1.
Martin Prodel Benoit Audry, Carine Jasseron
Rank Order of Candidates for Heart Transplantation in France: An Explainable Machine Learning Analysis Article de journal
Dans: American Journal of Transplantation, vol. x, 2026, ISSN: 1600-6135.
Résumé | Liens | BibTeX | Étiquettes: machine learning
@article{audry2026rank,
title = {Rank Order of Candidates for Heart Transplantation in France: An Explainable Machine Learning Analysis},
author = {Benoit Audry, Martin Prodel, Carine Jasseron, Christian Jacquelinet, Naissa Bacar, Pascal Leprince, Laurent Sebbag, Andre Vincentelli, Dahbia Horchi, Benoit Averland, William F. Parker, Michel Tsimaratos, Camille Legeai, Richard Dorent},
url = {https://www.amjtransplant.org/article/S1600-6135(26)02755-3/abstract},
doi = {https://doi.org/10.1016/j.ajt.2026.09.012},
issn = {1600-6135},
year = {2026},
date = {2026-09-11},
urldate = {2026-09-11},
journal = {American Journal of Transplantation},
volume = {x},
abstract = {The French composite cardiac allocation score is based on an algorithm that includes interacting components. This multi-step system may hinder understanding of the decision-making process leading to candidate ranking. This study aimed to provide insights into the variables contributing to the ranking of the French transplant candidates. This national cohort study included all candidates ranked between January 2018 and December 2023. The primary endpoint was the normalized ranking of candidates. After determining the most effective machine learning model for predicting ranking, we applied SHapley Additive exPlanations analysis to highlight the relationship between contributing factors and predicted ranking. During the study period, 365,310 rankings were generated, of which 197,684 rankings from 2,430 donors and 3,221 candidates, were other than zero. The post-transplant survival filter was activated in 7% of cases. The top five factors contributing to the ranking were the combined veno-arterial extracorporeal membrane oxygenation/natriuretic peptide variable, total bilirubin level, transport time from procurement to transplant center, candidate age and estimated glomerular filtration rate. The candidate's sex, blood type, and indication for transplantation had minimal impact on the ranking. These results show that our allocation system combines urgency and efficiency, while limiting influence of variables associated with reduced access to transplant.},
keywords = {machine learning},
pubstate = {published},
tppubtype = {article}
}
The French composite cardiac allocation score is based on an algorithm that includes interacting components. This multi-step system may hinder understanding of the decision-making process leading to candidate ranking. This study aimed to provide insights into the variables contributing to the ranking of the French transplant candidates. This national cohort study included all candidates ranked between January 2018 and December 2023. The primary endpoint was the normalized ranking of candidates. After determining the most effective machine learning model for predicting ranking, we applied SHapley Additive exPlanations analysis to highlight the relationship between contributing factors and predicted ranking. During the study period, 365,310 rankings were generated, of which 197,684 rankings from 2,430 donors and 3,221 candidates, were other than zero. The post-transplant survival filter was activated in 7% of cases. The top five factors contributing to the ranking were the combined veno-arterial extracorporeal membrane oxygenation/natriuretic peptide variable, total bilirubin level, transport time from procurement to transplant center, candidate age and estimated glomerular filtration rate. The candidate's sex, blood type, and indication for transplantation had minimal impact on the ranking. These results show that our allocation system combines urgency and efficiency, while limiting influence of variables associated with reduced access to transplant.
2.
Jan Kehrmann Olivier Dauwalder, Stéphane Corvec
Multicentre Evaluation of Cepheid Xpert® Research-Use-Only Panel for Molecular Rapid Detection of Gastrointestinal Pathogens Article de journal
Dans: Microorganisms, vol. 14, ISS. 9, non. 2011, 2026, ISSN: 2076-2607.
