Hotspots: Adherence in home foot temperature monitoring interventions for at-risk feet with diabetes-A narrative review.

BACKGROUND: Home foot temperature monitoring (HFTM) is recommended for those at moderate to high ulcer risk. Where a > 2.2°C difference in temperature between feet (hotspot) is detected, it is suggested that individuals (1) notify a healthcare professional (HCP); (2) reduce daily steps by 50%. We assess adherence to this and HFTM upon detecting a recurrent hotspot. METHODS: PubMed and Google Scholar were searched until 9 June 2023 for English-language peer-reviewed HFTM studies which reported adherence to HFTM, daily step reduction or HCP hotspot notification. The search returned 1030 results excluding duplicates of which 28 were shortlisted and 11 included. RESULTS: Typical adherence among HFTM study participants for >3 days per week was 61%-93% or >80% of study duration was 55.6%-83.1%. Monitoring foot temperatures >50% of the study duration was associated with decreased ulcer risk (Odds Ratio: 0.50, p < 0.001) in one study (n = 173), but no additional risk reduction was found for >80% adherence. Voluntary dropout was 5.2% (Smart mats); 8.1% (sock sensor) and 4.8%-35.8% (infrared thermometers). Only 16.9%-52.5% of participants notified an HCP upon hotspot detection. Objective evidence of adherence to 50% reduction in daily steps upon hotspot detection was limited to one study where the average step reduction was a pedometer-measured 51.2%. CONCLUSIONS: Ulcer risk reduction through HFTM is poorly understood given only half of the participants notify HCPs of recurrent hotspots and the number of reducing daily steps is largely unknown. HFTM adherence and dropout are variable and more research is needed to determine factors affecting adherence and those likely to adhere.

Contralateral foot temperature monitoring during Charcot immobilisation: A systematic review.

AIMS: Contralateral temperature difference (CTD) is a frequently used marker of healing in Charcot neuro-osteoarthropathy (CN). We aimed to determine whether there is a consistent approach to CTD measurement during healing and the decision-making process around cessation of immobilisation. MATERIALS AND METHODS: Medline, Scopus, and Web of Science were searched until February 2022 for peer-reviewed studies using keywords, including (('arthropathy' OR 'osteoarthropathy' OR 'osteopathy' OR 'neuroarthropathy') AND 'Charcot' AND ('temperature')), which returned 789 results excluding duplicates. Included studies monitored CTD in those with active CN to (i) assess the healing process and (ii) assist in determining the transition from immobilisation. RESULTS: Thirty four studies in total (n = 677 participants) were shortlisted and 19 were included after full paper review. Average CTD at presentation varied from 1.6 to 8.0°C with insufficient data to determine if CTD was proportional to severity of Charcot. Evidence of a relationship between CTD and radiographic or scintigraphy-based markers of healing varied depending on the methodology employed. Threshold CTD for the cessation of immobilisation ranged from <1°C to <2°C. Most frequently it was <2°C sustained for 2-3 visits. Temperature was monitored typically every 2-6 weeks using handheld thermometry at CN site(s) after resting the feet for 15 min. Device type, accuracy/reliability, and ambient temperature were rarely reported. CONCLUSIONS: Further research on CTD and radiographic and radiotracer markers is needed involving larger cohorts. Standardisation in reporting of thermometry device type, accuracy and reliability, foot resting times, and ambient temperature controls is essential to facilitate the comparison of studies, meta-analysis, and evaluation of different immobilisation interventions.

Feature selection for unsupervised machine learning of accelerometer data physical activity clusters - A systematic review.

BACKGROUND: Identifying clusters of physical activity (PA) from accelerometer data is important to identify levels of sedentary behaviour and physical activity associated with risks of serious health conditions and time spent engaging in healthy PA. Unsupervised machine learning models can capture PA in everyday free-living activity without the need for labelled data. However, there is scant research addressing the selection of features from accelerometer data. The aim of this systematic review is to summarise feature selection techniques applied in studies concerned with unsupervised machine learning of accelerometer-based device obtained physical activity, and to identify commonly used features identified through these techniques. Feature selection methods can reduce the complexity and computational burden of these models by removing less important features and assist in understanding the relative importance of feature sets and individual features in clustering. METHOD: We conducted a systematic search of Pubmed, Medline, Google Scholar, Scopus, Arxiv and Web of Science databases to identify studies published before January 2021 which used feature selection methods to derive PA clusters using unsupervised machine learning models. RESULTS: A total of 13 studies were eligible for inclusion within the review. The most popular feature selection techniques were Principal Component Analysis (PCA) and correlation-based methods, with k-means frequently used in clustering accelerometer data. Cluster quality evaluation methods were diverse, including both external (e.g. cluster purity) or internal evaluation measures (silhouette score most frequently). Only four of the 13 studies had more than 25 participants and only four studies included two or more datasets. CONCLUSION: There is a need to assess multiple feature selection methods upon large cohort data consisting of multiple (3 or more) PA datasets. The cut-off criteria e.g. number of components, pairwise correlation value, explained variance ratio for PCA, etc. should be expressly stated along with any hyperparameters used in clustering.

FilterK: A new outlier detection method for k-means clustering of physical activity.

In this paper, a new algorithm denoted as FilterK is proposed for improving the purity of k-means derived physical activity clusters by reducing outlier influence. We applied it to physical activity data obtained with body-worn accelerometers and clustered using k-means. We compared its performance with three existing outlier detection methods: Local Outlier Factor, Isolation Forests and KNN using the ground truth (class labels), average cluster and event purity (ACEP). FilterK provided comparable gains in ACEP (0.581 â†’ 0.596 compared to 0.580-0.617) whilst removing a lower number of outliers than the other methods (4% total dataset size vs 10% to achieve this ACEP). The main focus of our new outlier detection method is to improve the cluster purities of physical activity accelerometer data, but we also suggest it may be potentially applied to other types of dataset captured by k-means clustering. We demonstrate our method using a k-means model trained on two independent accelerometer datasets (training n = 90) and re-applied to an independent dataset (test n = 41). Labelled physical activities include lying down, sitting, standing, household chores, walking (laboratory and non-laboratory based), stairs and running. This type of clustering algorithm could be used to assist with identifying optimal physical activity patterns for health.

A fitting problem: Standardising shoe fit standards to reduce related diabetic foot ulcers.

AIMS: Incorrectly fitting shoes are implicated in callus formation and a significant proportion of diabetic foot ulcers, yet remain surprisingly prevalent. We review the current shoe fit guidelines for consistency and discuss ways in which technology may assist us in standardising methods of footwear assessment. METHODS: Narrative review. RESULTS: Incorrectly fitted shoes are implicated in the development of some diabetic foot ulcers yet surprisingly there's no consensus on shoe fit, despite substantial spending on prescription footwear. Suggested toe gaps vary from 6 to 20 mm and measurement methods also vary from Brannock Devices and callipers to manual measurement. CONCLUSIONS: To prevent fit-related foot ulceration, we need to standardise our biomechanical definition of fit. Future research should (1) evaluate the potential use of 3D scanning technology to provide a standardised means of capturing foot morphology; (2) develop a working biomechanical definition of fit, including toe gap through the identification of key physiological markers that capture and predict dynamic foot shape changes during different physical activities and body weight loading conditions; and (3) determine whether changes in dynamic foot shape of those with diabetes differs from those without, impacting on their shoe fitting needs, potentially necessitating specialist footwear at an earlier stage to avoid ulceration.