Mobile Big Data by Xiang Cheng & Luoyang Fang & Liuqing Yang & Shuguang Cui

Author:Xiang Cheng & Luoyang Fang & Liuqing Yang & Shuguang Cui
Language: eng
Format: epub
ISBN: 9783319961163
Publisher: Springer International Publishing

5.1.2 Pervasive Health Computing

With the multi-sensor data from wearable devices or smart phones, mobile health (mHealth) is a promising application. Such data from sensors can passively and unobtrusively collect the necessary information for health monitoring. In addition, smart phones with rich connectivity can provide a platform for active data collection in many scenarios, such as facial expression capturing with phone cameras and patient audio clip recording by microphones.

Actually, physical activity monitoring is an intuitive application of health monitoring [17] with such a rich set of sensors. For instance, fall detection coupled with an alert system [18] could be implemented to detect the fall and alert the authorities at the same time. In [19], Wu et al. collected the data of the accelerometer and the gyroscope for 16 participants on 13 activities, such as sitting, walking, jogging, and going upstairs and downstairs. Overall, the results of monitoring are claimed to achieve very good accuracies: 52.3–79.4% for up-and-down-stair walking, 91.7% for jogging, 90.1–94.1% for walking on the ground, and 100% for sitting.

In fact, mobile sensors equipped in smart devices can help doctors monitor their patients remotely and frequently. The sensing results will further help doctors customize the personalized medical treatment for each patient. In [20], Sharma et al. proposed a framework to help with monitoring Parkinson patients via both active and passive approaches, to understand the daily activities and manage the complex medication regimens personalized to individual needs. In particular, passive monitoring relies on accelerometers as well as gyroscopes to monitor motor aspects of Parkinson patients, such as walking, falls, balance, tremor, and so on, while active monitoring relies on the collection of contextual data, such as speech, facial tremors, etc., by interacting with patients.

Although physical activity monitoring can be considered as a modernized integration of biomedical sensors into personal mobile communication devices, the mental health monitoring hinges more upon context sensing and the development of emotion learning [21]. Based on where we have been, with whom we communicate, what applications we use, and how we use our mobile devices, various learning algorithms can be developed to exploit these mobile phone sensor values. These learning models can be adapted to predict a mobile user’s moods, emotions, cognitive/motivational states, activities, environmental contexts, and social contexts.

In [22], LiKamWa et al. proposed a mood monitor based on the logged data collected from 32 participants over 2 months. The authors showed that by analyzing the communication history and application usage patterns, a user’s daily mood curve could be statistically inferred with an initial accuracy of 66%, which gradually improves to an accuracy of 93% after a 2-month personalized training period. High-quality cameras equipped in smart phones [21] could easily capture our facial expressions indicating emotions and moods, and the new development on emotion learning techniques will further facilitate mobile emotion monitoring. In [23], the smile intensity mined from facial expressions was studied to help machines understand the emotion of human beings, by tracking the changes of facial muscles leading to a specific expression.

Besides psychological monitoring and

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.