Résumé | Liens | BibTeX | Étiquettes: biostatistics
@article{nokey,
title = {Multicentre Evaluation of Cepheid Xpert® Research-Use-Only Panel for Molecular Rapid Detection of Gastrointestinal Pathogens},
author = {Olivier Dauwalder, Jan Kehrmann, Stéphane Corvec, Maroussia Roelens, Tiphaine Roussel-Gaillard, Edoardo Bixio, Jan Buer, Elisa Baillemont, Martin Prodel, Valeria Cento},
url = {https://www.mdpi.com/2076-2607/14/9/2011},
doi = {10.3390/microorganisms14092011},
issn = {2076-2607},
year = {2026},
date = {2026-09-10},
journal = {Microorganisms},
volume = {14},
number = {2011},
issue = {9},
abstract = {Rapid and accurate identification of enteric pathogens is essential for the management of acute infectious diarrhoeal diseases. Multiplex PCR panels have emerged as first-line screening tools to conventional stool culture by providing broad pathogen coverage and shorter turnaround times. This study evaluated the diagnostic performance of the Cepheid Xpert gastrointestinal (GI) research-use-only (RUO) panel in routine clinical practice across multiple European centres. A multicentre study was conducted in four hospitals across France, Germany and Italy. Stool specimens collected as part of routine care were tested using local standard-of-care (SoC) methods, including conventional stool culture, BDMax® Enteric Panels, or BioFire FilmArray® GI Panel, and subsequently analysed using the Cepheid Xpert GI RUO panel. A total of 11 bacterial, viral and parasitic enteropathogens were targeted. Discordant results underwent additional testing, and a consensus result based on a two-out-of-three agreement rule was used as the reference standard. Diagnostic performance metrics, agreement with SoC methods, microbiological findings and turnaround times were assessed. Of 508 samples, 115 (22.6%) were identified as positive for at least one pathogen according to the consensus result, including 12 samples with co-detection of more than one pathogen (10.4% of positive samples), resulting in a total of 128 pathogens detected. The most frequently detected pathogens were Campylobacter spp. (37 specimens), Norovirus (26 specimens), and Salmonella spp. (25 specimens). Compared with the consensus reference standard, Xpert achieved an overall sensitivity of 98% (95% CI: 94–100%), specificity of 100% (95% CI: 100–100%), positive predictive value of 95% (95% CI: 90–98%), and negative predictive value of 100% (95% CI: 100–100%). Sensitivity was ≥95% and specificity ≥99% for all evaluable targets. Agreement with routine diagnostic methods was high, with overall positive and negative percent agreements exceeding 90% and a Cohen's kappa coefficient of 0.91. Median analytical turnaround time was 1 h 18 min for Xpert, compared with 3 h for BDMax, 1 h 15 min for BioFire and 48 h for conventional stool culture. The Cepheid Xpert GI (RUO) panel provided highly accurate detection of enteric pathogens across heterogeneous European diagnostic settings and reduced turnaround time by more than 95% compared with conventional stool culture. A future IVD version of this test could provide rapid and accurate detection of enteric pathogens for potential routine management of patients with suspected acute infectious gastroenteritis.},
keywords = {biostatistics},
pubstate = {published},
tppubtype = {article}
}
Rapid and accurate identification of enteric pathogens is essential for the management of acute infectious diarrhoeal diseases. Multiplex PCR panels have emerged as first-line screening tools to conventional stool culture by providing broad pathogen coverage and shorter turnaround times. This study evaluated the diagnostic performance of the Cepheid Xpert gastrointestinal (GI) research-use-only (RUO) panel in routine clinical practice across multiple European centres. A multicentre study was conducted in four hospitals across France, Germany and Italy. Stool specimens collected as part of routine care were tested using local standard-of-care (SoC) methods, including conventional stool culture, BDMax® Enteric Panels, or BioFire FilmArray® GI Panel, and subsequently analysed using the Cepheid Xpert GI RUO panel. A total of 11 bacterial, viral and parasitic enteropathogens were targeted. Discordant results underwent additional testing, and a consensus result based on a two-out-of-three agreement rule was used as the reference standard. Diagnostic performance metrics, agreement with SoC methods, microbiological findings and turnaround times were assessed. Of 508 samples, 115 (22.6%) were identified as positive for at least one pathogen according to the consensus result, including 12 samples with co-detection of more than one pathogen (10.4% of positive samples), resulting in a total of 128 pathogens detected. The most frequently detected pathogens were Campylobacter spp. (37 specimens), Norovirus (26 specimens), and Salmonella spp. (25 specimens). Compared with the consensus reference standard, Xpert achieved an overall sensitivity of 98% (95% CI: 94–100%), specificity of 100% (95% CI: 100–100%), positive predictive value of 95% (95% CI: 90–98%), and negative predictive value of 100% (95% CI: 100–100%). Sensitivity was ≥95% and specificity ≥99% for all evaluable targets. Agreement with routine diagnostic methods was high, with overall positive and negative percent agreements exceeding 90% and a Cohen's kappa coefficient of 0.91. Median analytical turnaround time was 1 h 18 min for Xpert, compared with 3 h for BDMax, 1 h 15 min for BioFire and 48 h for conventional stool culture. The Cepheid Xpert GI (RUO) panel provided highly accurate detection of enteric pathogens across heterogeneous European diagnostic settings and reduced turnaround time by more than 95% compared with conventional stool culture. A future IVD version of this test could provide rapid and accurate detection of enteric pathogens for potential routine management of patients with suspected acute infectious gastroenteritis.
3.
Martin Prodel Ludovic Lamarsalle, Stephane Roze
vol. 28, non. S539, 2025, ISSN: 1098-3015.
