Search

Physical rehabilitation is an important process for patients recovering after surgery. In this paper, we propose and develop a framework to monitor on-bed range of motion exercises that allows physical therapists to evaluate patient adherence to set exercise programs. Using a dense pressure sensitive bedsheet, a sequence of pressure maps are produced and analyzed using manifold learning techniques. We compare two methods, Local Linear Embedding and Isomap, to reduce the dimensionality of the pressure map data. Once the image sequences are converted into a low dimensional manifold, the manifolds can be compared to expected prior data for the rehabilitation exercises. Furthermore, a measure to compare the similarity of manifolds is presented along with experimental results for five on-bed rehabilitation exercises. The evaluation of this framework shows that exercise compliance can be tracked accurately according to prescribed treatment programs. [ABSTRACT FROM PUBLISHER]

Introduction: The American Heart Association (AHA) Life's Simple 7 (LS7) is a metric that is known to predict cardiovascular disease. What is unknown are the factors that predict cardiovascular health (CVH) scores over time, and what patterns of CVH scores are most common. Machine learning algorithms can identify these patterns using a data-driven approach. Our objective was to identify most common risk patterns over time and assess smartphone usage statistics to identify their predictors. Methods: We recruited 303 Northwestern University undergraduates and cluster randomized them to one of two health intervention groups. The intervention group focused on cardiovascular health behaviors (n=130) were provided with an app that delivered an intervention targeting the LS7 metrics smoking, physical activity, fruit and vegetable consumption, and weight. All participants reported levels of four behaviors weekly. Health scores were calculated as the sum of metrics meeting ideal criteria; CVH behaviors scored per the AHA LS7. Behavioral patterns over time used in the K-means unsupervised clustering algorithm include the mean, slope and variance of trends. K-Silhouette was used to determine optimal number of clusters, and predictors from the CVH intervention app use data were examined to determine which contributed to distinguishing between clusters. Results: We identified four clusters using K-silhouette criteria (Figure 1). Of the smartphone meta-data collected, 2 were significantly different between clusters. We observed greater number of interactions/week (p <.0001) and greater number of app-use days/week (p <.0001) for both "healthy-steady" and "improvers" as compared with "decliners" and "unhealthy-steady" clusters. Conclusions: Our findings suggest that frequency and consistency in health behavior monitoring predict healthier trends over time. These usage behaviors may be targets for behavioral interventions. Further work will investigate if these factors meaningfully predict CVH at the end of a 2 year intervention. In the future, predictors could be used as signals that an individual needs behavioral intervention. [ABSTRACT FROM AUTHOR]

Background: Predicting the likelihood of success of weight loss interventions using machine learning (ML) models may enhance intervention effectiveness by enabling timely and dynamic modification of intervention components for nonresponders to treatment. However, a lack of understanding and trust in these ML models impacts adoption among weight management experts. Recent advances in the field of explainable artificial intelligence enable the interpretation of ML models, yet it is unknown whether they enhance model understanding, trust, and adoption among weight management experts., Objective: This study aimed to build and evaluate an ML model that can predict 6-month weight loss success (ie, ≥7% weight loss) from 5 engagement and diet-related features collected over the initial 2 weeks of an intervention, to assess whether providing ML-based explanations increases weight management experts' agreement with ML model predictions, and to inform factors that influence the understanding and trust of ML models to advance explainability in early prediction of weight loss among weight management experts., Methods: We trained an ML model using the random forest (RF) algorithm and data from a 6-month weight loss intervention (N=419). We leveraged findings from existing explainability metrics to develop Prime Implicant Maintenance of Outcome (PRIMO), an interactive tool to understand predictions made by the RF model. We asked 14 weight management experts to predict hypothetical participants' weight loss success before and after using PRIMO. We compared PRIMO with 2 other explainability methods, one based on feature ranking and the other based on conditional probability. We used generalized linear mixed-effects models to evaluate participants' agreement with ML predictions and conducted likelihood ratio tests to examine the relationship between explainability methods and outcomes for nested models. We conducted guided interviews and thematic analysis to study the impact of our tool on experts' understanding and trust in the model., Results: Our RF model had 81% accuracy in the early prediction of weight loss success. Weight management experts were significantly more likely to agree with the model when using PRIMO (χ 2 =7.9; P=.02) compared with the other 2 methods with odds ratios of 2.52 (95% CI 0.91-7.69) and 3.95 (95% CI 1.50-11.76). From our study, we inferred that our software not only influenced experts' understanding and trust but also impacted decision-making. Several themes were identified through interviews: preference for multiple explanation types, need to visualize uncertainty in explanations provided by PRIMO, and need for model performance metrics on similar participant test instances., Conclusions: Our results show the potential for weight management experts to agree with the ML-based early prediction of success in weight loss treatment programs, enabling timely and dynamic modification of intervention components to enhance intervention effectiveness. Our findings provide methods for advancing the understandability and trust of ML models among weight management experts., (©Glenn J Fernandes, Arthur Choi, Jacob Michael Schauer, Angela F Pfammatter, Bonnie J Spring, Adnan Darwiche, Nabil I Alshurafa. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 06.09.2023.)

