Mohammad Arif Ul Alam

Automated Functional and Behavioral Health Assessment of Older Adults with Dementia

Dementia is a clinical syndrome of cognitive deficits that involves both memory and functional impairments. While disruptions in cognition is a striking feature of dementia, it is also closely coupled with changes in functional and behavioral health of older adults. In this paper, we investigate the challenges of improving the automatic assessment of dementia, by better exploiting the emerging physiological sensors in conjunction with ambient sensors in a real field environment with IRB approval. We hypothesize that the cognitive health of older individuals can be estimated by tracking their daily activities and mental arousal states. We employ signal processing on wearable sensor data streams (e.g., Electrodermal Activity (EDA), Photoplethysmogram (PPG), accelerometer (ACC)) and machine learning algorithms to assess cognitive impairments and its correlation with functional health decline. To validate our approach, we quantify the score of machine learning, survey and observation based Activities of Daily Living (ADLs) and signal processing based mental arousal state, respectively for functional and behavioral health measures among 17 older adults living in a continuing care retirement community in Baltimore. We compare clinically observed scores with technology guided automated scores using both machine learning and statistical techniques.

CACE: Exploiting Behavioral Interactions for Improved Activity Recognition in Multi-inhabitant Smart Homes

We propose CACE (Constraints And Correlations mining Engine) which investigates the challenges of improving the recognition of complex daily activities in multi-inhabitant smart homes, by better exploiting the spatiotemporal relationships across the activities of different individuals. We first propose and develop a loosely-coupled Hierarchical Dynamic Bayesian Network (HDBN), which both (a) captures the hierarchical inference of complex (macro-activity) contexts from lower-layer microactivity context (postural and improved oral gestural context), and (b) embeds the various types of behavioral correlations and constraints (at both micro-and macro-activity contexts) across the individuals. While this model is rich in terms of accuracy, it is computationally prohibitive, due to the explosive increase in the number of jointly-defined states. To tackle this challenge, we employ data mining to learn behaviorally-driven context correlations in the form of association rules, we then use such rules to prune the state space dramatically. To evaluate our framework, we build a customized smart home system and collected naturalistic multi-inhabitant smart home activities data. The system performance is illustrated with results from real-time system deployment experiences in a smart home environment reveals a radical (max 16 fold) reduction in the computational overhead compared to traditional hybrid classification approaches, as well as an improved activity recognition accuracy of max 95%.

Smart-energy group anomaly based behavioral abnormality detection

Monitoring behavioral abnormality of individuals living independently in their own homes is a key issue for building sustainable healthcare models in smart environments. While most of the efforts have been directed towards building ambient and wearable sensors-assisted activity recognition based behavioral analysis models for remote health monitoring, energy analytics assisted behavioral abnormality prediction have rarely been investigated. In this paper, we propose a data analytic approach that helps detect energy usage anomalies corresponding to the behavioral abnormality of the residents. Our approach relies on detecting everyday appliances usage from smart meter and smart plug data traces in regular activity days and then learning the unique time segment group of each appliances energy consumption. We focus on detecting behavioral anomalies over a set of energy source data points rather than pinpointing individual odd points. We employ hierarchical probabilistic model-based group anomaly detection [7] to interpret the anomalous behavior and therefore, detect potential tendency towards behavioral abnormality. We apply daily activity logs to evaluate our approach using two realworld energy datasets pertaining to staged functional behaviors, and show that it is possible to detect max. 97% of anomalous days with max. 87% of meaningful micro-behavioral abnormal events generating 1.1% of false alarms. However, we show that our detected abnormality can be meaningfully represented to different stakeholders such as caregivers and family members to understand the nature and severity of abnormal human behavior for sustaining better healthcare.