Résumé | Liens | BibTeX | Étiquettes:
@proceedings{lamarsalle2025value,
title = {What Is the Value of Large Language Models for Improving Extraction and Interpretation of Efficiency Data From French Commission Dévaluation Économique et de Santé Publique (CEESP): Decisions on Breast Cancer Medications?},
author = {Ludovic Lamarsalle, Martin Prodel, Stephane Roze},
url = {https://doi.org/10.1016/j.jval.2025.09.3495},
doi = {10.1016/j.jval.2025.09.3495},
issn = {1098-3015},
year = {2025},
date = {2025-12-01},
journal = {Value in Health},
volume = {28},
number = {S539},
issue = {12, Supplement 1},
abstract = {Objectives
Data insight of health-economic evidence is pivotal to understand reimbursement drivers. The economic committee (CEESP) of the French Haute Autorité de Santé conducts cost-effectiveness assessments for products that claim added clinical benefit (ASMR I-III) and are expected to exceed €20 million in annual sales. We assessed whether state-of-the-art Large Language Models (LLMs) can perform data extraction from recent CEESP breast-cancer opinions.
Methods
Sixteen opinions released between March 2014 and April 2024 were analyzed with three LLMs (ChatGPT-4o, Claude-3-5-Sonnet, Mistral-Large 24.11) using a unified prompt covering 65 data fields: 23 administrative and 42 health-economic variables (e.g., incremental cost-effectiveness ratio, modeling assumptions, utilities⋯). LLM outputs were compared pair-wise and against a manually curated reference created by senior health-economists.
Results
The analysis of 16 CEESP opinions revealed two performance patterns. Inter-LLM similarity analysis showed moderate concordance: perfect agreement among all three LLMs was achieved in 33.8% of fields, with partial agreement (≥2 LLMs) in 70.2% of cases. Structured administrative fields demonstrated high inter-model consistency: intervention type (100% agreement), therapeutic area (100%), and regulatory parameters (>90% agreement). However, when compared against expert-validated reference data, overall performance was lower. ChatGPT4o achieved the highest similarity score (0.550), followed by Claude-3.5-Sonnet (0.537) and Mistral-Large (0.519). Important discrepancies emerged in medico-economic fields: target population extraction, cost-effectiveness ratio values, and efficiency condition. Only 31% of fields achieved excellent reference concordance (≥90%), while 52.4% showed moderate performance (<70%). Complex numerical data and free-text clinical descriptions posed the greatest challenges, indicating that LLM consensus does not guarantee accuracy against expert validation.
Conclusions
LLMs demonstrate promising potential for automating health economic data extraction from regulatory documents. However, limitations persist for complex medico-economic parameters requiring specialized domain knowledge. A multi-LLM consensus approach with automated disagreement detection could trigger expert review only for discordant results, significantly reducing manual workload while maintaining quality standards.},
keywords = {},
pubstate = {published},
tppubtype = {proceedings}
}
Objectives
Data insight of health-economic evidence is pivotal to understand reimbursement drivers. The economic committee (CEESP) of the French Haute Autorité de Santé conducts cost-effectiveness assessments for products that claim added clinical benefit (ASMR I-III) and are expected to exceed €20 million in annual sales. We assessed whether state-of-the-art Large Language Models (LLMs) can perform data extraction from recent CEESP breast-cancer opinions.
Methods
Sixteen opinions released between March 2014 and April 2024 were analyzed with three LLMs (ChatGPT-4o, Claude-3-5-Sonnet, Mistral-Large 24.11) using a unified prompt covering 65 data fields: 23 administrative and 42 health-economic variables (e.g., incremental cost-effectiveness ratio, modeling assumptions, utilities⋯). LLM outputs were compared pair-wise and against a manually curated reference created by senior health-economists.
Results
The analysis of 16 CEESP opinions revealed two performance patterns. Inter-LLM similarity analysis showed moderate concordance: perfect agreement among all three LLMs was achieved in 33.8% of fields, with partial agreement (≥2 LLMs) in 70.2% of cases. Structured administrative fields demonstrated high inter-model consistency: intervention type (100% agreement), therapeutic area (100%), and regulatory parameters (>90% agreement). However, when compared against expert-validated reference data, overall performance was lower. ChatGPT4o achieved the highest similarity score (0.550), followed by Claude-3.5-Sonnet (0.537) and Mistral-Large (0.519). Important discrepancies emerged in medico-economic fields: target population extraction, cost-effectiveness ratio values, and efficiency condition. Only 31% of fields achieved excellent reference concordance (≥90%), while 52.4% showed moderate performance (<70%). Complex numerical data and free-text clinical descriptions posed the greatest challenges, indicating that LLM consensus does not guarantee accuracy against expert validation.
Conclusions
LLMs demonstrate promising potential for automating health economic data extraction from regulatory documents. However, limitations persist for complex medico-economic parameters requiring specialized domain knowledge. A multi-LLM consensus approach with automated disagreement detection could trigger expert review only for discordant results, significantly reducing manual workload while maintaining quality standards.
Data insight of health-economic evidence is pivotal to understand reimbursement drivers. The economic committee (CEESP) of the French Haute Autorité de Santé conducts cost-effectiveness assessments for products that claim added clinical benefit (ASMR I-III) and are expected to exceed €20 million in annual sales. We assessed whether state-of-the-art Large Language Models (LLMs) can perform data extraction from recent CEESP breast-cancer opinions.
