessed the feasibility of this approach by exploiting wearable inertial measurement units to estimate the whole-body kinematics of 43 healthy participants performing 18 reach-to-manipulate movements, which differed based on the object's position in the space and the type of manipulation required. Through unsupervised clustering, we identified multiple movement phenotypes graded by ergonomic performance. Furthermore, we determined which joints mostly contributed to instantiating the ergonomic differences across clusters, emphasizing the importance of monitoring this aspect during occupational gestures. Overall, our analysis suggests that movement phenotypes can be identified within occupational motor repertoires. Assigning individual performance to specific phenotypes has the potential to inform the development of more effective and tailored interventions.

Plantar pressure and falling risk in older individuals: a cross-sectional study

BACKGROUND

Falls are commonplace among elderly people. It is urgent to prevent falls. Previous studies have confirmed that there is a difference in plantar pressure between falls and non-falls in elderly people, but the relationship between fall risk and foot pressure has not been studied. In this study, the differences in dynamic plantar pressure between elderly people with high and low fall risk were preliminarily discussed, and the characteristic parameters of plantar pressure were determined.

METHODS

Twenty four high-fall-risk elderly individuals (HR) and 24 low-fall-risk elderly individuals (LR) were selected using the Berg Balance Scale 40 score. They wore wearable foot pressure devices to walk along a 20-m-long corridor. The peak pressure (PP), pressure time integral (PTI), pressure gradient (maximum pressure gradient (MaxPG), minimum pressure gradient (MinPG), full width at half maximum (FWHM)) and average pressure (AP) of their feet were measured for inter-group and intra-group analysis.

RESULTS

The foot pressure difference comparing the high fall risk with low fall risk groups was manifested in PP and MaxPG, concentrated in the midfoot and heel (p < 0.05), while the only time parameter, FWHM, was manifested in the whole foot (p < 0.05). The differences between the left and right foot were reflected in all parameters. The differences between the left and right foot in LR were mainly reflected in the heel (p < 0.05), while it in the HR was mainly reflected in the forefoot (p < 0.05).

CONCLUSIONS

The differences comparing the high fall risk with low fall risk groups were mostly reflected in the midfoot and heel. The HR may have been more cautious when landing. In the intra-group comparison, the difference between the right and left foot of the LR was mainly reflected during heel striking, while it was mainly reflected during pedalling in the HR. The sensitivity of PP, PTI and AP was lower and the newly introduced pressure gradient could better reflect the difference in foot pressure between the two groups. The pressure gradient can be used as a new foot pressure parameter in scientific research.

Footwear and insole design parameters to prevent occurrence and recurrence of neuropathic plantar forefoot ulcers in patients with diabetes: a series of N-of-1 trial study protocol

BACKGROUND

Foot complications occur in conjunction with poorly controlled diabetes. Plantar forefoot ulceration contributes to partial amputation in unstable diabetics, and the risk increases with concomitant neuropathy. Reducing peak plantar forefoot pressure reduces ulcer occurrence and recurrence. Footwear and insoles are used to offload the neuropathic foot, but the success of offloading is dependent on patient adherence. This study aims to determine which design and modification features of footwear and insoles improve forefoot plantar pressure offloading and adherence in people with diabetes and neuropathy.

METHODS

This study, involving a series of N-of-1 trials, included 21 participants who had a history of neuropathic plantar forefoot ulcers. Participants were recruited from two public hospitals and one private podiatry clinic in Sydney, New South Wales, Australia. This trial is non-randomised and unblinded. Participants will be recruited from three sites, including two high-risk foot services and a private podiatry clinic in Sydney, Australia. Mobilemat™ and F-Scan® plantar pressure mapping systems by TekScan® (Boston, USA) will be used to measure barefoot and in-shoe plantar pressures. Participants' self-reports will be used to quantify the wearing period over a certain period of between 2 and 4 weeks during the trial. Participant preference toward footwear, insole design and quality-of-life-related information will be collected and analysed. The descriptive and inferential statistical analyses will be performed using IBM SPSS Statistics (version 27). And the software NVivo (version 12) will be utilised for the qualitative data analysis.

DISCUSSION

This is the first trial assessing footwear and insole interventions in people with diabetes by using a series of N-of-1 trials. Reporting self-declared wearing periods and participants' preferences on footwear style and aesthetics are the important approaches for this trial. Patient-centric device designs are the key to therapeutic outcomes, and this study is designed with that strategy in mind.

TRIAL REGISTRATION

Australian New Zealand Clinical Trials Registry (ANZCTR) ACTRN12620000699965p. Registered on June 23, 2020.

