No abstract

Despite the fact that deep neural networks are powerful models and achieve appealing results on many tasks, they are too large to be deployed on edge devices like smartphones or embedded sensor nodes. There have been efforts to compress these networks, and a popular method is knowledge distillation, where a large (teacher) pre-trained network is used to train a smaller (student) network. However, in this paper, we show that the student network performance degrades when the gap between student and teacher is large. Given a fixed student network, one cannot employ an arbitrarily large teacher, or in other words, a teacher can effectively transfer its knowledge to students up to a certain size, not smaller. To alleviate this shortcoming, we introduce multi-step knowledge distillation, which employs an intermediate-sized network (teacher assistant) to bridge the gap between the student and the teacher. Moreover, we study the effect of teacher assistant size and extend the framework to multi-step distillation. Theoretical analysis and extensive experiments on CIFAR-10,100 and ImageNet datasets and on CNN and ResNet architectures substantiate the effectiveness of our proposed approach.

Background The use of remote monitoring systems (RMSs) in healthcare has grown exponentially and has improved the accessibility to and ability of patients to engage in treatment intensification. However, research describing the effects of RMSs on activation, self-care, and quality of life (QOL) in older patients with heart failure (HF) is limited. Objective The aim of this study was to compare the effects of a 3-month RMS intervention on activation, self-care, and QOL of older patients versus a reference group matched on age, gender, race, and functional status (ie, New York Heart Association classification) who received standard discharge instructions after an acute episode of HF exacerbation requiring hospitalization. Methods A total of 21 patients (mean age, 72.7 ± 8.9 years; range, 58–83 years; 52.4% women) provided consent and were trained to measure their weight, blood pressure, and heart rate at home with an RMS device and transmit this information every day for 3 months to a centralized information system. The system gathered all data and dispatched alerts when certain clinical conditions were met. Results The baseline sociodemographic and clinical characteristics of the 2 groups were comparable. Over time, participants in the RMS group showed greater improvements in activation, self-care, and QOL compared with their counterparts. Data showed moderately strong associations between increased activation, self-care, and QOL. Conclusion Our preliminary data show that the use of an RMS is feasible and effective in promoting activation, self-care, and QOL. A larger-scale randomized clinical trial is warranted to show that the RMS is a new and effective method for improving clinical management of older adults with chronic HF.

The evaluation of the postural control system (PCS) has applications in rehabilitation, sports medicine, gait analysis, fall detection, and diagnosis of many diseases associated with a reduction in balance ability. Standing involves significant muscle use to maintain balance, making standing balance a good indicator of the health of the PCS. Inertial sensor systems have been used to quantify standing balance by assessing displacement of the center of mass, resulting in several standardized measures. Electromyogram (EMG) sensors directly measure the muscle control signals. Despite strong evidence of the potential of muscle activity for balance evaluation, less study has been done on extracting unique features from EMG data that express balance abnormalities. In this paper, we present machine learning and statistical techniques to extract parameters from EMG sensors placed on the tibialis anterior and gastrocnemius muscles, which show a strong correlation to the standard parameters extracted from accelerometer data. This novel interpretation of the neuromuscular system provides a unique method of assessing human balance based on EMG signals. In order to verify the effectiveness of the introduced features in measuring postural sway, we conduct several classification tests that operate on the EMG features and predict significance of different balance measures.

Diet and physical activity are known as important lifestyle factors in self-management and prevention of many chronic diseases. Mobile sensors such as accelerometers have been used to measure physical activity or detect eating time. In many intervention studies, however, stringent monitoring of overall dietary composition and energy intake is needed. Currently, such a monitoring relies on self-reported data by either entering text or taking an image that represents food intake. These approaches suffer from limitations such as low adherence in technology adoption and time sensitivity to the diet intake context. In order to address these limitations, we introduce development and validation of Speech2Health, a voice-based mobile nutrition monitoring system that devises speech processing, natural language processing (NLP), and text mining techniques in a unified platform to facilitate nutrition monitoring. After converting the spoken data to text, nutrition-specific data are identified within the text using an NLP-based approach that combines standard NLP with our introduced pattern mapping technique. We then develop a tiered matching algorithm to search the food name in our nutrition database and accurately compute calorie intake values. We evaluate Speech2Health using real data collected with 30 participants. Our experimental results show that Speech2Health achieves an accuracy of 92.2% in computing calorie intake. Furthermore, our user study demonstrates that Speech2Health achieves significantly higher scores on technology adoption metrics compared to text-based and image-based nutrition monitoring. Our research demonstrates that new sensor modalities such as voice can be used either standalone or as a complementary source of information to existing modalities to improve the accuracy and acceptability of mobile health technologies for dietary composition monitoring.