Methods
Sixteen opinions released between March 2014 and April 2024 were analyzed with three LLMs (ChatGPT-4o, Claude-3-5-Sonnet, Mistral-Large 24.11) using a unified prompt covering 65 data fields: 23 administrative and 42 health-economic variables (e.g., incremental cost-effectiveness ratio, modeling assumptions, utilities⋯). LLM outputs were compared pair-wise and against a manually curated reference created by senior health-economists.
Results
The analysis of 16 CEESP opinions revealed two performance patterns. Inter-LLM similarity analysis showed moderate concordance: perfect agreement among all three LLMs was achieved in 33.8% of fields, with partial agreement (≥2 LLMs) in 70.2% of cases. Structured administrative fields demonstrated high inter-model consistency: intervention type (100% agreement), therapeutic area (100%), and regulatory parameters (>90% agreement). However, when compared against expert-validated reference data, overall performance was lower. ChatGPT4o achieved the highest similarity score (0.550), followed by Claude-3.5-Sonnet (0.537) and Mistral-Large (0.519). Important discrepancies emerged in medico-economic fields: target population extraction, cost-effectiveness ratio values, and efficiency condition. Only 31% of fields achieved excellent reference concordance (≥90%), while 52.4% showed moderate performance (<70%). Complex numerical data and free-text clinical descriptions posed the greatest challenges, indicating that LLM consensus does not guarantee accuracy against expert validation.
Conclusions
LLMs demonstrate promising potential for automating health economic data extraction from regulatory documents. However, limitations persist for complex medico-economic parameters requiring specialized domain knowledge. A multi-LLM consensus approach with automated disagreement detection could trigger expert review only for discordant results, significantly reducing manual workload while maintaining quality standards.
4.
Huguet, Marius; Pehlivan, Canan; Ballereau, François; Dodane-Loyenet, Antoine; Fontanili, Franck; Garaix, Thierry; Yordanov, Youri; Augusto, Vincent; Tazarourte, Karim; Redjaline, Abdesslam
Indoor positioning systems provide insight into emergency department systems enabling proposal of designs to improve workflow Article de journal
Dans: Communications Medicine, vol. 5, non. 1, 2025, (Cited by: 0; All Open Access, Gold Open Access).
@article{Huguet2025,
title = {Indoor positioning systems provide insight into emergency department systems enabling proposal of designs to improve workflow},
author = {Marius Huguet and Canan Pehlivan and François Ballereau and Antoine Dodane-Loyenet and Franck Fontanili and Thierry Garaix and Youri Yordanov and Vincent Augusto and Karim Tazarourte and Abdesslam Redjaline},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105000073700&doi=10.1038%2fs43856-025-00793-y&partnerID=40&md5=0c54a21314363953d989ec26aab4c7b5},
doi = {10.1038/s43856-025-00793-y},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Communications Medicine},
volume = {5},
number = {1},
note = {Cited by: 0; All Open Access, Gold Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
5.
Luu, VP; Denis, Hélène; Bensimon, L; Durand-Zaleski, I; Greillier, L; Thomas, P; Wislez, M; Panes, A; Prodel, Martin; Apert, M; others,
LUCIEN–Étude non interventionnelle sur les délais et les coûts des parcours diagnostiques du cancer du poumon primitif: analyse de processus (process mining) Article de journal
Dans: Revue des Maladies Respiratoires Actualités, vol. 17, non. 1, p. 245–246, 2025.
BibTeX | Étiquettes: pathways, process mining
@article{luu2025lucien,
title = {LUCIEN–Étude non interventionnelle sur les délais et les coûts des parcours diagnostiques du cancer du poumon primitif: analyse de processus (process mining)},
author = {VP Luu and Hélène Denis and L Bensimon and I Durand-Zaleski and L Greillier and P Thomas and M Wislez and A Panes and Martin Prodel and M Apert and others},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Revue des Maladies Respiratoires Actualités},
volume = {17},
number = {1},
pages = {245–246},
publisher = {Elsevier Masson},
keywords = {pathways, process mining},
pubstate = {published},
tppubtype = {article}
}
6.
Perrot, Aurore; Raguideau, Fanny; Denis, Helene; Prodel, Martin; Javelot, Matthieu; Pierres, Marie; Touzeau, Cyrille
Survival, Years of Life Lost and Attrition Rates in Multiple Myeloma Patients in France Article de journal
Dans: Clinical Lymphoma Myeloma and Leukemia, 2025, ISSN: 2152-2650.
