Skip to main content

Table 1 Definition of inputs and outputs of the H-FIS

From: Improving activity recognition using a wearable barometric pressure sensor in mobility-impaired stroke patients

Name

Description

Inputs

  PrevAct

Previous activity: the activity preceding the activity being evaluated by the algorithm

  CurrAct

Current activity: the activity being evaluated by the algorithm

  NextAct

Next activity: the activity following the activity being evaluated by the algorithm

  Transition

Transition detection probability: Computed through a logistic regression model, it provides a continuous value (PTr) from 0 to 1 representing the probability for a transition to be respectively “Not Detected” and “Detected”

Transition type

Transition type probability: Computed through a logistic regression model, it provides a continuous value (PType) from 0 to 1 representing the probability for a transition to be respectively of a type “Sit-to-stand” and “Stand-to-Sit”

  PrevDur

Duration of the previous activity being processed

  CurDur

Duration of the current activity being processed

  NextDur

Duration of the activity following the activity being processed

  AltitudeChange

Altitude change corresponds to the change in elevation around the transition time. It is computed on the barometric pressure signal through transition feature extraction algorithm [28]

  AltitudeIQR

Altitude Inter-Quartile Range is computed over the duration of the activity on the altitude signal (derived from the barometric pressure signal)

Outputs

  Event activity

Event activity represents the output of FIS I (this is the current activity recognized in this stage of classification)

  Behaviour activity

Behavior Activity represents the output of FIS II and thus the final output of the H-FIS (this is the current activity recognized at the end of second stage of classification)