Development of a field artificial intelligence triage tool: Confidence in the prediction of shock, transfusion, and definitive surgical therapy in patients with truncal gunshot wounds

Abstract

BACKGROUND:In-field triage tools for trauma patients are limited by availability of information, linear risk classification, and a lack of confidence reporting. We therefore set out to develop and test a machine learning algorithm that can overcome these limitations by accurately and confidently making predictions to support in-field triage in the first hours after traumatic injury. METHODS:Using an American College of Surgeons Trauma Quality Improvement Program-derived database of truncal and junctional gunshot wound (GSW) patients (aged 16-60 years), we trained an information-aware Dirichlet deep neural network (field artificial intelligence triage). Using supervised training, field artificial intelligence triage was trained to predict shock and the need for major hemorrhage control procedures or early massive transfusion (MT) using GSW anatomical locations, vital signs, and patient information available in the field. In parallel, a confidence model was developed to predict the true-class probability (scale of 0-1), indicating the likelihood that the prediction made was correct, based on the values and interconnectivity of input variables. RESULTS:A total of 29,816 patients met all the inclusion criteria. Shock, major surgery, and early MT were identified in 13.0%, 22.4%, and 6.3% of the included patients, respectively. Field artificial intelligence triage achieved mean areas under the receiver operating characteristic curve of 0.89, 0.86, and 0.82 for prediction of shock, early MT, and major surgery, respectively, for 80/20 train-test splits over 1,000 epochs. Mean predicted true-class probability for errors/correct predictions was 0.25/0.87 for shock, 0.30/0.81 for MT, and 0.24/0.69 for major surgery. CONCLUSION:Field artificial intelligence triage accurately identifies potential shock in truncal GSW patients and predicts their need for MT and major surgery, with a high degree of certainty. The presented model is an important proof of concept. Future iterations will use an expansion of databases to refine and validate the model, further adding to its potential to improve triage in the field, both in civilian and military settings.

Artificial intelligence and machine learning for hemorrhagic trauma care

“…In terms of imputation methods, mean imputation was the most used among the 33 studies which mention how the missing values were handled. Other imputation methods used were iterative or multiple imputation, ElasticNet regression, optimal imputation, chained equation imputation, and median imputation [ 30 , 35 , 44 , 62 , 70 , 71 , 80 , 94 , 97 , 110 , 113 ]. For dealing with imbalanced data, 6 studies addressed it with the most commonly used method being Synthetic Minority Over-Sampling Technique [ 49 , 63 , 72 , 81 , 91 , 99 ].…”

Science fiction or clinical reality: a review of the applications of artificial intelligence along the continuum of trauma care

“…Currently, remote triage takes time and relies on (1) EMS to contact hospitals when high-acuity patients are en route and (2) effective communication between the EMS team and the receiving physician. AI has been shown to predict the need for critical care/life-saving interventions to help stratify incoming trauma patients pre-hospital both generally [ 17 , 19 – 21 ] and in specific trauma subtypes, such as gunshot wounds and after resuscitation [ 22 , 33 ]. The ability to predict the need for life-saving interventions can help inform hospital selection, allowing EMS to route to hospitals with the capacity to handle the necessary care for their patient.…”

“…Algorithm inputs range in complexity from 6–8 inputs mostly comprising vitals, such as in the case of Liu et al and Kim et al [ 20 , 21 ], to more complex analyses that consider time to dispatch, basic laboratories, and injury characteristics [ 18 , 19 , 22 ]. Almost all the studies that were examined for this paper employed types of ANNs to elucidate this relationship, and studies that used a greater number of variables often (but not always) had greater accuracy (AUC 0.82–0.912) as compared to those with fewer input variables (AUC 0.71–0.88) [ 18 – 22 ].…”

Current knowledge and availability of machine learning across the spectrum of trauma science

“…All studies demonstrate the challenge to reach a time trade-off between feature input, meaning how long it takes a clinician to feed all features into the model and how this process can disturb the workflow and how the output integrates into time-critical decision making. Table 1 illustrates several examples [18 ▪ ,25 ▪ ,26–30].…”