Healthcare System Examples
For class projects in personal healthcare systems, students will
use the healthcare SIS testbed and the experimental
system with temperature sensor and Kinect sensor
to develop monitor components.
T1 Temperature sensor
T1 Single sensor (for class project only)
Improve the GUI so that it can specify various parameters
for temp sensor: sampling_rate, start_Month/Day/Hour/Min,
end_Month/Day/Hour/Min, high/low_threshold, reporting
options (all or alarm_only) and so on. Activate temp
component once parameters are set. Run scenario to
demo it works.
The temperature sensor will store data in a
speciic temperature reading file.
The T1 component access this file to processing temperature information.
T2 Multiple sensors (for class project Milestone 3)
T2 includes T1. Based upon T1, add multi-sensor processing
capabilities. For example, if T2 component receives
message from Kinect component that a person is lying
on T2 decides floor. the low temp threshold shoud
be raised, so that the temp component will be
triggered at a higher low temp threshold.
K1 Kinect sensor
K1 Single sensor (for class project only)
Improve the GUI so that it can specify various parameters
for Kinect. The recent report and codes for Kinect
FallDetection component is available. Improve the
K1 so that it can detect a person's fall as well as
a person lying motionless on the floor.
Fall Detection requirements
-
InputProcessor will process input video stream from Kinect at a specified speed
of N frame per second.
-
Uploader will upload images and send alert messages to the EMS through the Web Server.
-
Kinect images will be processed by PatientFall to detect the
fall of a patient.
After the fall, if the patient is motionless for a specified Na frames,
an alert message is sent to the EMS by Uploader.
-
Kinect images will be processed by PatNoMovement to detect the
lack of movement of a patient.
If the patient does not move for a specified Ns frames,
an alert message is sent to the EMS by Uploader.
-
GUI allows the user (EMS staff) to specify various
parameters such as N (frame rate per second),
Na (number of frames for absence to trigger an alarm),
and
Ns (number of frames for no movement to trigger an alarm).
K2 Multiple sensors (for class project Milestone 3)
K2 includes K1 and also supports multi-sensor processing.
If other sensors detects a person has high blood
pressure or irregular heart beat, we need to pay more
attention if he falls. When SIS sends messages to K2,
K2 will replace fall-detection algorithm by a more
sensitve one, i.e., K2 is a super-component that
enumerates and selects a more sensitive algorithm.
Hx Other healthcare sensors
Blood pressure and EKG sensors are available for
single sensor or multiple sensor projects.
The devil is in the details
If you are working on T1 or T2, please
read this document.
A completed T1 project,
its messages and
readme file are provided
here for your reference.
If you are working on K1 or K2 or other related projects, please
read this document.
The Kinect Alert message (Type 44)
detects the fall of a senior resident.
The Kinect data reading messages (Type 43) such as 43a.xml,
43b.xml
and 43c.xml
provide a sequence of point coordinates that refer
to several parts of human body such as head,
shoulder, elbow, foot and so on, where the position of each human-body-part
is represented as (x,y,z).
The KinectSensor Component
inputs raw Kinect sensor data and outputs Kinect data reading messages.
The Kinect Monitor K1
analyzes the Kinect data reading messages to detect certain events and
send out Kinect alert messages.
The Kinect Monitor K2 can then be built on top of K1.