Smoking is the leading cause of preventable death worldwide. Cigarette smoke includes thousands of chemicals that are harmful and cause tobacco-related diseases. To date, the causality between human exposure to specific compounds and the harmful effects is unknown. A first step in closing the gap in knowledge has been measuring smoking topography, or how the smoker smokes the cigarette (puffs, puff volume, and duration). However, current gold-standard approaches to smoking topography involve expensive, bulky, and obtrusive sensor devices, creating unnatural smoking behavior and preventing their potential for real-time interventions in the wild. Although motion-based wearable sensors and their corresponding machine-learned models have shown promise in unobtrusively tracking smoking gestures, they are notorious for confounding smoking with other similar hand-to-mouth gestures such as eating and drinking. In this paper, we present SmokeMon, a chest-worn thermal-sensing wearable system that can capture spatial, temporal, and thermal information around the wearer and cigarette all day to unobtrusively and passively detect smoking events. We also developed a deep learning-based framework to extract puffs and smoking topography. We evaluate SmokeMon in both controlled and free-living experiments with a total of 19 participants, more than 110 hours of data, and 115 smoking sessions achieving an F1-score of 0.9 for puff detection in the laboratory and 0.8 in the wild. By providing SmokeMon as an open platform, we provide measurement of smoking topography in free-living settings to enable testing of smoking topography in the real world, with potential to facilitate timely smoking cessation interventions.

Researchers have been leveraging wearable cameras to both visually confirm and automatically detect individuals' eating habits. However, energy-intensive tasks such as continuously collecting and storing RGB images in memory, or running algorithms in real-time to automate detection of eating, greatly impacts battery life. Since eating moments are spread sparsely throughout the day, battery life can be mitigated by recording and processing data only when there is a high likelihood of eating. We present a framework comprising a golf-ball sized wearable device using a low-powered thermal sensor array and real-time activation algorithm that activates high-energy tasks when a hand-to-mouth gesture is confirmed by the thermal sensor array. The high-energy tasks tested are turning on the RGB camera (Trigger RGB mode) and running inference on an on-device machine learning model (Trigger ML mode). Our experimental setup involved the design of a wearable camera, 6 participants collecting 18 hours of data with and without eating, the implementation of a feeding gesture detection algorithm on-device, and measures of power saving using our activation method. Our activation algorithm demonstrates an average of at-least 31.5% increase in battery life time, with minimal drop of recall (5%) and without impacting the accuracy of detecting eating (a slight 4.1% increase in F1-Score).