Effects of textile-fabricated insole on foot skin temperature and humidity for enhancing footwear thermal comfort

Traditional insole materials which trap heat and moisture inside footwear cause discomfort to the wearer. Here, a novel textile-fabricated insole material with a 3D structure that offers good porosity and breathability for improving the footwear microclimate is proposed. Changes in foot skin temperature and humidity when wearing the textile-fabricated insole throughout treadmill walking are collected from 21 female subjects (age: 25.5 ± 4.5) and compared with traditional and 3D printed insoles. Subjective assessment of their perceived thermal comfort with various insole conditions is also conducted. In comparison to polyurethane, 3D printed thermoplastic polyurethane and leather insoles, textile-fabricated insoles show no significant changes in foot skin temperature. Nevertheless, a significant reduction of the relative humidity of the skin of the sole (3.21%) and heel (24.41%) is found. The findings are a valuable reference for the fabrication of insoles with higher wear comfort.

New Distinct Component Patterns for Plantar Pressure Variables by Using Principal Component Analysis

BACKGROUND

It is important to determine the plantar pressure distribution of schoolchildren by applying static and dynamic foot analyses using a pedobarography device. However, it is difficult to obtain clear interpretations from results that can be explained by a large number of plantar pressure variables. The aim of this study was to use principal component analysis (PCA) to predict the main components for reducing the size of big data sets, provide a practical overview, and minimize information loss on the subject of plantar pressure assessment in youths.

METHODS

In total, 112 schoolchildren were included in the study (mean ± SD: age, 10.58 ± 1.27 years; body mass index, 18.86 ± 4.33). During the research, a pedobarography device was used to obtain plantar pressure data. Each foot was divided into six anatomical regions and evaluated. Global and regional plantar pressure distributions, load and surface areas, pressure-time integrals, weight ratios, and geometric foot properties were calculated.

RESULTS

The PCA yielded ten principal components that together account for 81.88% of the variation in the data set and represent new and distinct patterns. Thus, 137 variables affecting the subject were reduced to ten components.

CONCLUSIONS

The numerous variables that affect static and dynamic plantar pressure distributions can be reduced to ten components by PCA, making the research results more concise and understandable.

Recent Machine Learning Progress in Lower Limb Running Biomechanics With Wearable Technology: A Systematic Review

With the emergence of wearable technology and machine learning approaches, gait monitoring in real-time is attracting interest from the sports biomechanics community. This study presents a systematic review of machine learning approaches in running biomechanics using wearable sensors. Electronic databases were retrieved in PubMed, Web of Science, SPORTDiscus, Scopus, IEEE Xplore, and ScienceDirect. A total of 4,068 articles were identified via electronic databases. Twenty-four articles that met the eligibility criteria after article screening were included in this systematic review. The range of quality scores of the included studies is from 0.78 to 1.00, with 40% of articles recruiting participant numbers between 20 and 50. The number of inertial measurement unit (IMU) placed on the lower limbs varied from 1 to 5, mainly in the pelvis, thigh, distal tibia, and foot. Deep learning algorithms occupied 57% of total machine learning approaches. Convolutional neural networks (CNN) were the most frequently used deep learning algorithm. However, the validation process for machine learning models was lacking in some studies and should be given more attention in future research. The deep learning model combining multiple CNN and recurrent neural networks (RNN) was observed to extract different running features from the wearable sensors and presents a growing trend in running biomechanics.

Automatic Classification of Barefoot and Shod Populations Based on the Foot Metrics and Plantar Pressure Patterns

The human being’s locomotion under the barefoot condition enables normal foot function and lower limb biomechanical performance from a biological evolution perspective. No study has demonstrated the specific differences between habitually barefoot and shod cohorts based on foot morphology and dynamic plantar pressure during walking and running. The present study aimed to assess and classify foot metrics and dynamic plantar pressure patterns of barefoot and shod people via machine learning algorithms. One hundred and forty-six age-matched barefoot (n = 78) and shod (n = 68) participants were recruited for this study. Gaussian Naïve Bayes were selected to identify foot morphology differences between unshod and shod cohorts. The support vector machine (SVM) classifiers based on the principal component analysis (PCA) feature extraction and recursive feature elimination (RFE) feature selection methods were utilized to separate and classify the barefoot and shod populations via walking and running plantar pressure parameters. Peak pressure in the M1-M5 regions during running was significantly higher for the shod participants, increasing 34.8, 37.3, 29.2, 31.7, and 40.1%, respectively. The test accuracy of the Gaussian Naïve Bayes model achieved an accuracy of 93%. The mean 10-fold cross-validation scores were 0.98 and 0.96 for the RFE- and PCA-based SVM models, and both feature extract-based and feature select-based SVM models achieved an accuracy of 95%. The foot shape, especially the forefoot region, was shown to be a valuable classifier of shod and unshod groups. Dynamic pressure patterns during running contribute most to the identification of the two cohorts, especially the forefoot region.

