Prediction of therapeutic intensity level from automatic multiclass segmentation of traumatic brain injury lesions on CT-scans.

The prediction of the therapeutic intensity level (TIL) for severe traumatic brain injury (TBI) patients at the early phase of intensive care unit (ICU) remains challenging. Computed tomography images are still manually quantified and then underexploited. In this study, we develop an artificial intelligence-based tool to segment brain lesions on admission CT-scan and predict TIL within the first week in the ICU. A cohort of 29 head injured patients (87 CT-scans; Dataset1) was used to localize (using a structural atlas), segment (manually or automatically with or without transfer learning) 4 or 7 types of lesions and use these metrics to train classifiers, evaluated with AUC on a nested cross-validation, to predict requirements for TIL sum of 11 points or more during the 8 first days in ICU. The validation of the performances of both segmentation and classification tasks was done with Dice and accuracy scores on a sub-dataset of Dataset1 (internal validation) and an external dataset of 12 TBI patients (12 CT-scans; Dataset2). Automatic 4-class segmentation (without transfer learning) was not able to correctly predict the apparition of a day of extreme TIL (AUC = 60 ± 23%). In contrast, manual quantification of volumes of 7 lesions and their spatial location provided a significantly better prediction power (AUC = 89 ± 17%). Transfer learning significantly improved the automatic 4-class segmentation (DICE scores 0.63 vs 0.34) and trained more efficiently a 7-class convolutional neural network (DICE = 0.64). Both validations showed that segmentations based on transfer learning were able to predict extreme TIL with better or equivalent accuracy (83%) as those made with manual segmentations. Our automatic characterization (volume, type and spatial location) of initial brain lesions observed on CT-scan, publicly available on a dedicated computing platform, could predict requirements for high TIL during the first 8 days after severe TBI. Transfer learning strategies may improve the accuracy of CNN-based segmentation models.Trial registrations Radiomic-TBI cohort; NCT04058379, first posted: 15 august 2019; Radioxy-TC cohort; Health Data Hub index F20220207212747, first posted: 7 February 2022.

Contribution of CT-Scan Analysis by Artificial Intelligence to the Clinical Care of TBI Patients.

The gold standard to diagnose intracerebral lesions after traumatic brain injury (TBI) is computed tomography (CT) scan, and due to its accessibility and improved quality of images, the global burden of CT scan for TBI patients is increasing. The recent developments of automated determination of traumatic brain lesions and medical-decision process using artificial intelligence (AI) represent opportunities to help clinicians in screening more patients, identifying the nature and volume of lesions and estimating the patient outcome. This short review will summarize what is ongoing with the use of AI and CT scan for patients with TBI.

Effect of under triage on early mortality after major pediatric trauma: a registry-based propensity score matching analysis.

BACKGROUND: Little is known about the effect of under triage on early mortality in trauma in a pediatric population. Our objective is to describe the effect of under triage on 24-h mortality after major pediatric trauma in a regional trauma system. METHODS: This cohort study was conducted from January 2009 to December 2017. Data were obtained from the registry of the Northern French Alps Trauma System. The network guidelines triage pediatric trauma patients according to an algorithm shared with adult patients. Under triage was defined by the number of pediatric trauma patients that required specialized trauma care transported to a non-level I pediatric trauma center on the total number of injured patients with critical resource use. The effect of under triage on 24-h mortality was assessed with inverse probability treatment weighting (IPTW) and a propensity score (Ps) matching analysis. RESULTS: A total of 1143 pediatric patients were included (mean [SD], age 10 [5] years), mainly after a blunt trauma (1130 [99%]). Of the children, 402 (35%) had an ISS higher than 15 and 547 (48%) required specialized trauma care. Nineteen (1.7%) patients died within 24 h. Under triage rate was 33% based on the need of specialized trauma care. Under triage of children requiring specialized trauma care increased the risk of death in IPTW (risk difference 6.0 [95% CI 1.3-10.7]) and Ps matching analyses (risk difference 3.1 [95% CI 0.8-5.4]). CONCLUSIONS: In a regional inclusive trauma system, under triage increased the risk of early death after pediatric major trauma.

Diagnostic performance of thromboelastometry in trauma-induced coagulopathy: a comparison between two level I trauma centres using two different devices.