A smart segmentation technique towards improved infrequent non-speech gestural activity recognition model

Infrequent Non-Speech Gestural Activities (IGAs) such as coughing, deglutition and yawning help identify fine-grained physiological symptoms and chronic psychological conditions which are not directly observable from traditional daily activities. We propose a new wearable smart earring which is capable of differentiating IGAs in daily environment with single integrated accelerometer sensor signal processing. Our prior framework, GeSmart, shows significant improvement in IGAs recognition based on smart earring which necessitates users to replace the earring batteries frequently due to its energy hungry requirement (high sampling frequency) towards fine-grained IGAs recognition. In this improved work, we propose a new segmentation technique along with GeSmart which takes the advantages of change-point detection algorithm to segment sensor data streams, feature extraction and classification thus any machine learning technique can perform significantly well in low sampling rate. We also implement a baseline traditional graphical model based gesture recognition techniques and compare their performances with our model in terms of accuracy, energy consumption and degradation of sampling rate scenarios. Experimental results based on real data traces demonstrate that our approach improves the performances significantly compared to previously proposed solutions. We also apply our segmentation technique on two benchmark datasets to prove the superiority of our technique in low sampling rate scenario.

GeSmart: A gestural activity recognition model for predicting behavioral health

To promote independent living for elderly population activity recognition based approaches have been investigated deeply to infer the activities of daily living (ADLs) and instrumental activities of daily living (I-ADLs). Deriving and integrating the gestural activities (such as talking, coughing, and deglutition etc.) along with activity recognition approaches can not only help identify the daily activities or social interaction of the older adults but also provide unique insights into their long-term health care, wellness management and ambulatory conditions. Gestural activities (GAs), in general, help identify fine-grained physiological symptoms and chronic psychological conditions which are not directly observable from traditional activities of daily living. In this paper, we propose GeSmart, an energy efficient wearable smart earring based GA recognition model for detecting a combination of speech and non-speech events. To capture the GAs we propose to use only the accelerometer sensor inside our smart earring due to its energy efficient operations and ubiquitous presence in everyday wearable devices. We present initial results and insights based on a C4.5 classification algorithm to infer the infrequent GAs. Subsequently, we propose a novel change-point detection based hybrid classification method exploiting the emerging patterns in a variety of GAs to detect and infer infrequent GAs. Experimental results based on real data traces collected from 10 users demonstrate that this approach improves the accuracy of GAs classification by over 23%, compared to previously proposed pure classification-based solutions. We also note that the accelerometer sensor based earrings are surprisingly informative and energy efficient (by 2.3 times) for identifying different types of GAs.

Context-aware multi-inhabitant functional and physiological health assessment in smart home environment

Recognizing the human activity, behavior and physiological symptoms in smart home environments is of utmost importance for the functional, physiological and cognitive health assessment of the older adults. Unprecedented data from everyday devices such as smart wristbands, smart ornaments, smartphones, and ambient sensors provide opportunities for activity mining and inference, but pose fundamental research challenges in data processing, physiological feature extraction, activity learning and inference in the presence of multiple inhabitants. In this thesis, we develop micro-activity driven macro-activity recognition approaches while considering the underpinning spatiotemporal constraints and correlations across multiple inhabitants. We design novel signal processing and machine learning algorithms to detect physiological symptoms and infer macro-level activity of the inhabitants, respectively. We combine these activity recognition methodologies along with the physiological health assessment approaches to quantify the functional, behavioral, and cognitive health of the older adults. real-time data collected data from a continuing care retirement community center (IRB #HP-00064387) helped us to evaluate, compare, and benchmark our proposed computational approaches with the clinical tools used extensively for functional and cognitive health assessment.