Résumé | Liens | BibTeX | Étiquettes: biostatistics, pathways
@article{PERROT2025,
title = {Survival, Years of Life Lost and Attrition Rates in Multiple Myeloma Patients in France},
author = {Aurore Perrot and Fanny Raguideau and Helene Denis and Martin Prodel and Matthieu Javelot and Marie Pierres and Cyrille Touzeau},
url = {https://www.sciencedirect.com/science/article/pii/S2152265025000801},
doi = {https://doi.org/10.1016/j.clml.2025.03.004},
issn = {2152-2650},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Clinical Lymphoma Myeloma and Leukemia},
abstract = {Background
In a context of strong evolution of multiple myeloma (MM) treatment paradigm, real-world data allow a better understandingthe patients' medical needs.
Methods
The present analyses from the MYLORD study were designed to provide recent data on MM patients: characteristics, overall survival (OS), years of life lost (YLL) and attrition rates, using the French National Health Insurance Database (SNDS). It is based on a cohort of 33,032 MM patients who initiated a frontline therapy from 2014 to 2021 in France and who were followed until 2021.
Results
The 5-year OS rate is estimated at 51.9% (95% CI: 51.2–52.6). This rate is 78.3% (95% CI: 77.1–79.5) for patients with frontline autologous stem cell transplant (L1T) and 43.6% (95% CI: 42.8–44.4) for those without transplant (L1NT). The excess of mortality in MM patients is revealed by 12.3 YLL versus the matched general population and is higher in L1T patients than those L1NT (17 vs. 11 years). The attrition rate is important in MM patients with 38% of patients who did not benefit from at least 2 lines of treatment and is higher in L1NT patients than those in L1T (41% vs. 29%). Also, the majority of L1NT patients did not receive a L2 because of death, whereas it is because of stay in line status for L1T patients.
Conclusion
These results emphasize the urgency to treat MM patients with the best and tailored treatment strategy available, for all MM patients beneficiating or not from a transplant in frontline therapy. Further studies will be needed to assess the impact of new therapeutic options in real life.},
keywords = {biostatistics, pathways},
pubstate = {published},
tppubtype = {article}
}
Background
In a context of strong evolution of multiple myeloma (MM) treatment paradigm, real-world data allow a better understandingthe patients' medical needs.
Methods
The present analyses from the MYLORD study were designed to provide recent data on MM patients: characteristics, overall survival (OS), years of life lost (YLL) and attrition rates, using the French National Health Insurance Database (SNDS). It is based on a cohort of 33,032 MM patients who initiated a frontline therapy from 2014 to 2021 in France and who were followed until 2021.
Results
The 5-year OS rate is estimated at 51.9% (95% CI: 51.2–52.6). This rate is 78.3% (95% CI: 77.1–79.5) for patients with frontline autologous stem cell transplant (L1T) and 43.6% (95% CI: 42.8–44.4) for those without transplant (L1NT). The excess of mortality in MM patients is revealed by 12.3 YLL versus the matched general population and is higher in L1T patients than those L1NT (17 vs. 11 years). The attrition rate is important in MM patients with 38% of patients who did not benefit from at least 2 lines of treatment and is higher in L1NT patients than those in L1T (41% vs. 29%). Also, the majority of L1NT patients did not receive a L2 because of death, whereas it is because of stay in line status for L1T patients.
Conclusion
These results emphasize the urgency to treat MM patients with the best and tailored treatment strategy available, for all MM patients beneficiating or not from a transplant in frontline therapy. Further studies will be needed to assess the impact of new therapeutic options in real life.
In a context of strong evolution of multiple myeloma (MM) treatment paradigm, real-world data allow a better understandingthe patients' medical needs.
Methods
The present analyses from the MYLORD study were designed to provide recent data on MM patients: characteristics, overall survival (OS), years of life lost (YLL) and attrition rates, using the French National Health Insurance Database (SNDS). It is based on a cohort of 33,032 MM patients who initiated a frontline therapy from 2014 to 2021 in France and who were followed until 2021.
Results
The 5-year OS rate is estimated at 51.9% (95% CI: 51.2–52.6). This rate is 78.3% (95% CI: 77.1–79.5) for patients with frontline autologous stem cell transplant (L1T) and 43.6% (95% CI: 42.8–44.4) for those without transplant (L1NT). The excess of mortality in MM patients is revealed by 12.3 YLL versus the matched general population and is higher in L1T patients than those L1NT (17 vs. 11 years). The attrition rate is important in MM patients with 38% of patients who did not benefit from at least 2 lines of treatment and is higher in L1NT patients than those in L1T (41% vs. 29%). Also, the majority of L1NT patients did not receive a L2 because of death, whereas it is because of stay in line status for L1T patients.
Conclusion
These results emphasize the urgency to treat MM patients with the best and tailored treatment strategy available, for all MM patients beneficiating or not from a transplant in frontline therapy. Further studies will be needed to assess the impact of new therapeutic options in real life.
7.
Peng, Zhihao; Augusto, Vincent; Perrier, Lionel; Xie, Xiaolan
Optimal Process Mining of Traces with Events and Transition Attributes with Application to Care Pathways of Cancer Patients Article de journal
Dans: IEEE Transactions on Automation Science and Engineering, vol. 21, non. 3, p. 4364 – 4381, 2024, (Cited by: 2; All Open Access, Hybrid Gold Open Access).
