Mathieu Raux
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Mathieu Raux has a total of 443 co-authors. Here are the top co-authors by number of shared publications:
- Anatole Harrois
Paris-Saclay University, Le Kremlin-Bicetre, France
8 shared publications - Sophie Rym Hamada
Anaesthesia and Critical Care Department, AP-HP, Hôpital Bicêtre, Hôpitaux Universitaires Paris Sud, Université Paris Sud, Le Kremlin Bicêtre, France
8 shared publications - Thomas Similowski
Sorbonne Université, INSERM, UMRS1158 Neurophysiologie respiratoire expérimentale et clinique, Paris, France
8 shared publications - Tobias Gauss
Divison Anesthesia-Critical Care, Grenoble Alpes University Hospital, Grenoble, France
7 shared publications - Jacques Duranteau
Paris-Saclay University, Bicêtre Hospital, Assistance Publique Hôpitaux de Paris (AP-HP)
6 shared publications Bruno Riou
5 shared publications
Latest Publications
- Open Access
- 17 August 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (8) , e0354700
Predicting the Trauma and Injury Severity Score (TRISS) in patients admitted to trauma centers after trauma: A modeling study based on data from the French National Programme for Medicalization of Information
Background: Reliable mortality prediction is central to trauma research, quality assessment, and resource allocation. While the French national hospital discharge database (PMSI) provides comprehensive nationwide data, it does not capture clinical parameters required for established severity scores such as TRISS, which is used for predicting in-hospital survival after trauma and necessary for implementing national-level evaluations of health system performance. This study tested whether TRISS could nonetheless be accurately derived from PMSI data. Methods: We conducted a retrospective study of adult patients admitted for severe trauma to an academic trauma center between 2012 and 2023. Clinical data from medical records were linked with corresponding PMSI records. Surrogate predictors of TRISS were derived from PMSI codes. An extreme gradient boosting (XGBoost) model was used, model performance was assessed using bootstrap validation and interpretability was explored through SHapley Additive exPlanations analyses. Results: A total of 2,523 trauma patients were successfully matched between PMSI and medical records. Median age was 35 years, 79% were male, and in-hospital mortality was 15%. Correlation between Simplified Acute Physiology Score II (SAPS II) in PMSI and medical records was strong (ρ = 0.88), while concordance for shock and coma was weak (κ ≈ 0.25). The XGBoost model using individual PMSI codes achieved the best performance (optimism-adjusted R 2 = 0.72). SAPS II was the main contributor, with additional contributions from shock, coma, and anatomical injury codes. Conclusion: TRISS can be reliably approximated from PMSI data using an XGBoost model. This approach enables patient-specific assessment of trauma severity and mortality risk from PMSI, opening perspectives for nationwide surveillance and quality evaluation of trauma care in France.
- 17 August 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (8) , e0354700
Open AccessPredicting the Trauma and Injury Severity Score (TRISS) in patients admitted to trauma centers after trauma: A modeling study based on data from the French National Programme for Medicalization of Information
Background: Reliable mortality prediction is central to trauma research, quality assessment, and resource allocation. While the French national hospital discharge database (PMSI) provides comprehensive nationwide data, it does not capture clinical parameters required for established severity scores such as TRISS, which is used for predicting in-hospital survival after trauma and necessary for implementing national-level evaluations of health system performance. This study tested whether TRISS could nonetheless be accurately derived from PMSI data. Methods: We conducted a retrospective study of adult patients admitted for severe trauma to an academic trauma center between 2012 and 2023. Clinical data from medical records were linked with corresponding PMSI records. Surrogate predictors of TRISS were derived from PMSI codes. An extreme gradient boosting (XGBoost) model was used, model performance was assessed using bootstrap validation and interpretability was explored through SHapley Additive exPlanations analyses. Results: A total of 2,523 trauma patients were successfully matched between PMSI and medical records. Median age was 35 years, 79% were male, and in-hospital mortality was 15%. Correlation between Simplified Acute Physiology Score II (SAPS II) in PMSI and medical records was strong (ρ = 0.88), while concordance for shock and coma was weak (κ ≈ 0.25). The XGBoost model using individual PMSI codes achieved the best performance (optimism-adjusted R 2 = 0.72). SAPS II was the main contributor, with additional contributions from shock, coma, and anatomical injury codes. Conclusion: TRISS can be reliably approximated from PMSI data using an XGBoost model. This approach enables patient-specific assessment of trauma severity and mortality risk from PMSI, opening perspectives for nationwide surveillance and quality evaluation of trauma care in France.
