CONTINUOUS STRESS DETECTION USING A WRIST DEVICE -IN LABORATORY AND - - PowerPoint PPT Presentation

▶

Apr 11, 2023 9 likes •189 views

CONTINUOUS STRESS DETECTION USING A WRIST DEVICE -IN LABORATORY AND REAL LIFE- Martin Gjoreski, Hristijan Gjoreski, Mitja Lutrek, Matja Gams http://www.fit4work-aal.eu/index.html Workshop on Mental Health: Sensing and Intervention, UbiComp

SLIDE 1

CONTINUOUS STRESS DETECTION USING A WRIST DEVICE

IN LABORATORY AND REAL LIFE-

Martin Gjoreski, Hristijan Gjoreski, Mitja Luštrek, Matjaž Gams

13 September 2016 1

http://www.fit4work-aal.eu/index.html

Workshop on Mental Health: Sensing and Intervention, UbiComp

SLIDE 2

Motivation

2 26 September 2016

SLIDE 3

Motivation

Chronical stress:

raised blood pressure
bad sleep
infections
decreased performance
slower recovery

EU, work-related stress costs €20 billion a year.

26 September 2016

SLIDE 4

Definition

■ Definition (Ice and James) - “Stress is considered a process by which a stimulus elicits an emotional, behavioral and/or physiological response, which is conditioned by an individual’s personal, biological and cultural context”.

4 26 September 2016

SLIDE 5

The method

Activity recognizer Stress Activity

Context-based stress detector

Real-life stress monitoring Bio data

Acc. data

Laboratory

Real-life method

Lab stress detector

Aggregate predictions

Simple contexts (e.g., date-time)

Real-life data

26 September 2016 5

SLIDE 6

The method

Activity recognizer Stress Activity

Context-based stress detector

Real-life stress monitoring Bio data

Acc. data

Laboratory

Real-life method Aggregate predictions

Simple contexts (e.g., date-time)

Real-life data

Lab stress detector

26 September 2016 6

SLIDE 7

Lab stress detector – Lab data

■ Stress inducing – math task under time and evaluation pressure (200 EUR reward for motivation) ■ 21 participants x 75 minutes of data per sensor (Heart Rate, Skin

Temperature, Blood Volume Pulse, Inter-beat-interval, Electrodermal Activity, Acceleration)

7 26 September 2016

SLIDE 8

Lab stress detector – Lab data

Timing Anxiety score Before 10.95 After Easy 13.33 After Medium 14.05 After Hard (End) 13.81 Labelled Data # Participants 21 Age Mean 28+-4 No Stress 840 minutes Low Stress 356 minutes High Stress 368 minutes

EDA BVP ACC HR ST

10.95 13.33 14.05 13.81

26 September 2016

SLIDE 9

FILTERING

Lab stress detector –The method

SENSOR DATA SEGMENTATION FEATURE EXTRACTION MODEL LEARNING SAMPLE DATA EVALUATION FEATURE SELECTION LABORATORY STRESS DETECTOR

9 26 September 2016

SLIDE 10

Lab stress detector

The method

Stress Activity

Context-based stress detector

Real-life stress monitoring Bio data

Acc. data

Laboratory

Real-life method Aggregate predictions

Simple contexts (e.g., date-time)

Real-life data

Activity recognizer

26 September 2016 10

SLIDE 11

Activity recognizer Lab stress detector

The method

Stress Activity

Context-based stress detector

Real-life stress monitoring Laboratory

Real-life method Aggregate predictions

Simple contexts (e.g., date-time) Bio data

Acc. data

Real-life data

26 September 2016 11

SLIDE 12

Context-based stress detector – Real-life data

■ No constraints at all ■ Smartphone application

– For assessing stress levels at random periods of the day – Logger for stressful events Real-life labeled data # Participants 5 Age Mean 28+-4.3 No Stress 1216 hours Low Stress 70 hours High Stress 41 hours

Stress event

26 September 2016

SLIDE 13

Activity recognizer Lab stress detector

The method

Context-based stress detector

Real-life stress monitoring Laboratory

Real-life method

Simple contexts (e.g., date-time) Bio data

Acc. data

Real-life data

Stress Activity

Aggregate predictions

26 September 2016 13

SLIDE 14

Activity recognizer Lab stress detector

The method

Real-life stress monitoring Laboratory Bio data

Acc. data

Real-life data

Stress Activity

Aggregate predictions Context-based stress detector Real-life method

Simple contexts (e.g., date-time)

26 September 2016 14

SLIDE 15

Activity recognizer Lab stress detector

The method

Laboratory Bio data

Acc. data

Real-life data

Stress Activity

Aggregate predictions Context-based stress detector Real-life method

Simple contexts (e.g., date-time) Real-life stress monitoring

26 September 2016 15

SLIDE 16

Context-based stress detector – Experiments (2)

Confusion matrices and evaluation metrics for No-Context vs. With-Context classification. Each number represents an instance/event. No-Context With Context No Stress Stress No Stress Stress No Stress 3308 1630 4932 6 Stress 34 125 47 112 F1 score 0.80 0.13 0.99 + 0.19 0.81 + 0.68

16 26 September 2016

SLIDE 17

Limitations and Future work

■ Limitations

– Sample size – Age structure

■ Future work

– More real-life data – Richer context – Personalization – Energy efficiency

http://www.fit4work-aal.eu/index.html

26 September 2016

SLIDE 18

Conclusions

■ We addressed the problem of stress detection in real-life.

■ Data-preprocessing, feature extraction and feature selection methods. ■ Machine learning methods for stress detection in a constrained environments. ■ Context-based method for stress detection in an unconstrained environments.

■ The key idea is to use context information. ■ Evaluated the proposed method on a real-life data.

– the presented context-based method for stress detection detects (recalls) 70% of the stress events with a precision of 95%.

18 26 September 2016