@article{Peng20244364,
title = {Optimal Process Mining of Traces with Events and Transition Attributes with Application to Care Pathways of Cancer Patients},
author = {Zhihao Peng and Vincent Augusto and Lionel Perrier and Xiaolan Xie},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85165398199&doi=10.1109%2fTASE.2023.3295947&partnerID=40&md5=8c10f6733aeb6a0ddac21c29ac9ce5ef},
doi = {10.1109/TASE.2023.3295947},
year = {2024},
date = {2024-01-01},
journal = {IEEE Transactions on Automation Science and Engineering},
volume = {21},
number = {3},
pages = {4364 – 4381},
note = {Cited by: 2; All Open Access, Hybrid Gold Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
8.
Dauphinot, Virginie; Laurent, Marie; Prodel, Martin; Civet, Alexandre; Vainchtock, Alexandre; Moutet, Claire; Krolak-Salmon, Pierre; Garnier-Crussard, Antoine
Identification of profiles associated with conversions between the Alzheimer’s disease stages, using a machine learning approach Article de journal
Dans: Alzheimer's Research & Therapy, vol. 16, non. 1, p. 166, 2024.
BibTeX | Étiquettes: machine learning
@article{dauphinot2024identification,
title = {Identification of profiles associated with conversions between the Alzheimer’s disease stages, using a machine learning approach},
author = {Virginie Dauphinot and Marie Laurent and Martin Prodel and Alexandre Civet and Alexandre Vainchtock and Claire Moutet and Pierre Krolak-Salmon and Antoine Garnier-Crussard},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
journal = {Alzheimer's Research & Therapy},
volume = {16},
number = {1},
pages = {166},
publisher = {BioMed Central London},
keywords = {machine learning},
pubstate = {published},
tppubtype = {article}
}
9.
Tula, Ana Lucia; Augusto, Vincent; Boucher, Xavier; Sarazin, Marianne
2024, (Cited by: 0).
@conference{Tula2024964,
title = {AUTOMATIC POPULATION-BASED RESPONSIBILITY MODELING USING PROCESS MINING: APPLICATION TO CHRONIC OBSTRUCTIVE PULMONARY DISEASE},
author = {Ana Lucia Tula and Vincent Augusto and Xavier Boucher and Marianne Sarazin},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85217623165&doi=10.1109%2fWSC63780.2024.10838908&partnerID=40&md5=4807768a1bdbc1fd564a9012b46fda00},
doi = {10.1109/WSC63780.2024.10838908},
year = {2024},
date = {2024-01-01},
journal = {Proceedings - Winter Simulation Conference},
pages = {964 – 975},
note = {Cited by: 0},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
10.
Oliveira, Hugo De; Prodel, Martin; Lamarsalle, Ludovic; Augusto, Vincent; Xie, Xiaolan
Correction: Explaining predictive factors in patient pathways using autoencoders Article de journal
Dans: PloS one, vol. 19, non. 12, p. e0315113, 2024.
BibTeX | Étiquettes: auto-encoders, pathways
@article{de2024correction,
title = {Correction: Explaining predictive factors in patient pathways using autoencoders},
author = {Hugo De Oliveira and Martin Prodel and Ludovic Lamarsalle and Vincent Augusto and Xiaolan Xie},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
journal = {PloS one},
volume = {19},
number = {12},
pages = {e0315113},
keywords = {auto-encoders, pathways},
pubstate = {published},
tppubtype = {article}
}
11.
Gonzalez, Javier-Leonardo -Rodriguez; Franco, Carlos; Pinzón-Espitia, Olga; Caballer, Vicent; Alfonso-Lizarazo, Edgar; Augusto, Vincent
Prediction of pharmaceutical and nonpharmaceutical expenditures associated with Diabetes Mellitus type II based on clinical risk Article de journal
Dans: PLoS ONE, vol. 19, non. 6 June, 2024, (Cited by: 0; All Open Access, Gold Open Access, Green Open Access).
@article{Gonzalez2024,
title = {Prediction of pharmaceutical and nonpharmaceutical expenditures associated with Diabetes Mellitus type II based on clinical risk},
author = {Javier-Leonardo -Rodriguez Gonzalez and Carlos Franco and Olga Pinzón-Espitia and Vicent Caballer and Edgar Alfonso-Lizarazo and Vincent Augusto},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85195037874&doi=10.1371%2fjournal.pone.0301860&partnerID=40&md5=d0b89241ceabcb96f1f6fc32085e589a},
doi = {10.1371/journal.pone.0301860},
year = {2024},
date = {2024-01-01},
journal = {PLoS ONE},
volume = {19},
number = {6 June},
note = {Cited by: 0; All Open Access, Gold Open Access, Green Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
12.
Uhl, Laura; Augusto, Vincent; Dalmas, Benjamin; Alexandre, Youenn; Bercelli, Paolo; Jardinaud, Fanny; Aloui, Saber
Evaluating the Bias in Hospital Data: Automatic Preprocessing of Patient Pathways Algorithm Development and Validation Study Article de journal
Dans: JMIR Medical Informatics, vol. 12, 2024, (Cited by: 0; All Open Access, Gold Open Access, Green Open Access).
@article{Uhl2024,
title = {Evaluating the Bias in Hospital Data: Automatic Preprocessing of Patient Pathways Algorithm Development and Validation Study},
author = {Laura Uhl and Vincent Augusto and Benjamin Dalmas and Youenn Alexandre and Paolo Bercelli and Fanny Jardinaud and Saber Aloui},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85206501788&doi=10.2196%2f58978&partnerID=40&md5=ba7dc6edb8480a6c38aff3f1810f7ff4},
doi = {10.2196/58978},
year = {2024},
date = {2024-01-01},
journal = {JMIR Medical Informatics},
volume = {12},
note = {Cited by: 0; All Open Access, Gold Open Access, Green Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
13.
Raguideau, Fanny; Pierotti, Anthony; Panes, Arnaud; Prodel, Martin; Lamarsalle, Ludovic
Automation of Study Report Abstract Generation by Large Language Models, Methodology and Application Example Proceedings Article
Dans: PHARMACOEPIDEMIOLOGY AND DRUG SAFETY, p. 126–126, WILEY 111 RIVER ST, HOBOKEN 07030-5774, NJ USA 2024.
@inproceedings{raguideau2024automation,
title = {Automation of Study Report Abstract Generation by Large Language Models, Methodology and Application Example},
author = {Fanny Raguideau and Anthony Pierotti and Arnaud Panes and Martin Prodel and Ludovic Lamarsalle},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
booktitle = {PHARMACOEPIDEMIOLOGY AND DRUG SAFETY},
volume = {33},
pages = {126–126},
organization = {WILEY 111 RIVER ST, HOBOKEN 07030-5774, NJ USA},
keywords = {LLM},
pubstate = {published},
tppubtype = {inproceedings}
}
14.
Lay, Jules Le; Augusto, Vincent; Alfonso-Lizarazo, Edgar; Masmoudi, Malek; Gramont, Baptiste; Xie, Xiaolan; Bongue, Bienvenu; Celarier, Thomas
COVID-19 Bed Management Using a Two-Step Process Mining and Discrete-Event Simulation Approach Article de journal
Dans: IEEE Transactions on Automation Science and Engineering, vol. 21, non. 3, p. 3080 – 3091, 2024, (Cited by: 1; All Open Access, Green Open Access).
@article{LeLay20243080,
title = {COVID-19 Bed Management Using a Two-Step Process Mining and Discrete-Event Simulation Approach},
author = {Jules Le Lay and Vincent Augusto and Edgar Alfonso-Lizarazo and Malek Masmoudi and Baptiste Gramont and Xiaolan Xie and Bienvenu Bongue and Thomas Celarier},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85160232631&doi=10.1109%2fTASE.2023.3274847&partnerID=40&md5=5a55e8579bdbbf60985bc8f1d9e63da8},
doi = {10.1109/TASE.2023.3274847},
year = {2024},
date = {2024-01-01},
journal = {IEEE Transactions on Automation Science and Engineering},
volume = {21},
number = {3},
pages = {3080 – 3091},
note = {Cited by: 1; All Open Access, Green Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
15.
Oliveira, Hugo De; Prodel, Martin; Lamarsalle, Ludovic; Augusto, Vincent; Xie, Xiaolan
Erratum: Correction: Explaining predictive factors in patient pathways using autoencoders (PloS one (2022) 17 11 DOI: 10.1371/journal.pone.0277135) Article de journal
Dans: PloS one, vol. 19, non. 12, p. e0315113, 2024, (Cited by: 0; All Open Access, Gold Open Access).
@article{DeOliveira2024e0315113,
title = {Erratum: Correction: Explaining predictive factors in patient pathways using autoencoders (PloS one (2022) 17 11 DOI: 10.1371/journal.pone.0277135)},
author = {Hugo De Oliveira and Martin Prodel and Ludovic Lamarsalle and Vincent Augusto and Xiaolan Xie},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85211410722&doi=10.1371%2fjournal.pone.0315113&partnerID=40&md5=6025795bd011eaf99ab0cadcc87a650d},
doi = {10.1371/journal.pone.0315113},
year = {2024},
date = {2024-01-01},
journal = {PloS one},
volume = {19},
number = {12},
pages = {e0315113},
note = {Cited by: 0; All Open Access, Gold Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
16.
