Your browser doesn't support javascript.
loading
Machine learning for the prediction of post-ERCP pancreatitis risk: A proof-of-concept study.
Archibugi, Livia; Ciarfaglia, