Database Reference
In-Depth Information
different information fusion schemes. The authors use selective fusion
of detection results exploiting independent sources of error to filter out
false positives and obtain an event classification in the same step, and
achieve highly accurate activity recognition.
Recent work [85] has also focused on the use of body sensors for patient
authentication. Credential based authentication methods (e.g., pass-
words, certificates) are not well-suited for remote healthcare as these may
be compromised. One-time authentication using credentials or trait-
based biometrics (e.g., face, fingerprints, iris) do not cover the entire
monitoring period and may lead to unauthorized post-authentication
use in some situations. Recent studies have shown that the human elec-
trocardiogram (ECG) exhibits unique patterns that can be used to dis-
criminate individuals. However, perturbations of the ECG signal due
to physical activity in real-world situations can lead to authentication
failures. The authors in [85] build an activity-aware biometric authenti-
cation system that combines ECG information with accelerometer data
to handle the variability that arises from physical activity. The authors
use the SHIMMER [86] sensing platform (with an integrated 3-axis ac-
celerometer) developed by Intel Digital Health Advanced Technology
Group to combine the motion and activity data with the ECG signal
using a direct cable to a commercially available Polar WearLink Plus
ECG chest strap. The sensor data is transmitted via Bluetooth device
to a commputer running the BioMOBIOUS software for analysis. The
mining uses different types of feature cleaning and preprocessing (beat-
based linear interpolation)combined with K-Nearest Neighbor (KNN)
and Bayesian network(BN) classification to obtain accurate user authen-
tication under different activity levels.
The MIThril [87] project is focused on developing a next-generation
wearable sensor research platform. The project includes the develop-
ment and prototyping of new techniques of human-computer interaction
for body-worn applications, through the application of human factors,
machine learning, hardware engineering, and software engineering. The
MIThril project also involves research into constructing a new computing
environment and developing prototype applications for health, communi-
cations, and just-in-time information delivery. The MIThril LiveNet [88]
is a flexible distributed mobile platform that can be deployed for a va-
riety of proactive healthcare applications. The LiveNet system allows
people to receive real-time feedback from their continuously monitored
and analyzed health state, as well as communicate health information
with care-givers and other members of the social network of an indi-
vidual for support and interaction. Key components of this system in-
clude a PDA-centric mobile wearable platform, the Enchantment soft-
Search WWH ::




Custom Search