Machine learning applied to multi-sensor information to reduce false alarm rate in the ICU
Studies reveal that the false alarm rate (FAR) demonstrated by intensive care unit (ICU) vital signs monitors ranges from 0.72 to 0.99. We applied machine learning (ML) to ICU multi-sensor information to imitate a medical specialist in diagnosing patient condition. We hypothesized that applying this data-driven approach to medical monitors will help reduce the FAR even when data from sensors are missing. An expert-based rules algorithm identified and tagged in our dataset seven clinical alarm scenarios. We compared a random forest (RF) ML model trained using the tagged data, where parameters (e.g., heart rate or blood pressure) were (deliberately) removed, in detecting ICU signals with the full expert-based rules (FER), our ground truth, and partial expert-based rules (PER), missing these parameters. When all alarm scenarios were examined, RF and FER were almost identical. However, in the absence of one to three parameters, RF maintained its values of the Youden index (0.94-0.97) and positive predictive value (PPV) (0.98-0.99), whereas PER lost its value (0.54-0.8 and 0.76-0.88, respectively). While the FAR for PER with missing parameters was 0.17-0.39, it was only 0.01-0.02 for RF. When scenarios were examined separately, RF showed clear superiority in almost all combinations of scenarios and numbers of missing parameters. When sensor data are missing, specialist performance worsens with the number of missing parameters, whereas the RF model attains high accuracy and low FAR due to its ability to fuse information from available sensors, compensating for missing parameters.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2020 |
|---|---|
| Erschienen: |
2020 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:34 |
|---|---|
| Enthalten in: |
Journal of clinical monitoring and computing - 34(2020), 2 vom: 06. Apr., Seite 339-352 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Hever, Gal [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
False alarms |
|---|
| Anmerkungen: |
Date Completed 04.06.2021 Date Revised 15.03.2023 published: Print-Electronic Citation Status MEDLINE |
|---|
| doi: |
10.1007/s10877-019-00307-x |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM295767219 |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM295767219 | ||
| 003 | DE-627 | ||
| 005 | 20231225084354.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231225s2020 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s10877-019-00307-x |2 doi | |
| 028 | 5 | 2 | |a pubmed24n0985.xml |
| 035 | |a (DE-627)NLM295767219 | ||
| 035 | |a (NLM)30955160 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Hever, Gal |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Machine learning applied to multi-sensor information to reduce false alarm rate in the ICU |
| 264 | 1 | |c 2020 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 04.06.2021 | ||
| 500 | |a Date Revised 15.03.2023 | ||
| 500 | |a published: Print-Electronic | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a Studies reveal that the false alarm rate (FAR) demonstrated by intensive care unit (ICU) vital signs monitors ranges from 0.72 to 0.99. We applied machine learning (ML) to ICU multi-sensor information to imitate a medical specialist in diagnosing patient condition. We hypothesized that applying this data-driven approach to medical monitors will help reduce the FAR even when data from sensors are missing. An expert-based rules algorithm identified and tagged in our dataset seven clinical alarm scenarios. We compared a random forest (RF) ML model trained using the tagged data, where parameters (e.g., heart rate or blood pressure) were (deliberately) removed, in detecting ICU signals with the full expert-based rules (FER), our ground truth, and partial expert-based rules (PER), missing these parameters. When all alarm scenarios were examined, RF and FER were almost identical. However, in the absence of one to three parameters, RF maintained its values of the Youden index (0.94-0.97) and positive predictive value (PPV) (0.98-0.99), whereas PER lost its value (0.54-0.8 and 0.76-0.88, respectively). While the FAR for PER with missing parameters was 0.17-0.39, it was only 0.01-0.02 for RF. When scenarios were examined separately, RF showed clear superiority in almost all combinations of scenarios and numbers of missing parameters. When sensor data are missing, specialist performance worsens with the number of missing parameters, whereas the RF model attains high accuracy and low FAR due to its ability to fuse information from available sensors, compensating for missing parameters | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a False alarms | |
| 650 | 4 | |a Intensive care unit | |
| 650 | 4 | |a Machine learning | |
| 650 | 4 | |a Missing data | |
| 650 | 4 | |a Random forest | |
| 700 | 1 | |a Cohen, Liel |e verfasserin |4 aut | |
| 700 | 1 | |a O'Connor, Michael F |e verfasserin |4 aut | |
| 700 | 1 | |a Matot, Idit |e verfasserin |4 aut | |
| 700 | 1 | |a Lerner, Boaz |e verfasserin |4 aut | |
| 700 | 1 | |a Bitan, Yuval |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Journal of clinical monitoring and computing |d 1998 |g 34(2020), 2 vom: 06. Apr., Seite 339-352 |w (DE-627)NLM095818405 |x 1573-2614 |7 nnns |
| 773 | 1 | 8 | |g volume:34 |g year:2020 |g number:2 |g day:06 |g month:04 |g pages:339-352 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1007/s10877-019-00307-x |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 34 |j 2020 |e 2 |b 06 |c 04 |h 339-352 | ||