Automated detection and validation of fine-grained human activities from egocentric vision has gained increased attention in recent years due to the rich information afforded by RGB images. However, it is not easy to discern how much rich information is necessary to detect the activity of interest reliably. Localization of hands and objects in the image has proven helpful to distinguishing between hand-related fine-grained activities. This paper describes the design of a hand-object-based mask obfuscation method (HOBM) and assesses its effect on automated recognition of fine-grained human activities. HOBM masks all pixels other than the hand and object in-hand, improving the protection of personal user information (PUI). We test a deep learning model trained with and without obfuscation using a public egocentric activity dataset with 86 class labels and achieve almost similar classification accuracies (2% decrease with obfuscation). Our findings show that it is possible to protect PUI at smaller image utility costs (loss of accuracy).

Screen time is associated with several health risk behaviors including mindless eating, sedentary behavior, and decreased academic performance. Screen time behavior is traditionally assessed with self-report measures, which are known to be burdensome, inaccurate, and imprecise. Recent methods to automatically detect screen time are geared more towards detecting television screens from wearable cameras that record high-resolution video. Activity-oriented wearable cameras (i.e., cameras oriented towards the wearer with a fisheye lens) have recently been designed and shown to reduce privacy concerns, yet pose a greater challenge in capturing screens due to their orientation and fewer pixels on target. Methods that detect screens from low-power, low-resolution wearable camera video are needed given the increased adoption of such devices in longitudinal studies. We propose a method that leverages deep learning algorithms and lower-resolution images from an activity-oriented camera to detect screen presence from multiple types of screens with high variability of pixel on target (e.g., near and far TV, smartphones, laptops, and tablets). We test our system in a real-world study comprising 10 individuals, 80 hours of data, and 1.2 million low-resolution RGB frames. Our results outperform existing state-of-the-art video screen detection methods yielding an F1-score of 81%. This paper demonstrates the potential for detecting screen-watching behavior in longitudinal studies using activity-oriented cameras, paving the way for a nuanced understanding of screen time's relationship with health risk behaviors.

Wearable cameras provide an informative view of wearer activities, context, and interactions. Video obtained from wearable cameras is useful for life-logging, human activity recognition, visual confirmation, and other tasks widely utilized in mobile computing today. Extracting foreground information related to the wearer and separating irrelevant background pixels is the fundamental operation underlying these tasks. However, current wearer foreground extraction methods that depend on image data alone are slow, energy- in efficient, and even inaccurate in some cases, making many tasks-like activity recognition- challenging to implement in the absence of significant computational resources. To fill this gap, we built ActiSight, a wearable RGB-Thermal video camera that uses thermal information to make wearer segmentation practical for body-worn video. Using ActiSight, we collected a total of 59 hours of video from 6 participants, capturing a wide variety of activities in a natural setting. We show that wearer foreground extracted with ActiSight achieves a high dice similarity score while significantly lowering execution time and energy cost when compared with an RGB-only approach.

We present the design, implementation, and evaluation of a multi-sensor, low-power necklace, NeckSense , for automatically and unobtrusively capturing fine-grained information about an individual's eating activity and eating episodes, across an entire waking day in a naturalistic setting. NeckSense fuses and classifies the proximity of the necklace from the chin, the ambient light, the Lean Forward Angle, and the energy signals to determine chewing sequences, a building block of the eating activity. It then clusters the identified chewing sequences to determine eating episodes. We tested NeckSense on 11 participants with and 9 participants without obesity, across two studies, where we collected more than 470 hours of data in a naturalistic setting. Our results demonstrate that NeckSense enables reliable eating detection for individuals with diverse body mass index (BMI) profiles, across an entire waking day, even in free-living environments. Overall, our system achieves an F1-score of 81.6% in detecting eating episodes in an exploratory study. Moreover, our system can achieve an F1-score of 77.1% for episodes even in an all-day-long free-living setting. With more than 15.8 hours of battery life, NeckSense will allow researchers and dietitians to better understand natural chewing and eating behaviors. In the future, researchers and dietitians can use NeckSense to provide appropriate real-time interventions when an eating episode is detected or when problematic eating is identified.