A Wearable Gait Analysis System Used in Type 2 Diabetes Mellitus Patients: A Case-Control Study

BACKGROUND

Previous studies have shown that the gait of patients with type-2 diabetes mellitus is abnormal compared with the healthy group. Currently, a three-dimensional motion analyzer system is commonly used for gait analysis. However, it is challenging to collect data and use in clinical study due to extensive experimental conditions and high price. In this study, we used a wearable gait analysis system (Gaitboter) to investigate the spatial and temporal parameters, and kinematic data of gait in diabetic patients, especially those with peripheral neuropathy. The aim of the study is to evaluate the wearable gait analysis system in diabetic study.

MATERIALS AND METHODS

We conducted a case-control study to analyze the gait of type 2 diabetes mellitus. Gaitboter was used to detect and collect gait data in the ward of Beijing Chao-yang Hospital, Capital Medical University from June 2018 to October 2018. We collected the gait data of participants (N= 146; 73 patients with type 2 diabetes, 16 with peripheral neuropathy and 57 without peripheral neuropathy, and 73 matched controls). The gait data (stance phase, swing phase, double-foot stance phase, single-foot stance phase, walking cadence, stride length, walking speed, off-ground angle, landing angle, maximum swing angle, minimum swing angle, and foot progression angle) in diabetic patients were recorded and compared with controls. SPSS 22.0 statistical software was used to analyzed the gait parameter data.

RESULTS

We found that the landing angle and the maximum swing angle of diabetes patients with or without peripheral neuropathy were significantly less than those of the control group (P < 0.05). The walking speed of diabetes patients with peripheral neuropathy is significantly less than those of the control group (P < 0.05).

CONCLUSION

This study confirms that the wearable gait analysis system (Gaitboter) is an ideal system to identify abnormal gait in patients with type 2 diabetes and provides a new device and method for diabetes-related gait research.

Predicting ground reaction and tibiotalar contact forces after total ankle arthroplasty during walking

The motion capture and force plates data are essential inputs for musculoskeletal multibody dynamics models to predict in vivo tibiotalar contact forces. However, it could be almost impossible to obtain valid force plates data in old patients undergoing total ankle arthroplasty under some circumstances, such as smaller gait strides and inconsistent walking speeds during gait analysis. To remove the dependence of force plates, this study has established a patient-specific musculoskeletal multibody dynamics model with total ankle arthroplasty by combining a foot-ground contact model based on elastic contact elements. And the established model could predict ground reaction forces, ground reaction moments and tibiotalar contact forces simultaneously. Three patients' motion capture and force plates data during their normal walking were used to establish the patient-specific musculoskeletal models and evaluate the predicted ground reaction forces and ground reaction moments. Reasonable accuracies were achieved for the predicted and measured ground reaction forces and ground reaction moments. The predicted tibiotalar contact forces for all patients using the foot-ground contact model had good consistency with those using force plates data. These findings suggested that the foot-ground contact model could take the place of the force plates data for predicting the tibiotalar contact forces in other total ankle arthroplasty patients, thus providing a simplified and valid platform for further study of the patient-specific prosthetic designs and clinical problems of total ankle arthroplasty in the absence of force plates data.

Footwear and insole design features that reduce neuropathic plantar forefoot ulcer risk in people with diabetes: a systematic literature review

BACKGROUND

In people with diabetes, offloading high-risk foot regions by optimising footwear, or insoles, may prevent ulceration. This systematic review aimed to summarise and evaluate the evidence for footwear and insole features that reduce pathological plantar pressures and the occurrence of diabetic neuropathy ulceration at the plantar forefoot in people with diabetic neuropathy.

METHODS

Six electronic databases (Medline, Cinahl, Amed, Proquest, Scopus, Academic Search Premier) were searched in July 2019. The search period was from 1987 to July 2019. Articles, in English, using footwear or insoles as interventions in patients with diabetic neuropathy were reviewed. Any study design was eligible for inclusion except systematic literature reviews and case reports. Search terms were diabetic foot, physiopathology, foot deformities, neuropath*, footwear, orthoses, shoe, footwear prescription, insole, sock*, ulcer prevention, offloading, foot ulcer, plantar pressure.

RESULTS

Twenty-five studies were reviewed. The included articles used repeated measure (n = 12), case-control (n = 3), prospective cohort (n = 2), randomised crossover (n = 1), and randomised controlled trial (RCT) (n = 7) designs. This involved a total of 2063 participants. Eleven studies investigated footwear, and 14 studies investigated insoles as an intervention. Six studies investigated ulcer recurrence; no study investigated the first occurrence of ulceration. The most commonly examined outcome measures were peak plantar pressure, pressure-time integral and total contact area. Methodological quality varied. Strong evidence existed for rocker soles to reduce peak plantar pressure. Moderate evidence existed for custom insoles to offload forefoot plantar pressure. There was weak evidence that insole contact area influenced plantar pressure.