- Open Access
- 25 March 2026
- Published by Elsevier in Respiratory Physiology & Neurobiology
- Vol. 342, 104566
Association between dyspnea and cognitive task performance in amyotrophic lateral sclerosis: An exploratory study
- 25 March 2026
- Published by Elsevier in Respiratory Physiology & Neurobiology
- Vol. 342, 104566
Open AccessAssociation between dyspnea and cognitive task performance in amyotrophic lateral sclerosis: An exploratory study
- Open Access
- 3 March 2026
- Published by Elsevier in Médecine de Catastrophe - urgences collectives
- Vol. 10 (2) , 132-134
Réserves et formation
- 3 March 2026
- Published by Elsevier in Médecine de Catastrophe - urgences collectives
- Vol. 10 (2) , 132-134
Open AccessRéserves et formation
- Open Access
- 30 January 2026
- Published by Springer Nature in Journal of NeuroEngineering and Rehabilitation
- Vol. 23 (1)
SSVEP-based brain–computer interface enabling graded dyspnoea self-report: proof-of-concept study in healthy volunteers
Background: Mechanically ventilated patients may experience respiratory suffering, which is difficult to assess when verbal communication is impaired. We evaluated the performance of a steady-state visual evoked potential (SSVEP)-based brain–computer interface (BCI) designed to enable self-reporting of dyspnoea in this context. Methods: Forty-nine healthy volunteers were studied under five respiratory conditions: normal breathing (NB), inspiratory resistive loading (IRL), inspiratory threshold loading (ITL), CO₂ inhalation (CO₂), and a return to NB as wash-out (NBWO). Respiratory discomfort was evaluated using a visual analogue scale (VAS). Two BCIs models were tested: a detection BCI (D-BCI), designed to discriminate between ‘breathing is OK’ and ‘breathing is difficult’, and a quantification BCI in the form of a LED-based analogue scale (LAS), composed of five light-emitting diodes. Visual stimuli were delivered at different frequency sets: 12–15 Hz, 15–20 Hz, and 20–30 Hz for the D-BCI; low frequencies (13–17–19–23–29 Hz) and high frequencies (41–43–47–53–59 Hz) for the LAS. Performance was assessed using receiver operating characteristic (ROC) curves; the area under the ROC curve (AUC) was the primary outcome. Results: Participants reported significant respiratory discomfort during IRL, ITL, and CO₂ conditions in the D-BCI groups, and during ITL and CO₂ in the LAS groups, as reflected by higher dyspnoea VAS scores compared to NB. The best-performing frequency sets were 20–30 Hz for the D-BCI (AUC 0.89 [0.89–0.90]) and low frequencies for the LAS (AUC 0.84 [0.83–0.85]). Conclusions: This study demonstrates that an SSVEP-based BCI can sucessfully detect and quantify experimentally induced dyspnoea in healthy individuals. Further research is needed to evaluate its clinical applicability for assessing dyspnoea in non-communicative patients.
- 30 January 2026
- Published by Springer Nature in Journal of NeuroEngineering and Rehabilitation
- Vol. 23 (1)
Open AccessSSVEP-based brain–computer interface enabling graded dyspnoea self-report: proof-of-concept study in healthy volunteers
Background: Mechanically ventilated patients may experience respiratory suffering, which is difficult to assess when verbal communication is impaired. We evaluated the performance of a steady-state visual evoked potential (SSVEP)-based brain–computer interface (BCI) designed to enable self-reporting of dyspnoea in this context. Methods: Forty-nine healthy volunteers were studied under five respiratory conditions: normal breathing (NB), inspiratory resistive loading (IRL), inspiratory threshold loading (ITL), CO₂ inhalation (CO₂), and a return to NB as wash-out (NBWO). Respiratory discomfort was evaluated using a visual analogue scale (VAS). Two BCIs models were tested: a detection BCI (D-BCI), designed to discriminate between ‘breathing is OK’ and ‘breathing is difficult’, and a quantification BCI in the form of a LED-based analogue scale (LAS), composed of five light-emitting diodes. Visual stimuli were delivered at different frequency sets: 12–15 Hz, 15–20 Hz, and 20–30 Hz for the D-BCI; low frequencies (13–17–19–23–29 Hz) and high frequencies (41–43–47–53–59 Hz) for the LAS. Performance was assessed using receiver operating characteristic (ROC) curves; the area under the ROC curve (AUC) was the primary outcome. Results: Participants reported significant respiratory discomfort during IRL, ITL, and CO₂ conditions in the D-BCI groups, and during ITL and CO₂ in the LAS groups, as reflected by higher dyspnoea VAS scores compared to NB. The best-performing frequency sets were 20–30 Hz for the D-BCI (AUC 0.89 [0.89–0.90]) and low frequencies for the LAS (AUC 0.84 [0.83–0.85]). Conclusions: This study demonstrates that an SSVEP-based BCI can sucessfully detect and quantify experimentally induced dyspnoea in healthy individuals. Further research is needed to evaluate its clinical applicability for assessing dyspnoea in non-communicative patients.
