Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population
Abstract Determination of prognosis in the triage process after traumatic brain injury (TBI) is difficult to achieve. Current severity measures like the Trauma and injury severity score (TRISS) and revised trauma score (RTS) rely on additional information from the Glasgow Coma Scale (GCS) and the In...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-03-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-57538-5 |
| _version_ | 1797219808995966976 |
|---|---|
| author | Shiming Yang Peter Hu Konstantinos Kalpakis Bradford Burdette Hegang Chen Gunjan Parikh Ryan Felix Jamie Podell Neeraj Badjatia |
| author_facet | Shiming Yang Peter Hu Konstantinos Kalpakis Bradford Burdette Hegang Chen Gunjan Parikh Ryan Felix Jamie Podell Neeraj Badjatia |
| author_sort | Shiming Yang |
| collection | DOAJ |
| description | Abstract Determination of prognosis in the triage process after traumatic brain injury (TBI) is difficult to achieve. Current severity measures like the Trauma and injury severity score (TRISS) and revised trauma score (RTS) rely on additional information from the Glasgow Coma Scale (GCS) and the Injury Severity Score (ISS) which may be inaccurate or delayed, limiting their usefulness in the rapid triage setting. We hypothesized that machine learning based estimations of GCS and ISS obtained through modeling of continuous vital sign features could be used to rapidly derive an automated RTS and TRISS. We derived variables from electrocardiograms (ECG), photoplethysmography (PPG), and blood pressure using continuous data obtained in the first 15 min of admission to build machine learning models of GCS and ISS (ML-GCS and ML-ISS). We compared the TRISS and RTS using ML-ISS and ML-GCS and its value using the actual ISS and GCS in predicting in-hospital mortality. Models were tested in TBI with systemic injury (head abbreviated injury scale (AIS) ≥ 1), and isolated TBI (head AIS ≥ 1 and other AIS ≤ 1). The area under the receiver operating characteristic curve (AUROC) was used to evaluate model performance. A total of 21,077 cases (2009–2015) were in the training set. 6057 cases from 2016 to 2017 were used for testing, with 472 (7.8%) severe TBI (GCS 3–8), 223 (3.7%) moderate TBI (GCS 9–12), and 5913 (88.5%) mild TBI (GCS 13–15). In the TBI with systemic injury group, ML-TRISS had similar AUROC (0.963) to TRISS (0.965) in predicting mortality. ML-RTS had AUROC (0.823) and RTS had AUROC 0.928. In the isolated TBI group, ML-TRISS had AUROC 0.977, and TRISS had AUROC 0.983. ML-RTS had AUROC 0.790 and RTS had AUROC 0.957. Estimation of ISS and GCS from machine learning based modeling of vital sign features can be utilized to provide accurate assessments of the RTS and TRISS in a population of TBI patients. Automation of these scores could be utilized to enhance triage and resource allocation during the ultra-early phase of resuscitation. |
| first_indexed | 2024-04-24T12:39:32Z |
| format | Article |
| id | doaj.art-8caa2b11db5b4ca68e217737084d498b |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-24T12:39:32Z |
| publishDate | 2024-03-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-8caa2b11db5b4ca68e217737084d498b2024-04-07T11:18:56ZengNature PortfolioScientific Reports2045-23222024-03-0114111110.1038/s41598-024-57538-5Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury populationShiming Yang0Peter Hu1Konstantinos Kalpakis2Bradford Burdette3Hegang Chen4Gunjan Parikh5Ryan Felix6Jamie Podell7Neeraj Badjatia8Program in Trauma, University of Maryland School of MedicineProgram in Trauma, University of Maryland School of MedicineDepartment of Computer Science and Electrical Engineering, University of Maryland, Baltimore CountyProgram in Trauma, University of Maryland School of MedicineDepartment of Epidemiology and Public Health, University of Maryland School of MedicineProgram in Trauma, University of Maryland School of MedicineFischell Department of Bioengineering, University of MarylandProgram in Trauma, University of Maryland School of MedicineProgram in Trauma, University of Maryland School of MedicineAbstract Determination of prognosis in the triage process after traumatic brain injury (TBI) is difficult to achieve. Current severity measures like the Trauma and injury severity score (TRISS) and revised trauma score (RTS) rely on additional information from the Glasgow Coma Scale (GCS) and the Injury Severity Score (ISS) which may be inaccurate or delayed, limiting their usefulness in the rapid triage setting. We hypothesized that machine learning based estimations of GCS and ISS obtained through modeling of continuous vital sign features could be used to rapidly derive an automated RTS and TRISS. We derived variables from electrocardiograms (ECG), photoplethysmography (PPG), and blood pressure using continuous data obtained in the first 15 min of admission to build machine learning models of GCS and ISS (ML-GCS and ML-ISS). We compared the TRISS and RTS using ML-ISS and ML-GCS and its value using the actual ISS and GCS in predicting in-hospital mortality. Models were tested in TBI with systemic injury (head abbreviated injury scale (AIS) ≥ 1), and isolated TBI (head AIS ≥ 1 and other AIS ≤ 1). The area under the receiver operating characteristic curve (AUROC) was used to evaluate model performance. A total of 21,077 cases (2009–2015) were in the training set. 6057 cases from 2016 to 2017 were used for testing, with 472 (7.8%) severe TBI (GCS 3–8), 223 (3.7%) moderate TBI (GCS 9–12), and 5913 (88.5%) mild TBI (GCS 13–15). In the TBI with systemic injury group, ML-TRISS had similar AUROC (0.963) to TRISS (0.965) in predicting mortality. ML-RTS had AUROC (0.823) and RTS had AUROC 0.928. In the isolated TBI group, ML-TRISS had AUROC 0.977, and TRISS had AUROC 0.983. ML-RTS had AUROC 0.790 and RTS had AUROC 0.957. Estimation of ISS and GCS from machine learning based modeling of vital sign features can be utilized to provide accurate assessments of the RTS and TRISS in a population of TBI patients. Automation of these scores could be utilized to enhance triage and resource allocation during the ultra-early phase of resuscitation.https://doi.org/10.1038/s41598-024-57538-5Traumatic brain injuryGlasgow coma scaleInjury severity scoreMortalityMachine learning |
| spellingShingle | Shiming Yang Peter Hu Konstantinos Kalpakis Bradford Burdette Hegang Chen Gunjan Parikh Ryan Felix Jamie Podell Neeraj Badjatia Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population Scientific Reports Traumatic brain injury Glasgow coma scale Injury severity score Mortality Machine learning |
| title | Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population |
| title_full | Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population |
| title_fullStr | Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population |
| title_full_unstemmed | Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population |
| title_short | Utilizing ultra-early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population |
| title_sort | utilizing ultra early continuous physiologic data to develop automated measures of clinical severity in a traumatic brain injury population |
| topic | Traumatic brain injury Glasgow coma scale Injury severity score Mortality Machine learning |
| url | https://doi.org/10.1038/s41598-024-57538-5 |
| work_keys_str_mv | AT shimingyang utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT peterhu utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT konstantinoskalpakis utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT bradfordburdette utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT hegangchen utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT gunjanparikh utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT ryanfelix utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT jamiepodell utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation AT neerajbadjatia utilizingultraearlycontinuousphysiologicdatatodevelopautomatedmeasuresofclinicalseverityinatraumaticbraininjurypopulation |