CONCLUSION

Rocker soles, custom-made insoles with metatarsal additions and a high degree of contact between the insole and foot reduce plantar pressures in a manner that may reduce ulcer occurrence. Most studies rely on reduction in plantar pressure measures as an outcome, rather than the occurrence of ulceration. There is limited evidence to inform footwear and insole interventions and prescription in this population. Further high-quality studies in this field are required.

Clustering classification of diabetic walking abnormalities: a new approach taking into account intralimb coordination patterns

BACKGROUND

It is well recognized that diabetes and peripheral neuropathy have a detrimental effect on gait. However, there are large variations in the results of studies addressing this aspect due to the heterogeneity of diabetic population in relation to presence and severity of diabetes complications. The aim of this study is to adopt an unsupervised classification technique to better elucidate the gait changes throughout the entire spectrum of diabetes and neuropathy.

METHODS

Sixty subjects were assessed and classified into four groups using a fuzzy logic model: 13 controls (55 ± 7years), 18 diabetics subjects without neuropathy (59 ± 6 years, 11 ± 7 diabetes years), 7 with mild neuropathy (56 ± 4years, 19 ± 7 diabetes years), and 22 with moderate to severe neuropathy (57 ± 5 years, 14 ± 8 diabetes years). Data were gathered by six infrared cameras at 100 Hz regarding lower limb joint kinematics (angles and angular velocities) and the relative phase for the hip-ankle, hip-knee, and knee-ankle were calculated. The K-means clustering algorithm was adopted to classify subjects considering the whole kinematics time series. A one-way ANOVA test was used to compare both clinical and kinematics parameters across clusters.

RESULTS

Only the classification based on the intralimb coordination variables succeeded in defining 5 well separated clusters with the following clinical characteristics: controls were grouped mainly in Cluster 2, diabetics in Cluster 4, and neuropathic subjects in Cluster 5 (which included various degrees of severity). Hip-ankle coordination in Clusters 4 and 5 were significantly different (p < 0.05) with respect to Cluster 2, mainly in the stance phase. During the swing phase, differences were observed in the ankle-knee coordination (p < 0.05) across clusters.

CONCLUSION

Classification based on intralimb coordination patterns succeeded in efficiently categorize gait alterations in diabetic subjects. It can be speculated that variables extracted from sagittal plane kinematics might be adopted as a support to clinical decision making in diabetes.

Exploring the Application of Pattern Recognition and Machine Learning for Identifying Movement Phenotypes During Deep Squat and Hurdle Step Movements

Background

Movement screens are increasingly used in sport and rehabilitation to evaluate movement competency. However, common screens are often evaluated using subjective visual detection of a priori prescribed discrete movement features (e.g., spine angle at maximum squat depth) and may not account for whole-body movement coordination, or associations between different discrete features.

Objective

To apply pattern recognition and machine learning techniques to identify whole-body movement pattern phenotypes during the performance of exemplar functional movement screening tasks; the deep squat and hurdle step. Additionally, we also aimed to compare how discrete kinematic measures, commonly used to score movement competency, differed between emergent groups identified via pattern recognition and machine learning.

Methods

Principal component analysis (PCA) was applied to 3-dimensional (3D) trajectory data from participant's deep squat (DS) and hurdle step performance, identifying emerging features that describe orthogonal modes of inter-trial variance in the data. A gaussian mixture model (GMM) was fit and used to cluster the principal component scores as an unsupervised machine learning approach to identify emergent movement phenotypes. Between group features were analyzed using a one-way ANOVA to determine if the objective classifications were significantly different from one another.

Results

Three clusters (i.e., phenotypes) emerged for the DS and right hurdle step (RHS) and 4 phenotypes emerged for the left hurdle step (LHS). Selected discrete points commonly used to score DS and hurdle step movements were different between emergent groups. In regard to the select discrete kinematic measures, 4 out of 5, 7 out of 7 and 4 out of 7, demonstrated a main effect (p < 0.05) between phenotypes for the DS, RHS, and LHS respectively.

Conclusion

Findings support that whole-body movement analysis, pattern recognition and machine learning techniques can objectively identify movement behavior phenotypes without the need to a priori prescribe movement features. However, we also highlight important considerations that can influence outcomes when using machine learning for this purpose.