Wearable sensory devices are becoming the enabling technology for many applications in healthcare and well-being, where computational elements are tightly coupled with the human body to monitor specific events about their subjects. Classification algorithms are the most commonly used machine learning modules that detect events of interest in these systems. The use of accurate and resource-efficient classification algorithms is of key importance because wearable nodes operate on limited resources on one hand and intend to recognize critical events (e.g., falls) on the other hand. These algorithms are used to map statistical features extracted from physiological signals onto different states such as health status of a patient or type of activity performed by a subject. Conventionally selected features may lead to rapid battery depletion, mainly due to the absence of computing complexity criterion while selecting prominent features. In this paper, we introduce the notion of power-aware feature selection, which aims at minimizing energy consumption of the data processing for classification applications such as action recognition. Our approach takes into consideration the energy cost of individual features that are calculated in real-time. A graph model is introduced to represent correlation and computing complexity of the features. The problem is formulated using integer programming and a greedy approximation is presented to select the features in a power-efficient manner. Experimental results on thirty channels of activity data collected from real subjects demonstrate that our approach can significantly reduce energy consumption of the computing module, resulting in more than 30% energy savings while achieving 96.7% classification accuracy.

BackgroundAs commercially available activity trackers are being utilized in clinical trials, the research community remains uncertain about reliability of the trackers, particularly in studies that involve walking aids and low-intensity activities. While these trackers have been tested for reliability during walking and running activities, there has been limited research on validating them during low-intensity activities and walking with assistive tools.ObjectiveThe aim of this study was to (1) determine the accuracy of 3 Fitbit devices (ie, Zip, One, and Flex) at different wearing positions (ie, pants pocket, chest, and wrist) during walking at 3 different speeds, 2.5, 5, and 8 km/h, performed by healthy adults on a treadmill; (2) determine the accuracy of the mentioned trackers worn at different sites during activities of daily living; and (3) examine whether intensity of physical activity (PA) impacts the choice of optimal wearing site of the tracker.MethodsWe recruited 15 healthy young adults to perform 6 PAs while wearing 3 Fitbit devices (ie, Zip, One, and Flex) on their chest, pants pocket, and wrist. The activities include walking at 2.5, 5, and 8 km/h, pushing a shopping cart, walking with aid of a walker, and eating while sitting. We compared the number of steps counted by each tracker with gold standard numbers. We performed multiple statistical analyses to compute descriptive statistics (ie, ANOVA test), intraclass correlation coefficient (ICC), mean absolute error rate, and correlation by comparing the tracker-recorded data with that of the gold standard.ResultsAll the 3 trackers demonstrated good-to-excellent (ICC>0.75) correlation with the gold standard step counts during treadmill experiments. The correlation was poor (ICC<0.60), and the error rate was significantly higher in walker experiment compared to other activities. There was no significant difference between the trackers and the gold standard in the shopping cart experiment. The wrist worn tracker, Flex, counted several steps when eating (P<.01). The chest tracker was identified as the most promising site to capture steps in more intense activities, while the wrist was the optimal wearing site in less intense activities.ConclusionsThis feasibility study focused on 6 PAs and demonstrated that Fitbit trackers were most accurate when walking on a treadmill and least accurate during walking with a walking aid and for low-intensity activities. This may suggest excluding participants with assistive devices from studies that focus on PA interventions using commercially available trackers. This study also indicates that the wearing site of the tracker is an important factor impacting the accuracy performance. A larger scale study with a more diverse population, various activity tracker vendors, and a larger activity set are warranted to generalize our results.

No abstract