Lequertier, Vincent; Wang, Tao; Fondrevelle, Julien; Augusto, Vincent; Polazzi, Stéphanie; Duclos, Antoine
Length of Stay Prediction with Standardized Hospital Data from Acute and Emergency Care Using a Deep Neural Network Article de journal
Dans: Medical Care, vol. 62, non. 4, p. 225 – 234, 2024, (Cited by: 2).
@article{Lequertier2024225,
title = {Length of Stay Prediction with Standardized Hospital Data from Acute and Emergency Care Using a Deep Neural Network},
author = {Vincent Lequertier and Tao Wang and Julien Fondrevelle and Vincent Augusto and Stéphanie Polazzi and Antoine Duclos},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85187543743&doi=10.1097%2fMLR.0000000000001975&partnerID=40&md5=84afd026ed5e2d2eb1d23152e5ec3085},
doi = {10.1097/MLR.0000000000001975},
year = {2024},
date = {2024-01-01},
journal = {Medical Care},
volume = {62},
number = {4},
pages = {225 – 234},
note = {Cited by: 2},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
17.
Laurent, Marie; Raguideau, Fanny; Prodel, Martin; Assi, Nada; Herquelot, Eléonore; Grenier, Benjamin
Use of a Clustering Algorithm for Treatment Sequences (TAK®) as a Pre-processing Step to Explanatory and Predictive Models Article de journal
Dans: Group, vol. 2, non. 1.6, p. 1–8, 2023.
BibTeX | Étiquettes: machine learning
@article{laurent2use,
title = {Use of a Clustering Algorithm for Treatment Sequences (TAK®) as a Pre-processing Step to Explanatory and Predictive Models},
author = {Marie Laurent and Fanny Raguideau and Martin Prodel and Nada Assi and Eléonore Herquelot and Benjamin Grenier},
year = {2023},
date = {2023-08-23},
urldate = {2023-08-23},
journal = {Group},
volume = {2},
number = {1.6},
pages = {1–8},
keywords = {machine learning},
pubstate = {published},
tppubtype = {article}
}
18.
Perrot, Aurore; Javelot, Matthieu; Deslandes, Marine; Prodel, Martin; Denis, Hélène; Touzeau, Cyrille
Épidémiologie et prise en charge thérapeutique des patients atteints d'un myélome multiple: Données issues de la cohorte MYLORD jusqu'à 2020 Présentation
29.03.2023.
BibTeX | Étiquettes: machine learning
@misc{perrot15epidemiologie,
title = {Épidémiologie et prise en charge thérapeutique des patients atteints d'un myélome multiple: Données issues de la cohorte MYLORD jusqu'à 2020},
author = {Aurore Perrot and Matthieu Javelot and Marine Deslandes and Martin Prodel and Hélène Denis and Cyrille Touzeau},
year = {2023},
date = {2023-03-29},
urldate = {2023-03-29},
journal = {Diabète},
volume = {15},
pages = {8},
keywords = {machine learning},
pubstate = {published},
tppubtype = {presentation}
}
19.
Gasmi, Asma; Augusto, Vincent; Faucheu, Jenny; Morin, Claire; Serpaggi, Xavier
Anomaly Detection in Sleep Habits Using Deep Learning Conférence
vol. 2023-August, 2023, (Cited by: 1).
@conference{Gasmi2023,
title = {Anomaly Detection in Sleep Habits Using Deep Learning},
author = {Asma Gasmi and Vincent Augusto and Jenny Faucheu and Claire Morin and Xavier Serpaggi},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85174400181&doi=10.1109%2fCASE56687.2023.10260617&partnerID=40&md5=8151a5daabcaea60ddc044580b85fd8d},
doi = {10.1109/CASE56687.2023.10260617},
year = {2023},
date = {2023-01-01},
journal = {IEEE International Conference on Automation Science and Engineering},
volume = {2023-August},
note = {Cited by: 1},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
20.
Fleck, Julia L.; Hooijenga, Daniëlle; Phan, Raksmey; Xie, Xiaolan; Augusto, Vincent; Heudel, Pierre-Etienne
Adjuvant therapeutic strategy decision support for an elderly population with localized breast cancer: A monocentric cohort retrospective study Article de journal
Dans: PLoS ONE, vol. 18, non. 8 AUGUST, 2023, (Cited by: 0; All Open Access, Gold Open Access, Green Open Access).
@article{Fleck2023,
title = {Adjuvant therapeutic strategy decision support for an elderly population with localized breast cancer: A monocentric cohort retrospective study},
author = {Julia L. Fleck and Daniëlle Hooijenga and Raksmey Phan and Xiaolan Xie and Vincent Augusto and Pierre-Etienne Heudel},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85168746984&doi=10.1371%2fjournal.pone.0290566&partnerID=40&md5=b4c794b47cbb4008a24a4495cb1df548},
doi = {10.1371/journal.pone.0290566},
year = {2023},
date = {2023-01-01},
journal = {PLoS ONE},
volume = {18},
number = {8 AUGUST},
note = {Cited by: 0; All Open Access, Gold Open Access, Green Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}