Differentiating Sitting, Standing, and Walking Through Regional Plantar Pressure Characteristics

Prolonged static weight bearing (WBR) is thought to aggravate plantar heel pain and is common in the workplace, which may put employees at greater risk of developing plantar heel pain. However, objective measures of physical activity and sedentary behaviors in the workplace are lacking, making it difficult to establish or refute the connection between work exposure and plantar heel pain. Characterizing loading patterns during common workplace postures will enhance the understanding of foot function and inform the development of new measurement tools. Plantar pressure data during periods of sitting, standing, and walking were measured in ten healthy participants using the F-Scan in-shoe measurement system (Tekscan Inc, Boston, MA). Peak and average pressure, peak and average contact area, and average pressure differential were analyzed in ten different regions of the foot. A two-way repeated measures analysis of variance (ANOVA) assessed the posture by foot region interaction for each measurement parameter; significant effects of posture by foot region were identified for all five measurement parameters. Ten foot region by measurement parameter combinations were found to significantly differentiate all three postures simultaneously; seven used pressure measures to differentiate while three used area measures. The heel, lateral midfoot (LM), and medial and central forefoot (CFF) encompassed nine of ten areas capable of differentiating all postures simultaneously. This work demonstrates that plantar pressure is a viable means to characterize and differentiate three common workplace postures. The results of this study can inform the development of measurement tools for quantifying posture duration at work.

PAPPI: Personalized analysis of plantar pressure images using statistical modelling and parametric mapping

Quantitative analyses of plantar pressure images typically occur at the group level and under the assumption that individuals within each group display homogeneous pressure patterns. When this assumption does not hold, a personalized analysis technique is required. Yet, existing personalized plantar pressure analysis techniques work at the image level, leading to results that can be unintuitive and difficult to interpret. To address these limitations, we introduce PAPPI: the Personalized Analysis of Plantar Pressure Images. PAPPI is built around the statistical modelling of the relationship between plantar pressures in healthy controls and their demographic characteristics. This statistical model then serves as the healthy baseline to which an individual’s real plantar pressures are compared using statistical parametric mapping. As a proof-of-concept, we evaluated PAPPI on a cohort of 50 hallux valgus patients. PAPPI showed that plantar pressures from hallux valgus patients did not have a single, homogeneous pattern, but instead, 5 abnormal pressure patterns were observed in sections of this population. When comparing these patterns to foot pain scores (i.e. Foot Function Index, Manchester-Oxford Foot Questionnaire) and radiographic hallux angle measurements, we observed that patients with increased pressure under metatarsal 1 reported less foot pain than other patients in the cohort, while patients with abnormal pressures in the heel showed more severe hallux valgus angles and more foot pain. Also, incidences of pes planus were higher in our hallux valgus cohort compared to the modelled healthy controls. PAPPI helped to clarify recent discrepancies in group-level plantar pressure studies and showed its unique ability to produce quantitative, interpretable, and personalized analyses for plantar pressure images.

Gait changes in persons with diabetes: Early risk marker for diabetic foot ulcer

Background Increasing prevalence of diabetic foot ulcer (DFU) and subsequent foot amputation in persons with type 2 diabetic neuropathy is a well known fact. The present study was aimed to identify the initial risk marker for DFU. Methods Dynamic plantar pressure analysis was done for persons with type 2 diabetes mellitus (T2DM) without neuropathy (D), patients with diabetic neuropathy (DN) with normal foot profile and healthy persons with normal foot profile (C). Results The data showed a significant difference in dynamic peak plantar pressure between C and DN (P = 0.035) and no significant difference between D and DN (P = 0.997). The dynamic segmental peak plantar pressure results showed significant difference only in the medial heel region (P = 0.009) among the three groups. Conclusions Gait variations and restrictions in subtalar and first metatarsophalangeal joint are found in persons with diabetic neuropathy even before the onset of foot deformity.

A state-of-the-art review of foot pressure

The science of foot pressure studies the forces acting on the bottom and different regions of the foot along with the pressure exerted on the plantar surface with the interacting surface in contact. The information derived gave impact to human biomechanical assessment on body balance and ergonomics posture during gait. Various experiments designed at generating foot pressure data returns only with limited knowledge generated. Obviously, the procedure for experiment design needs to be properly understood from the foot morphology aspects; healthiness, footwear, surface in contact, load and forces impacts, and the foot sensitivity as well as the specification for the foot pressure. This paper reviews the proper preliminary experimental setups for foot pressure measurement analysis during static or dynamic gait. The strength and limitations of recent devices used and considerable variables are also discussed. The overall review explains that the comfortable natural gait in relation to the aspects of sensitivity, load, time duration, and stability are the standard considerations for plantar pressure experiments.

Pattern detection from seating pressure distribution during wheelchair motion using deep embedded clustering

To minimize the occurrence of ulcers in high-risk mobile individuals such as wheelchair users, it is necessary to detect all typical distribution patterns and to indentify the patterns that may be associated with pressure ulcers. However, pattern detection is difficult because the pressure distribution during motion includes a variety of patterns compared to those of static postures. Thus, the establishment of a method to detect typical patterns based on distribution patterns is important. We utilized deep embedded clustering for identification purposes. This clustering technique extracts features using auto-encoder and simultaneously optimizes data points into the clusters, which might realize good clustering performance due to the detected optimal features. We used a pressure distribution dataset that was pre-labeled by nursing experts. The dataset consisted of a total of 26944 distribution images with ten class annotations. The clustering method including traditional approaches (k-means and principal component analysis plus k-means) were compared with deep embedded clustering while the threshold to noise reduction was changing. The deep embedded clustering with 80 mmHg threshold achieved the best performance. This approach also tended to be less dependent on the threshold values.