PURPOSE: The implementation of a ROTEM®-based algorithm requires reliable thresholds to mirror a prothrombin time (PT) ratio > 1.2 and/or a fibrinogen concentration < 1.5 g l-1. Our goal was to compare the diagnostic performances of two devices (ROTEM® Sigma and Delta, IL Werfen, Munich, Germany) in two level-I trauma centres for the diagnostic of post-traumatic coagulopathy. METHODS: We conducted a retrospective analysis of two registries across two periods of time: from September 2014 to December 2015 in Lyon-Sud university trauma centre and from April 2016 to January 2018 in the Grenoble Alps Trauma Centre. Accuracies of EXTEM and FIBTEM assays to detect patients with coagulation disorders were tested for each device using receiver operating characteristic (ROC) analyses. RESULTS: Within the study period, 74 trauma patients in the Grenoble cohort and 75 trauma patients in the Lyon cohort had concomitant ROTEM® and standard coagulation testing on admission. No statistically significant difference was found between the two ROC curves for FIBTEM amplitude at 5 min (A5), FIBTEM maximum clot firmness, EXTEM clotting time (CT) and EXTEM A5 for ROTEM® Sigma and Delta to diagnose post-traumatic coagulation disorders. The best threshold for FIBTEM A5 to predict low fibrinogen concentration was 7 mm for each device. EXTEM CT thresholds to diagnose PT ratio > 1.2 were 78 s and 74 s for ROTEM® Sigma and Delta, respectively. CONCLUSIONS: These results suggest that ROTEM®-based algorithms may be transposed from one trauma centre to another independently of the setting and the ROTEM® device in use.

Performance of point-of-care international normalized ratio measurement to diagnose trauma-induced coagulopathy.

BACKGROUND: Trauma-induced coagulopathy (TIC) is a common feature after severe trauma. Detection of TIC is based upon classic coagulation tests including international normalized ratio (INR) value. Point-of-care (POC) devices have been developed to rapidly measure INR at the bedside on whole blood. The aim of the study was to test the precision of the Coagucheck® XS Pro device for INR measurement at hospital admission after severe trauma. METHODS: We conducted a prospective observational study in a French level I trauma center. From January 2015 to May 2016, 98 patients with a suspicion of a post-traumatic acute hemorrhage had POC-INR measurement on whole blood concomitantly to classic laboratory INR determination (lab-INR) on plasma at hospital admission. The agreement between the two methods in sorting three predefined categories of INR (normal coagulation, moderate TIC and severe TIC) was evaluated using the Cohen's kappa test with a quadratic weighting. The correlation between POC-INR and lab-INR was measured using the Pearson's coefficient. We also performed a Bland and Altman analysis. RESULTS: The agreement between the lab-INR and the POC-INR was moderate (Kappa = 0.45 [95% CI 0.36-0.50]) and the correlation between the two measurements was also weak (Pearson's coefficient = 0.44 [95% CI 0.27-0.59]). Using a Bland and Altman analysis, the mean difference (bias) for INR was 0.22 [95% CI 0.02-0.42], and the standard deviation (precision) of the difference was 1.01. DISCUSSION/CONCLUSION: POC Coagucheck® XS Pro device is not reliable to measure bedside INR. Its moderate agreement with lab-INR weakens the usefulness of such device after severe trauma. TRIAL REGISTRATION: NCT02869737 . Registered 9 August 2016.

Does continuous wound infiltration enhance baseline intravenous multimodal analgesia after posterior spinal fusion surgery? A randomized, double-blinded, placebo-controlled study.

PURPOSE: There has been a growing interest in continuous local anaesthetic wound infiltration as a non-opioid technique for postoperative pain relief. The impact of this modality on baseline analgesia after spinal fusion surgery has however been inconclusive. We tested whether continuous wound infiltration with ropivacaine can enhance postoperative analgesia compared to a baseline intravenous multimodal analgesia protocol after spinal fusion surgery. METHODS: In this randomized, double-blinded, placebo-controlled study, a multiholed 19-gauge catheter was placed at the end of the surgical procedure through the wound to permit the continuous administration (8 ml/h) of ropivacaine 0.2 % (ropivacaine group; n = 19 patients) or saline (control group; n = 20 patients) during the first 48 postoperative hours (H48). Both groups received intraoperative low-dose ketamine, a combination of acetaminophen, non-steroidal anti-inflammatory drug, and nefopam over the same postoperative period, and morphine delivered by a patient-controlled analgesia (PCA) device. RESULTS: Morphine consumption was comparable between the two groups both at H48, 38 mg (26:52) (median, 25th:75th percentile) (control group) versus 43 mg (19:74) (ropivacaine group), and at H24, 18 mg (16:22) versus 22 mg (9:35) respectively. Pain scores at rest and during mobilization, quality of postoperative sleep, and morphine-related side effects were comparable between the two groups at H24 and H48. CONCLUSION: Our findings indicate that no additional analgesia was provided with continuous wound infiltration of ropivacaine compared to a baseline intravenous multimodal analgesia protocol after spinal fusion surgery. TRIAL REGISTRATION: Clinicaltrials.gov #NCT01743794.