High levels of stress during pregnancy increase the chances of having a premature or low-birthweight baby. Perceived self-reported stress does not often capture or align with the physiological and behavioral response. But what if there was a self-report measure that could better capture the physiological response? Current perceived stress self-report assessments require users to answer multi-item scales at different time points of the day. Reducing it to one question, using microinteraction-based ecological momentary assessment (micro-EMA, collecting a single in situ self-report to assess behaviors) allows us to identify smaller or more subtle changes in physiology. It also allows for more frequent responses to capture perceived stress while at the same time reducing burden on the participant. We propose a framework for selecting the optimal micro-EMA that combines unbiased feature selection and unsupervised Agglomerative clustering. We test our framework in 18 women performing 16 activities in-lab wearing a Biostamp, a NeuLog, and a Polar chest strap. We validated our results in 17 pregnant women in real-world settings. Our framework shows that the question "How worried were you?" results in the highest accuracy when using a physiological model. Our results provide further in-depth exposure to the challenges of evaluating stress models in real-world situations.

One in five US adults will be diagnosed with skin cancer. As most skin cancers are attributable to sun exposure, this risk factor is an important target for research and intervention. Most sun exposure measures assess frequency of specific sun-protection behaviors, which does not account for the use of multiple, potentially overlapping sun-protection methods. In contrast, the Daily Minutes of Unprotected Sun Exposure (MUSE) Inventory assesses sun-protection behavior during self-reported activities, providing several useful metrics, including duration of unprotected sun exposure on 17 body sites, combined to yield an overall MUSE score weighted by percent of body exposed. The present study was conducted July-September 2017, in Chicago, IL USA. For 10 days, participants (39 melanoma survivors; M age  = 58.59, 64.5% female) wore an ultraviolet radiation (UVR) sensor and completed the Daily MUSE Inventory each evening. The Sun Habits Survey was completed at the end of the study. Outdoor time reported in the MUSE Inventory significantly predicted outdoor time recorded by UVR sensors, B  = 0.53, p  < .001. For all sun-protection behaviors except shade, reports from the Daily MUSE Inventory (i.e., percentage of outdoor time a particular strategy was used) correlated with frequency ratings of the same strategy from the Sun Habits Survey ( r s = 0.66-0.75, p  < .05). In sum, the Daily MUSE Inventory corresponds with sensor and survey data, and provides a novel metric of unprotected sun exposure that will be useful for evaluating overall extent of sun exposure, including exposure on several smaller body sites that are at high risk for skin cancer.

Objective: Studies have revealed that non-adherence to prescribed medication can lead to hospital readmissions, clinical complications, and other negative patient outcomes. Though many techniques have been proposed to improve patient adherence rates, they suffer from low accuracy. Our objective is to develop and test a novel system for assessment of medication adherence., Methods: Recently, several smart pill bottle technologies have been proposed, which can detect when the bottle has been opened, and even when a pill has been retrieved. However, very few systems can determine if the pill is subsequently ingested or discarded. We propose a system for detecting user adherence to medication using a smart necklace, capable of determining if the medication has been ingested based on the skin movement in the lower part of the neck during a swallow. This, coupled with existing medication adherence systems that detect when medicine is removed from the bottle, can detect a broader range of use-cases with respect to medication adherence., Results: Using Bayesian networks, we were able to correctly classify between chewable vitamins, saliva swallows, medication capsules, speaking, and drinking water, with average precision and recall of 90.17% and 88.9%, respectively. A total of 135 instances were classified from a total of 20 subjects., Conclusion: Our experimental evaluations confirm the accuracy of the piezoelectric necklace for detecting medicine swallows and disambiguating them from related actions. Further studies in real-world conditions are necessary to evaluate the efficacy of the proposed scheme., (Copyright © 2016 Elsevier B.V. All rights reserved.)