Single BSN-based multi-label activity recognition

Wearable physical sensor signal processing-based activity recognition has profound impacts on context-aware remote healthcare and cognitive reasoning. Different nodes of Body Sensor Network (BSN) are involved with different contexts that limits single BSN sensor based context recognition models into a sole extreme. On the other hand, multiple physical sensors based context recognition systems are suffered with disrupted BSN network connectivity, multi-modal sensor signal processing complexity, multi-sensor device implantation and fault intolerance. To mitigate these problems, we postulate a sparse-deconvolution aided single BSN sensor-based multi-label physical activity recognition framework. We first consider a single physical sensor device is attached to individuals wrist node i.e., one of the upper extreme body nodes. We hypothesize that the final sensor signals of upper extreme nodes are affected by the lower extreme nodes contexts with an approximately sparse factor (ASF). Based on the hypothesis, we postulate (a) a sparse deconvolution method on the upper extreme node signals to disaggregate ASF and original signals; and design (b) a hybrid classification model to detect both upper extreme (such as, hand waving, hand shaking etc.) and lower extreme (such as, walking, standing, running etc.) activities. We evaluate the performance of our proposed framework with three real-time dataset with distinct characteristics (a real-time collected activity dataset in a controlled lab environment, a real-time smart home system deployed in a retirement community center -(IRB #HP-00064387) and a publicly available dataset) which corroborates a radical improvement in recognizing multi-label human activities.

Unseen Activity Recognitions: A Hierarchical Active Transfer Learning Approach

Human activity recognition (AR) is an essential element for user-centric and context-aware applications. While previous studies showed promising results using various machine learning algorithms, most of them can only recognize the activities that were previously seen in the training data. We investigate the challenges of improving the recognition of unseen daily activities in smart home environment, by better exploiting the hierarchical taxonomy of complex daily activities. We first (a) design a hierarchical representation of complex activity taxonomy in terms of human-readable semantic attributes, and (b) develop a hierarchy of classifiers which incorporates a cluster tree built on the domain knowledge from training samples. Though this model is rich in recognizing complex activities that are previously seen in training data, it is not well versed to recognize unseen complex activities without new training samples. To tackle this challenge, we extend Hierarchical Active Transfer Learning (HATL) approach that exploits semantic attribute cluster structure of complex activities shared between seen (source) and unseen (target) activity domains. Our approach employs transfer and active learning to help label target domain unlabeled data by spawning the most effective queries. We evaluated our approach with two real-time smart home systems (IRB #HP-00064387) which corroborates radical improvements in recognizing unseen complex activities.

Infrequent Non-speech Gestural Activity Recognition Using Smart Jewelry: Challenges and Opportunities for Large-Scale Adaptation

Wearable Body Area Network (BAN) based activity recognition is one of the fastest growing research areas in activity recognition and context reasoning. However, wearable physical sensor based Infrequent Non-Speech Gestural Activity (IGA) recognition is not well studied problem because IGAs are not directly observable from BAN sensor devices. Due to the recent proliferation of smart jewelries capable of monitoring locomotive and physiological signals from certain specific human body positions which are currently hitherto impossible to measure by traditional fitness and smart wristwatch devices opens up unprecedented research and development opportunities in anatomical gestural activity recognition. Inspired by this, we propose a new wearable smart earring based framework which is capable of differentiating IGAs in a daily environment with a single integrated accelerometer sensor. The natural gestures associated with the first portion of the human alimentary canal, i.e., human mouth can broadly be categorized in two types; frequent (talking, silence etc.) or infrequent (coughing, deglutition, yawning) gestures. Infrequent Gestural Activities (IGAs) help create an abrupt but distinct change in accelerometer sensor signal streams of an earring pertaining to specific activities. Mining and classifying the abrupt changes in sensor signal streams require high sampling frequency which in turn depletes the limited battery life of any smart ornaments. Extending the battery life of smartened designer jewelry requires probing those devices less which in turn prohibits of achieving high precision and recall for non-frequent gestural activity discovery and recognition. In this book chapter, we propose a novel data segmentation technique that harnesses the power of change-point detection algorithm to detect and quantify any abrupt changes in sensor data streams of smart earrings. This helps to distinguish between frequent and infrequent gestural acclivities at a high precision with a low sampling frequency, energy, and computational overhead. Experimental evaluation on one real-time and two publicly available benchmark datasets attests the scalability and adaptation of our techniques for both IGAs and postural activities in large-scale participatory sensing health applications.