- Open Access
- 16 January 2026
- Published by Elsevier in Annals of Intensive Care
- Vol. 16, 100004
Opioid-associated modulation of respiratory-related cortical activity in dyspnoeic mechanically ventilated patients: An electroencephalographic study
- 16 January 2026
- Published by Elsevier in Annals of Intensive Care
- Vol. 16, 100004
Open AccessOpioid-associated modulation of respiratory-related cortical activity in dyspnoeic mechanically ventilated patients: An electroencephalographic study
- 3 August 2024
- Published by Elsevier in Journal of Visceral Surgery
- Vol. 161 (6) , 356-363
French practice of emergency resuscitative thoracotomy. A study based on the Traumabase Registry
- 3 August 2024
- Published by Elsevier in Journal of Visceral Surgery
- Vol. 161 (6) , 356-363
French practice of emergency resuscitative thoracotomy. A study based on the Traumabase Registry
- 30 July 2024
- Published by Elsevier in Anaesthesia Critical Care & Pain Medicine
- Vol. 43 (5) , 101409
Lessons learned from the war in Ukraine for the anesthesiologist and intensivist: A scoping review
- 30 July 2024
- Published by Elsevier in Anaesthesia Critical Care & Pain Medicine
- Vol. 43 (5) , 101409
Lessons learned from the war in Ukraine for the anesthesiologist and intensivist: A scoping review
- Open Access
- 22 September 2023
- Published by Springer Nature in Critical Care
- Vol. 27 (1) , 1-10
Heterogeneity in defining multiple trauma: a systematic review of randomized controlled trials
Introduction: While numerous randomized controlled trials (RCTs) have been conducted in the field of trauma, a substantial portion of them are yielding negative results. One potential contributing factor to this trend could be the lack of agreement regarding the chosen definitions across different trials. The primary objective was to identify the terminology and definitions utilized for the characterization of multiple trauma patients within randomized controlled trials (RCTs). Methods: A systematic review of the literature was performed in MEDLINE, EMBASE and clinicaltrials.gov between January 1, 2002, and July 31, 2022. RCTs or RTCs protocols were eligible if they included multiple trauma patients. The terms employed to characterize patient populations were identified, and the corresponding definitions for these terms were extracted. The subsequent impact on the population recruited was then documented to expose clinical heterogeneity. Results: Fifty RCTs were included, and 12 different terms identified. Among these terms, the most frequently used were “multiple trauma” (n = 21, 42%), "severe trauma" (n = 8, 16%), "major trauma" (n = 4, 8%), and trauma with hemorrhagic shock" (n = 4, 8%). Only 62% of RCTs (n = 31) provided a definition for the terms used, resulting a total of 21 different definitions. These definitions primarily relied on the injury severity score (ISS) (n = 15, 30%), displaying an important underlying heterogeneity. The choice of the terms had an impact on the study population, affecting both the ISS and in-hospital mortality. Eleven protocols were included, featuring five different terms, with "severe trauma" being the most frequent, occurring six times (55%). Conclusion: This systematic review uncovers an important heterogeneity both in the terms and in the definitions employed to recruit trauma patients within RCTs. These findings underscore the imperative of promoting the use of a unique and consistent definition.
- 22 September 2023
- Published by Springer Nature in Critical Care
- Vol. 27 (1) , 1-10
Open AccessHeterogeneity in defining multiple trauma: a systematic review of randomized controlled trials
Introduction: While numerous randomized controlled trials (RCTs) have been conducted in the field of trauma, a substantial portion of them are yielding negative results. One potential contributing factor to this trend could be the lack of agreement regarding the chosen definitions across different trials. The primary objective was to identify the terminology and definitions utilized for the characterization of multiple trauma patients within randomized controlled trials (RCTs). Methods: A systematic review of the literature was performed in MEDLINE, EMBASE and clinicaltrials.gov between January 1, 2002, and July 31, 2022. RCTs or RTCs protocols were eligible if they included multiple trauma patients. The terms employed to characterize patient populations were identified, and the corresponding definitions for these terms were extracted. The subsequent impact on the population recruited was then documented to expose clinical heterogeneity. Results: Fifty RCTs were included, and 12 different terms identified. Among these terms, the most frequently used were “multiple trauma” (n = 21, 42%), "severe trauma" (n = 8, 16%), "major trauma" (n = 4, 8%), and trauma with hemorrhagic shock" (n = 4, 8%). Only 62% of RCTs (n = 31) provided a definition for the terms used, resulting a total of 21 different definitions. These definitions primarily relied on the injury severity score (ISS) (n = 15, 30%), displaying an important underlying heterogeneity. The choice of the terms had an impact on the study population, affecting both the ISS and in-hospital mortality. Eleven protocols were included, featuring five different terms, with "severe trauma" being the most frequent, occurring six times (55%). Conclusion: This systematic review uncovers an important heterogeneity both in the terms and in the definitions employed to recruit trauma patients within RCTs. These findings underscore the imperative of promoting the use of a unique and consistent definition.
- 1 June 2022
- Published by Elsevier in The Surgeon
- Vol. 20 (3) , e20-e25
Depacked patients who underwent a shortened perihepatic packing for severe blunt liver trauma have a high survival rate: 20 years of experience in a level I trauma center
- 1 June 2022
- Published by Elsevier in The Surgeon
- Vol. 20 (3) , e20-e25
Depacked patients who underwent a shortened perihepatic packing for severe blunt liver trauma have a high survival rate: 20 years of experience in a level I trauma center