Maintaining appropriate levels of food intake and developing regularity in eating habits is crucial to weight loss and the preservation of a healthy lifestyle. Moreover, awareness of eating habits is an important step towards portion control and weight loss. In this paper, we introduce a novel food-intake monitoring system based around a wearable wireless-enabled necklace. The proposed necklace includes an embedded piezoelectric sensor, small Arduino-compatible microcontroller, Bluetooth LE transceiver, and Lithium-Polymer battery. Motion in the throat is captured and transmitted to a mobile application for processing and user guidance. Results from data collected from 30 subjects indicate that it is possible to detect solid and liquid foods, with an F-measure of 0.837 and 0.864, respectively, using a naive Bayes classifier. Furthermore, identification of extraneous motions such as head turns and walking are shown to significantly reduce the false positive rate of swallow detection., (Copyright © 2015 Elsevier Ltd. All rights reserved.)

In this paper, we propose a novel methodology for utilizing disease diagnostic information to predict severity of condition for Congestive Heart Failure (CHF) patients. Our methodology relies on a novel, clustering-based, feature extraction framework using disease diagnostic information. To reduce the dimensionality we identify disease clusters using cooccurence frequencies. We then utilize these clusters as features to predict patient severity of condition. We build our clustering and feature extraction algorithm using the 2012 National Inpatient Sample (NIS), Healthcare Cost and Utilization Project (HCUP) which contains 7 million discharge records and ICD-9-CM codes. The proposed framework is tested on Ronald Reagan UCLA Medical Center Electronic Health Records (EHR) from 3041 patients. We compare our cluster-based feature set with another that incorporates the Charlson comorbidity score as a feature and demonstrate an accuracy improvement of up to 14% in the predictability of the severity of condition.

Current studies have produced a plethora of remote health monitoring (RHM) systems designed to enhance the care of patients with chronic diseases. Many RHM systems are designed to improve patient risk factors for cardiovascular disease, including physiological parameters such as body mass index (BMI) and waist circumference, and lipid profiles such as low density lipoprotein (LDL) and high density lipoprotein (HDL). There are several patient characteristics that could be determining factors for a patient's RHM outcome success, but these characteristics have been largely unidentified. In this paper, we analyze results from an RHM system deployed in a six month Women's Heart Health study of 90 patients, and apply advanced feature selection and machine learning algorithms to identify patients' key baseline contextual features and build effective prediction models that help determine RHM outcome success. We introduce Wanda-CVD, a smartphone-based RHM system designed to help participants with cardiovascular disease risk factors by motivating participants through wireless coaching using feedback and prompts as social support. We analyze key contextual features that secure positive patient outcomes in both physiological parameters and lipid profiles. Results from the Women's Heart Health study show that health threat of heart disease, quality of life, family history, stress factors, social support, and anxiety at baseline all help predict patient RHM outcome success.

Diabetes is the seventh leading cause of death in the United States, but careful symptom monitoring can prevent adverse events. A real-time patient monitoring and feedback system is one of the solutions to help patients with diabetes and their healthcare professionals monitor health-related measurements and provide dynamic feedback. However, data-driven methods to dynamically prioritize and generate tasks are not well investigated in the domain of remote health monitoring. This paper presents a wireless health project (WANDA) that leverages sensor technology and wireless communication to monitor the health status of patients with diabetes. The WANDA dynamic task management function applies data analytics in real-time to discretize continuous features, applying data clustering and association rule mining techniques to manage a sliding window size dynamically and to prioritize required user tasks. The developed algorithm minimizes the number of daily action items required by patients with diabetes using association rules that satisfy a minimum support, confidence and conditional probability thresholds. Each of these tasks maximizes information gain, thereby improving the overall level of patient adherence and satisfaction. Experimental results from applying EM-based clustering and Apriori algorithms show that the developed algorithm can predict further events with higher confidence levels and reduce the number of user tasks by up to 76.19 %.