Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)
-
Upload
marco-altini -
Category
Technology
-
view
222 -
download
0
Transcript of Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)
![Page 1: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/1.jpg)
Variable-length accelerometer features and electromyography to improve accuracy of fetal kicks detection during pregnancy using a single wearable device
BHI, 2017 bloomlife.com
Marco Altini, Elisa Rossetti, Michiel Rooijakkers, Julien Penders, Dorien Lanssens, Lars Grieten and Wilfried Gyselaers
![Page 2: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/2.jpg)
2 |
FETAL MOVEMENT
BHI, 2017 bloomlife.com
Monitoring fetal movement during pregnancy is the mostpractical and widespread method to assess fetal wellbeing, one of the most important and complex tasks of modern obstetrics.!!As birth outcomes are strongly linked to the development of fetal conditions during pregnancy, several techniques have been developed to monitor fetal movement up to date!
![Page 3: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/3.jpg)
3 |
CURRENT CLINICAL PRACTICE
BHI, 2017 bloomlife.com
Ultrasound: relies on high frequency sound!waves being used to generate an image of the fetus and!can be used only for a limited amount of time due to!safety concerns. Require hospital stays or trained personnel.!!Continuous cardiotocography: require cumbersome infrastructure and hospital visits, also involving trained personnel to set up the device and process the produced information.!
![Page 4: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/4.jpg)
4 |
CURRENT CLINICAL PRACTICE
BHI, 2017 bloomlife.com
Ultrasound: relies on high frequency sound!waves being used to generate an image of the fetus and!can be used only for a limited amount of time due to!safety concerns. Require hospital stays or trained personnel.!!Continuous cardiotocography: require cumbersome infrastructure and hospital visits, also involving trained personnel to set up the device and process the produced information.!
-> Only sporadic checks in the hospital environment
![Page 5: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/5.jpg)
5 |
NEW PASSIVE SOLUTIONS
BHI, 2017 bloomlife.com
Accelerometers: Most studies to date involved one single accelerometer placed on the abdomen and reported rather low sensitivity and specificity.!!Other researchers added a reference accelerometerwith the rationale that by monitoring maternal movement!artifacts using an accelerometer placed outside of the abdominal area, fetal movement should be separable from!maternal movement and therefore detected more accurately. !
![Page 6: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/6.jpg)
6 |
NEW PASSIVE SOLUTIONS: REFERENCE ACCELEROMETER
BHI, 2017 bloomlife.com
0
100
200
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
sen
sor 1
labelnothingkick
Multi − data sample
0
50
100
150
200
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
sen
sor 3
0
10
20
30
40
50
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
sen
sor 6
Reference accelerometer on the back
Fetal movements do not appear on the reference accelerometer, while maternal movements typically do!
![Page 7: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/7.jpg)
7 |
NEW PASSIVE SOLUTIONS
BHI, 2017 bloomlife.com
-> Promising results, still limited practical applicability
![Page 8: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/8.jpg)
8 |
SINGLE SENSOR
BHI, 2017 bloomlife.com
Performance: consistently lower with respect to multiple sensors and reference accelerometers outside of the abdomen area. Higher false positives (harder to discriminate between maternal movements / artifacts and fetal movements)!! !
How do we reduce false positives?
![Page 9: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/9.jpg)
9 |
VARIABLE-LENGTH ACCELEROMETER FEATURES AND PHYSIOLOGICAL DATA
BHI, 2017 bloomlife.com
The proposed techniques aim at reducing false positives by!providing more contextual information related to maternal!movement while still using a single wearable device to cope!with the absence of a reference accelerometer or a more!obtrusive system.!!To account for different dynamics in maternal and fetal movement, we computed features over two time windows of 0.5 and 4 seconds. The rationale is that short fetal movements should be averaged out over longer time windows but captured over short ones, while maternal movements should appear over windows of both durations.!
