new configuration. The separation of the virtual sensors as independent subscribers
ensures that the modification or the elimination of a device does not affect the whole
system.
• Data management configuration
Added to the devices configuration possibility, our system integrates an automation
algorithm for the calibration of the data storage and the score calculation frequency.
For each score level, the maximum value is defined as score level limit. The
configuration manager reads frequently the calculated score of each vital sign from the
local database and compare it with the corresponding predefined limits.
If it captures successive abnormal values, the frequency of the collection of this
data is automatically increased to be able to monitor the data concerned with more
precision. The new frequency value is saved in the local database and published as a
message to the corresponding virtual sensor.
When the measurements become more stable and less than the predefined limits,
the administrator can decide to change this frequency.
The values of required frequencies are defined depending on the medical experts’
requirements (Fig. 3).
The self-adaptivity process is illustrated in the following algorithm:
Algorithm 1 Self-adaptivity algorithm
Inputs: (SC) NEWS2 calculated score, (Freq) Corresponding frequency, (VST) Virtual
Sensor Topic
Variables: (Score0Limit, Score1Limit) NEWS2 scores limits
Output: NewFreq (new data collection frequency).
0: begin
1: If ((Freq == FreqLevel1) and (SC ==0)) then
2:
Score0Limit= GetScoreLimit (VST, 0)
3:
If (verifEgalityOfLast2Values (VST, SC, Score0Limit)
4:
== false) then
5:
NewFreq = FreqLevel2
6:
End if
7: Else if ((Freq == 1min) and ((SC == 1) or (SC == 2))) then
8:
Score1Limit = GetScoreLimit (VST, 1)
9:
If (verifEgalityOfLast3Values(VST, SC, Score1Limit)
10:
== false) then
11:
NewFreq = FreqLevel3
12:
End if
13: End if
14: Publish (VST, NewFreq)
15: end
Fig. 3. Self-adaptative configuration algorithm.
Self-adaptative Early Warning Scoring System for Smart Hospital
23
ensures that the modification or the elimination of a device does not affect the whole
system.
• Data management configuration
Added to the devices configuration possibility, our system integrates an automation
algorithm for the calibration of the data storage and the score calculation frequency.
For each score level, the maximum value is defined as score level limit. The
configuration manager reads frequently the calculated score of each vital sign from the
local database and compare it with the corresponding predefined limits.
If it captures successive abnormal values, the frequency of the collection of this
data is automatically increased to be able to monitor the data concerned with more
precision. The new frequency value is saved in the local database and published as a
message to the corresponding virtual sensor.
When the measurements become more stable and less than the predefined limits,
the administrator can decide to change this frequency.
The values of required frequencies are defined depending on the medical experts’
requirements (Fig. 3).
The self-adaptivity process is illustrated in the following algorithm:
Algorithm 1 Self-adaptivity algorithm
Inputs: (SC) NEWS2 calculated score, (Freq) Corresponding frequency, (VST) Virtual
Sensor Topic
Variables: (Score0Limit, Score1Limit) NEWS2 scores limits
Output: NewFreq (new data collection frequency).
0: begin
1: If ((Freq == FreqLevel1) and (SC ==0)) then
2:
Score0Limit= GetScoreLimit (VST, 0)
3:
If (verifEgalityOfLast2Values (VST, SC, Score0Limit)
4:
== false) then
5:
NewFreq = FreqLevel2
6:
End if
7: Else if ((Freq == 1min) and ((SC == 1) or (SC == 2))) then
8:
Score1Limit = GetScoreLimit (VST, 1)
9:
If (verifEgalityOfLast3Values(VST, SC, Score1Limit)
10:
== false) then
11:
NewFreq = FreqLevel3
12:
End if
13: End if
14: Publish (VST, NewFreq)
15: end
Fig. 3. Self-adaptative configuration algorithm.
Self-adaptative Early Warning Scoring System for Smart Hospital
23