Distribution of the Highest Plantar Pressure Regions in Patients with Diabetes and Its Association with Peripheral Neuropathy, Gender, Age, and BMI: One Centre Study

The abnormal plantar pressure distribution and value play a key role in the formation of plantar calluses and diabetic foot ulcer. The prevalence of the highest pressure different distribution and its association with various factors among patients with diabetes is not well known. The study purpose was to evaluate the prevalence of different regions for the highest pressure on the sole and its association with selected factors among patients with diabetes. Medical records of nonulcer patients were retrospectively analysed. The relationship between pressure patterns on the sole obtained during a pedobarographic test as a semiquantitative assessment with colourful print analysis and neuropathy, gender, age, and BMI was searched. The most common location of the highest pressure was the central part of the forefoot. No association was found between the different highest pressure regions and age, sensory neuropathy, calluses, and foot deformities. The highest pressure on the lateral part of the foot and midfoot was observed more often in females and in patients with a BMI ≥ 35. The prevalence of the highest pressure on the forefoot was more common in patients with a BMI < 35. Conclusions. The most frequent regions of the highest pressure on the sole in patients with diabetes were the central part of the forefoot (2-3 metatarsal heads) with no simple relationship to the assessed variables other than BMI < 35. Female gender and higher BMI seem to be responsible for shifting the place of the highest pressure to other places of the foot.

STAPP: Spatiotemporal analysis of plantar pressure measurements using statistical parametric mapping

BACKGROUND

Pedobarography produces large sets of plantar pressure samples that are routinely subsampled (e.g. using regions of interest) or aggregated (e.g. center of pressure trajectories, peak pressure images) in order to simplify statistical analysis and provide intuitive clinical measures.

RESEARCH QUESTION

We hypothesize that these data reductions discard gait information that can be used to differentiate between groups or conditions.

METHODS

To test the hypothesis of null information loss, we created an implementation of statistical parametric mapping (SPM) for dynamic plantar pressure datasets (i.e. plantar pressure videos). Our SPM software framework brings all plantar pressure videos into anatomical and temporal correspondence, then performs statistical tests at each sampling location in space and time. Novelly, we introduce non-linear temporal registration into the framework in order to normalize for timing differences within the stance phase. We refer to our software framework as STAPP: spatiotemporal analysis of plantar pressure measurements. Using STAPP, we tested our hypothesis on plantar pressure videos from 33 healthy subjects walking at different speeds.

RESULTS

As walking speed increased, STAPP was able to identify significant decreases in plantar pressure at mid-stance from the heel through the lateral forefoot. The extent of these plantar pressure decreases has not previously been observed using existing plantar pressure analysis techniques.

SIGNIFICANCE

We therefore conclude that the subsampling of plantar pressure videos - a task which led to the discarding of gait information in our study - can be avoided using STAPP.

Technologies to monitor the health of loaded skin tissues

There are many situations where the skin and underlying soft tissues are compromised by mechanical loading in the form or pressure, or pressure in combination with shear. If sustained, this can lead to damage in the tissues particularly adjacent to bony prominences, resulting in chronic wounds. An array of bioengineering technologies have been adopted to assess the integrity of loaded soft tissues. This paper aims to review these approaches for the quantification, simulation and early detection of mechanically-induced skin damage. The review considers different measurements at the interface between the skin and support surface/medical device, involving pressure, shear, friction and the local microclimate. The potential of the techniques to monitor the physiological response of the skin to these external stimuli including biophysical measurement devices and sampling of biofluids are critically analysed. In addition, it includes an analysis of medical imaging technologies and computational modelling to provide a means by which tissue deformation can be quantified and thresholds for tissue damage defined. Bioengineering measurement and imaging technologies have provided an insight into the temporal status of loaded skin. Despite the advances in technology, to date, the translation to clinical tools which are robust and cost effective has been limited. There is a need to adapt existing technologies and simulation platforms to enable patients, carers and clinicians to employ appropriate intervention strategies to minimise soft tissue damage.

Validation of plantar pressure simulations using finite and discrete element modelling in healthy and diabetic subjects

Plantar pressure simulation driven by integrated 3D motion capture data, using both a finite element and a discrete element model, is compared for ten healthy and ten diabetic neuropathic subjects. The simulated peak pressure deviated on average between 16.7 and 34.2% from the measured peak pressure. The error in the position of the peak pressure was on average smaller than 4.2 cm. No method was more accurate than the other although statistical differences were found between them. Both techniques are thus complementary and useful tools to better understand the alteration of diabetic foot biomechanics during gait.