![Page 10: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/10.jpg)
0.0
0.1
0.2
0.3
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
labelnothingkick
Single sensor, short time window
0.0
0.1
0.2
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
Single sensor, long time window
0
100
200
300
0 5 10 15Timestamp (minutes)
EHG
inte
nsity
Single sensor, EHG data
10 |
VARIABLE-LENGTH ACCELEROMETER FEATURES AND PHYSIOLOGICAL DATA
BHI, 2017 bloomlife.com
![Page 11: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/11.jpg)
0.0
0.1
0.2
0.3
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
labelnothingkick
Single sensor, short time window
0.0
0.1
0.2
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
Single sensor, long time window
0
100
200
300
0 5 10 15Timestamp (minutes)
EHG
inte
nsity
Single sensor, EHG data
11 |
VARIABLE-LENGTH ACCELEROMETER FEATURES AND PHYSIOLOGICAL DATA
BHI, 2017 bloomlife.com
Maternal movements appear on both traces!
![Page 12: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/12.jpg)
0.0
0.1
0.2
0.3
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
labelnothingkick
Single sensor, short time window
0.0
0.1
0.2
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
Single sensor, long time window
0
100
200
300
0 5 10 15Timestamp (minutes)
EHG
inte
nsity
Single sensor, EHG data
12 |
VARIABLE-LENGTH ACCELEROMETER FEATURES AND PHYSIOLOGICAL DATA
BHI, 2017 bloomlife.com
Fetal movements appear on the short window trace only!
![Page 13: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/13.jpg)
0.0
0.1
0.2
0.3
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
labelnothingkick
Single sensor, short time window
0.0
0.1
0.2
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
Single sensor, long time window
0
100
200
300
0 5 10 15Timestamp (minutes)
EHG
inte
nsity
Single sensor, EHG data
0.0
0.1
0.2
0.3
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
labelnothingkick
Single sensor, short time window
0.0
0.1
0.2
0 5 10 15Timestamp (minutes)
Mot
ion
inte
nsity
Single sensor, long time window
0
100
200
300
0 5 10 15Timestamp (minutes)
EHG
inte
nsity
Single sensor, EHG data
13 |
VARIABLE-LENGTH ACCELEROMETER FEATURES AND PHYSIOLOGICAL DATA
BHI, 2017 bloomlife.com
Maternal movements are more likely to trigger EMG activity!
![Page 14: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/14.jpg)
14 |
STUDY DESIGN
BHI, 2017 bloomlife.com
Twenty-two recordings of about 60 minutes duration were collected from 22 pregnant women at different gestational ages during pregnancy, all from week 30 onwards. !!Fetal movements ranged between 0 for inactive babies to 315 for hiccups cases. !!Measurements were performed using two devices. A research version of the Bloomlife wearable device, configured to acquire two channels EMG at 4096 Hz and triaxial accelerometer data at 128 Hz and the TMSi system including 6 accelerometers, five placed on the abdomen and one on the back.!
![Page 15: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/15.jpg)
15 |
FEATURES, CLASSIFIER AND CLASS-IMBALANCE
BHI, 2017 bloomlife.com
Features: low-complexity time domain features (mean, standard!deviation, interquartile range, correlation between axis, sum,!min, max and magnitude).!!Classification: Random forests. We set the number of features to select at each iteration to the square root of the total number of features.!!Class imbalance: small number of kicks with respect to the total available data. The optimal ratio between reference class (kicks) and majority class (non-kicks) was determined by cross-validating and optimizing for F-score. Our optimal balance included all data from the minority class and one fifth of the majority class data !
![Page 16: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/16.jpg)
16 |
COMPARISONS AND CROSS-VALIDATION
BHI, 2017 bloomlife.com
We compared four feature sets associated to the two systems used in this study in order to highlight the impact of the novel methods proposed to improve accuracy of a single wearable device:!1. TMSi (6 accelerometer system) and variable-length features.!2. Bloomlife (single wearable sensor) and features computed
over a short time window only Bloomlife and features computed over both short and long time windows. !
3. Bloomlife and features computed over both short and long time windows plus EMG features. !
All models were derived and validated using leave one!participant out cross-validation!
