Table 1 Description and categorization of the feature set.

From: Passive detection of COVID-19 with wearable sensors and explainable machine learning algorithms

Feature category

Feature description

Total COVID-19 individual tests

Number of individuals

Fitbit users [%]

Available days median [IQR]

Symptoms features

Fatigue

1149

1091

Headache

1104

1061

Difficulty breathing

208

206

Diarrhea or vomiting

378

368

Decrease in taste smell

247

247

Cough

892

859

Fever chills or sweating

660

645

Congestion Or runny nose

1152

1097

Neck pain

409

396

Body aches

823

795

Sore throat

973

922

Stomach ache

312

302

Sensor features (activity)

Total number of daily steps

9348

6983

82%

673 [507–707]

Total daily distance traveled on foot

9281

6938

82%

676 [554–708]

Calories burned from periods above sedentary level

7629

5679

100%

676 [554–708]

Calories burned inclusive of BMR

8279

6142

100%

676 [554–708]

Minutes spent fairly active

7291

5450

100%

676 [554–708]

Minutes spent sedentary

8196

6082

100%

676 [554–708]

Minutes spent very active

7171

5370

100%

676 [554–708]

Minutes spent lightly active

7628

5678

100%

676 [554–708]

Sensor features (Heart)

Daily resting heart rate

9105

6810

81%

492 [341–653]

Maximum daily heart rate variability

1812

1407

0%

363 [185–499]

Minimum daily heart rate variability

1812

1407

0%

363 [185–499]

Average daily heart rate variability

1812

1407

0%

363 [185–499]

Sensor features (Sleep)

Total daily sleep time

7473

5603

94%

445 [240–629]

Total daily time spent in bed

7473

5603

94%

445 [240–629]

Sleep efficiency of the main sleep

7473

5603

94%

445 [240–629]

Sleep time of the main sleep

7473

5603

94%

445 [240–629]

Anthropometrics

Body mass index

8478

6303

Self-reported height

1673

1277

Body weight

9240

6896

Body fat percentage

4594

3431

Basal metabolic rate (BMR) only calories

8279

6142

Demographic features

Self-reported gender

10,494

7853

Age at the time of test

10,479

7841

  1. The set of features for each category is reported in the table. The same individual may report multiple tests. The total number of individual COVID-19 tests is the number of tests with the corresponding feature value available for the analysis, after exclusion criteria. The number of corresponding unique individuals is also reported. The available days for Sensor Features represent the median number of days available for all the participants (IQR reported in brackets).