Discrete sensors distribution for accurate plantar pressure analyses

The aim of this study was to determine the distribution of discrete sensors under the footprint for accurate plantar pressure analyses. For this purpose, two different sensor layouts have been tested and compared, to determine which was the most accurate to monitor plantar pressure with wireless devices in research and/or clinical practice. Ten healthy volunteers participated in the study (age range: 23-58 years). The barycenter of pressures (BoP) determined from the plantar pressure system (W-inshoe®) was compared to the center of pressures (CoP) determined from a force platform (AMTI) in the medial-lateral (ML) and anterior-posterior (AP) directions. Then, the vertical ground reaction force (vGRF) obtained from both W-inshoe® and force platform was compared for both layouts for each subject. The BoP and vGRF determined from the plantar pressure system data showed good correlation (SCC) with those determined from the force platform data, notably for the second sensor organization (ML SCC= 0.95; AP SCC=0.99; vGRF SCC=0.91). The study demonstrates that an adjusted placement of removable sensors is key to accurate plantar pressure analyses. These results are promising for a plantar pressure recording outside clinical or laboratory settings, for long time monitoring, real time feedback or for whatever activity requiring a low-cost system.

Efficacy measures associated to a plantar pressure based classification system in diabetic foot medicine

AIMS

The concept of 'classification' has, similar to many other diseases, been found to be fundamental in the field of diabetic medicine. In the current study, we aimed at determining efficacy measures of a recently published plantar pressure based classification system.

METHODS

Technical efficacy of the classification system was investigated by applying a high resolution, pixel-level analysis on the normalized plantar pressure pedobarographic fields of the original experimental dataset consisting of 97 patients with diabetes and 33 persons without diabetes. Clinical efficacy was assessed by considering the occurence of foot ulcers at the plantar aspect of the forefoot in this dataset. Classification efficacy was assessed by determining the classification recognition rate as well as its sensitivity and specificity using cross-validation subsets of the experimental dataset together with a novel cohort of 12 patients with diabetes.

RESULTS

Pixel-level comparison of the four groups associated to the classification system highlighted distinct regional differences. Retrospective analysis showed the occurence of eleven foot ulcers in the experimental dataset since their gait analysis. Eight out of the eleven ulcers developed in a region of the foot which had the highest forces. Overall classification recognition rate exceeded 90% for all cross-validation subsets. Sensitivity and specificity of the four groups associated to the classification system exceeded respectively the 0.7 and 0.8 level in all cross-validation subsets.

CONCLUSIONS

The results of the current study support the use of the novel plantar pressure based classification system in diabetic foot medicine. It may particularly serve in communication, diagnosis and clinical decision making.

Comparative Clustering of Plantar Pressure Distributions in Diabetics with Polyneuropathy May Be Applied to Reveal Inappropriate Biomechanical Stress

In diabetic patients, excessive peak plantar pressure has been identified as major risk factor for ulceration. Analyzing plantar pressure distributions potentially improves the identification of patients with a high risk for foot ulceration development. The goal of this study was to classify regional plantar pressure distributions. By means of a sensor-equipped insole, pressure recordings of healthy controls (n = 18) and diabetics with severe polyneuropathy (n = 25) were captured across eight foot regions. The study involved a controlled experimental protocol with multiple sessions, where a session contained several cycles of pressure exposure. Clustering was used to identify subgroups of study participants that are characterized by similar pressure distributions. For both analyzed groups, the number of clusters to best describe the pressure profiles was four. When both groups were combined, analysis again led to four distinct clusters. While three clusters did not separate between healthy and diabetic volunteers the fourth cluster was only represented by diabetics. Here the pressure distribution pattern is characterized by a focal point of pressure application on the forefoot and low pressure on the lateral region. Our data suggest that pressure clustering is a feasible means to identify inappropriate biomechanical plantar stress.

Intralimb Coordination Patterns in Absent, Mild, and Severe Stages of Diabetic Neuropathy: Looking Beyond Kinematic Analysis of Gait Cycle

AIM

Diabetes Mellitus progressively leads to impairments in stability and joint motion and might affect coordination patterns, mainly due to neuropathy. This study aims to describe changes in intralimb joint coordination in healthy individuals and patients with absent, mild and, severe stages of neuropathy.

METHODS

Forty-seven diabetic patients were classified into three groups of neuropathic severity by a fuzzy model: 18 without neuropathy (DIAB), 7 with mild neuropathy (MILD), and 22 with moderate to severe neuropathy (SVRE). Thirteen healthy subjects were included as controls (CTRL). Continuous relative phase (CRP) was calculated at each instant of the gait cycle for each pair of lower limb joints. Analysis of Variance compared each frame of the CRP time series and its standard deviation among groups (α = 5%).