![Page 17: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/17.jpg)
17 |
RESULTS: SENSITIVITY
BHI, 2017 bloomlife.com
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 18: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/18.jpg)
18 |
RESULTS: SENSITIVITY
BHI, 2017 bloomlife.com
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 19: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/19.jpg)
19 |
RESULTS: SENSITIVITY
BHI, 2017 bloomlife.com
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
Sensitivity does not change much by introducing variable-length and EMG features as the aim of these features is to reduce false positives. !
![Page 20: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/20.jpg)
20 |
RESULTS: SENSITIVITY
BHI, 2017 bloomlife.com
Sensitivity does not change much by introducing variable-length and EMG features as the aim of these features is to reduce false positives. !
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 21: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/21.jpg)
21 |
RESULTS: POSITIVE PREDICTIVE VALUE
BHI, 2017 bloomlife.com
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 22: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/22.jpg)
22 |
RESULTS: POSITIVE PREDICTIVE VALUE
BHI, 2017 bloomlife.com
On the other hand, PPV was 0.75 for the 6 sensors system and increased between 0.65 to 0.75 when including variable-length and EMG features
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 23: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/23.jpg)
23 |
RESULTS: POSITIVE PREDICTIVE VALUE
BHI, 2017 bloomlife.com
On the other hand, PPV was 0.75 for the 6 sensors system and increased between 0.65 to 0.75 when including variable-length and EMG features
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 24: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/24.jpg)
24 |
RESULTS: POSITIVE PREDICTIVE VALUE
BHI, 2017 bloomlife.com
On the other hand, PPV was 0.75 for the 6 sensors system and increased between 0.65 to 0.75 when including variable-length and EMG features
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
Sensitivity
0
25
50
75
Multi SL Single S Single SLSingle SLEModel
Perc
enta
ge (%
)
PPV
![Page 25: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/25.jpg)
25 |
RESULTS: TOTAL NUMBER OF KICKS PER RECORDING
BHI, 2017 bloomlife.com
●●●●●●● ●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Multi SL
●●●●●●●●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single S
●●●●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SL
●●● ●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SLE
![Page 26: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/26.jpg)
26 |
RESULTS: TOTAL NUMBER OF KICKS PER RECORDING
BHI, 2017 bloomlife.com
●●●●●●● ●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Multi SL
●●●●●●●●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single S
●●●●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SL
●●● ●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SLE
![Page 27: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/27.jpg)
27 |
RESULTS: TOTAL NUMBER OF KICKS PER RECORDING
BHI, 2017 bloomlife.com
●●●●●●● ●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Multi SL
●●●●●●●●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single S
●●●●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SL
●●● ●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SLE
![Page 28: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/28.jpg)
28 |
RESULTS: TOTAL NUMBER OF KICKS PER RECORDING
BHI, 2017 bloomlife.com
Less overdetections at the recording level as well
●●●●●●● ●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Multi SL
●●●●●●●●
●
●
●
●
●●
●●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single S
●●●●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SL
●●● ●●●●●
●
●
●
●
●●●
●●
●
●
●
●
●
0
100
200
300
0 100 200 300Detected kicks
Actu
al k
icks
Actual vs Detected kicks, Single SLE
![Page 29: Talk at the International Conference on Biomedical and Health Informatics (BHI 2017)](https://reader031.fdocuments.us/reader031/viewer/2022030313/58ac28c11a28abf03a8b6287/html5/thumbnails/29.jpg)
29 |
CONCLUSIONS
BHI, 2017 bloomlife.com
We proposed a method to improve the accuracy of fetal kicks detection during pregnancy using a single wearable device placed on the abdomen. !!Including variable-length accelerometer features, short fetal movement is averaged out over longer time windows but captured over short ones, while maternal movements of greater intensity appear over windows of both durations. As a result, a single wearable device can be used to better discriminate fetal and maternal movement without the need for a reference accelerometer (11% improvement in PPV).!