<i>ShockModes:</i>A Multimodal Model for Prognosticating Intensive Care Outcomes from Physician Notes and Vitals
ABSTRACT Objective Shock Index (SI) is widely used for prognosticating outcomes in ICU and emergency settings. We aimed to create a multi-modal early warning system (EWS) for development of abnormal shock index using routinely available vitals and clinical notes.Material and Methods 17,294 ICU-stays in MIMIC-III data were scored for SI. A new episode of abnormal SI was defined as SI > 0.7 for >30 minutes AND preceded by >=24 hours of normal SI. ICU stays with <24 hours admission, or SI >0.7 within the first 24 hours of admission, or missing SI in >50% in the 24 hour early warning window were excluded, leaving a final cohort of 337 normal and 84 abnormal SI instances. 3117 features from vitals time-series combined with BERT-based features from clinical notes were used to train a battery of machine learning models. The best multimodal pipeline (ShockModes) was assessed for interpretability using SHAP features.Results Vitals-based, notes-based and multi-modal classifiers achieved the best sensitivity of 0.81, 0.81, and 0.83 with corresponding specificity of 0.92, 0.99, and 0.94 respectively, thus demonstrating the potential ofShockModesfor early detection, while preventing false alarms. Global SHAP values revealed Fourier-features of heart rate and heparin sodium prophylaxis as top features. Sensitivity of early detection was highest in acute respiratory failure and chronic kidney disease patients.Conclusion The multimodal, interpretable early warning systemShockModescan be used for prognosticating SI based outcomes in ICU and emergency settings..
| Medienart: |
|
|---|
Preprint| Erscheinungsjahr: |
2022 |
|---|---|
| Erschienen: |
2022 |
| Enthalten in: |
bioRxiv.org - (2022) vom: 22. Dez. Zur Gesamtaufnahme - year:2022 |
|---|
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Pal, Ridam [VerfasserIn] |
|---|
| Links: |
Volltext [kostenfrei] |
|---|
| Themen: |
|---|
| doi: |
10.1101/2022.12.16.22283559 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
XBI038213818 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | XBI038213818 | ||
| 003 | DE-627 | ||
| 005 | 20230429100413.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 221220s2022 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1101/2022.12.16.22283559 |2 doi | |
| 035 | |a (DE-627)XBI038213818 | ||
| 035 | |a (biorXiv)10.1101/2022.12.16.22283559 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Pal, Ridam |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a <i>ShockModes:</i>A Multimodal Model for Prognosticating Intensive Care Outcomes from Physician Notes and Vitals |
| 264 | 1 | |c 2022 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a ABSTRACT Objective Shock Index (SI) is widely used for prognosticating outcomes in ICU and emergency settings. We aimed to create a multi-modal early warning system (EWS) for development of abnormal shock index using routinely available vitals and clinical notes.Material and Methods 17,294 ICU-stays in MIMIC-III data were scored for SI. A new episode of abnormal SI was defined as SI > 0.7 for >30 minutes AND preceded by >=24 hours of normal SI. ICU stays with <24 hours admission, or SI >0.7 within the first 24 hours of admission, or missing SI in >50% in the 24 hour early warning window were excluded, leaving a final cohort of 337 normal and 84 abnormal SI instances. 3117 features from vitals time-series combined with BERT-based features from clinical notes were used to train a battery of machine learning models. The best multimodal pipeline (ShockModes) was assessed for interpretability using SHAP features.Results Vitals-based, notes-based and multi-modal classifiers achieved the best sensitivity of 0.81, 0.81, and 0.83 with corresponding specificity of 0.92, 0.99, and 0.94 respectively, thus demonstrating the potential ofShockModesfor early detection, while preventing false alarms. Global SHAP values revealed Fourier-features of heart rate and heparin sodium prophylaxis as top features. Sensitivity of early detection was highest in acute respiratory failure and chronic kidney disease patients.Conclusion The multimodal, interpretable early warning systemShockModescan be used for prognosticating SI based outcomes in ICU and emergency settings. | ||
| 650 | 4 | |a Biology |7 (dpeaa)DE-84 | |
| 650 | 4 | |a 570 |7 (dpeaa)DE-84 | |
| 700 | 1 | |a Patel, Shaswat |e verfasserin |4 aut | |
| 700 | 1 | |a Bhatnagar, Akshala |e verfasserin |4 aut | |
| 700 | 1 | |a Garg, Hardik |e verfasserin |4 aut | |
| 700 | 1 | |a Singh, Pradeep |e verfasserin |4 aut | |
| 700 | 1 | |a Soun, Ritesh Singh |e verfasserin |4 aut | |
| 700 | 1 | |a Agarwal, Aditya |e verfasserin |4 aut | |
| 700 | 1 | |a Nagori, Aditya |e verfasserin |4 aut | |
| 700 | 1 | |a Khanna, Ashish |e verfasserin |4 aut | |
| 700 | 1 | |a Lodha, Rakesh |e verfasserin |4 aut | |
| 700 | 1 | |a Mathur, Piyush |e verfasserin |4 aut | |
| 700 | 1 | |a Sethi, Tavpritesh |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t bioRxiv.org |g (2022) vom: 22. Dez. |
| 773 | 1 | 8 | |g year:2022 |g day:22 |g month:12 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1101/2022.12.16.22283559 |z kostenfrei |3 Volltext |
| 912 | |a GBV_XBI | ||
| 912 | |a SSG-OLC-PHA | ||
| 951 | |a AR | ||
| 952 | |j 2022 |b 22 |c 12 | ||