RESULTS

For the ankle-hip CRP, the SVRE group presented increased variability at the propulsion phase and a distinct pattern at the propulsion and initial swing phases compared to the DIAB and CTRL groups. For the ankle-knee CRP, the 3 diabetic groups presented more anti-phase ratios than the CTRL group at the midstance, propulsion, and terminal swing phases, with decreased variability at the early stance phase. For the knee-hip CRP, the MILD group showed more in-phase ratio at the early stance and terminal swing phases and lower variability compared to all other groups. All diabetic groups were more in-phase at early the midstance phase (with lower variability) than the control group.

CONCLUSION

The low variability and coordination differences of the MILD group showed that gait coordination might be altered not only when frank evidence of neuropathy is present, but also when neuropathy is still incipient. The ankle-knee CRP at the initial swing phase showed distinct patterns for groups from all degrees of neuropathic severity and CTRLs. The ankle-hip CRP pattern distinguished the SVRE patients from other diabetic groups, particularly in the transitional phase from stance to swing.

Plantar loading reflects ulceration risks of diabetic foot with toe deformation

Diabetes has been one of the most common chronic diseases all over the world. The purpose of this study was to quantitatively assess the foot loading characteristics of diabetic patients with fifth-toe deformity through a comparative analysis with diabetic patients with healthy and normal feet. Six neuropathic diabetic female subjects with the fifth-toe deformation and six age-matched neuropathic diabetic controls without any feet deformities participated in the walking test. Dynamic barefoot plantar pressure was measured with Novel EMED force plate. Peak pressure and pressure-time integral for all 7 foot regions (rearfoot, midfoot, lateral forefoot, central forefoot, medial forefoot, great toe, and other toes) were collected. Peak pressure was significantly higher in the patients with toe deformity in rearfoot, central forefoot, and great toe regions compared with the control group. Meanwhile, loading sustaining period extended longer in great toe region of deformed group than in that of the control group, and the center of pressure was nearly in the big toe region during toe offstage. Diabetic patients with fifth-toe deformity could have plantar contact area reduction in the other toes part and increased loading to the great toe part. The result showed that fifth-toe deformity was associated with potential ulceration risk especially in hallux region.

Abnormalities of plantar pressure distribution in early, intermediate, and late stages of diabetic neuropathy

Inconsistent findings with regard to plantar pressure while walking in the diabetic population may be due to the heterogeneity of the studied groups resulting from the classification/grouping criteria adopted. The clinical diagnosis and classification of diabetes have inherent uncertainties that compromise the definition of its onset and the differentiation of its severity stages. A fuzzy system could improve the precision of the diagnosis and classification of diabetic neuropathy because it takes those uncertainties into account and combines different assessment methods. Here, we investigated how plantar pressure abnormalities evolve throughout different severity stages of diabetic polyneuropathy (absent, n=38; mild, n=20; moderate, n=47; severe, n=24). Pressure distribution was analysed over five areas while patients walked barefoot. Patients with mild neuropathy displayed an increase in pressure-time integral at the forefoot and a lower peak pressure at the heel. The peak and pressure-time integral under the forefoot and heel were aggravated in later stages of the disease (moderate and severe) compared with early stages of the disease (absent and mild). In the severe group, lower pressures at the lateral forefoot and hallux were observed, which could be related to symptoms that develop with the aggravation of neuropathy: atrophy of the intrinsic foot muscles, reduction of distal muscle activity, and joint stiffness. Although there were clear alterations over the forefoot and in a number of plantar areas with higher pressures within each severity stage, they did not follow the aggravation evolution of neuropathy classified by the fuzzy model. Based on these results, therapeutic interventions should begin in the early stages of this disease to prevent further consequences of the disease.

Classification of forefoot plantar pressure distribution in persons with diabetes: a novel perspective for the mechanical management of diabetic foot?

Background

The aim of this study was to identify groups of subjects with similar patterns of forefoot loading and verify if specific groups of patients with diabetes could be isolated from non-diabetics.

Methodology/Principal Findings

Ninety-seven patients with diabetes and 33 control participants between 45 and 70 years were prospectively recruited in two Belgian Diabetic Foot Clinics. Barefoot plantar pressure measurements were recorded and subsequently analysed using a semi-automatic total mapping technique. Kmeans cluster analysis was applied on relative regional impulses of six forefoot segments in order to pursue a classification for the control group separately, the diabetic group separately and both groups together. Cluster analysis led to identification of three distinct groups when considering only the control group. For the diabetic group, and the computation considering both groups together, four distinct groups were isolated. Compared to the cluster analysis of the control group an additional forefoot loading pattern was identified. This group comprised diabetic feet only. The relevance of the reported clusters was supported by ANOVA statistics indicating significant differences between different regions of interest and different clusters.

Conclusion/s Significance

There seems to emerge a new era in diabetic foot medicine which embraces the classification of diabetic patients according to their biomechanical profile. Classification of the plantar pressure distribution has the potential to provide a means to determine mechanical interventions for the prevention and/or treatment of the diabetic foot.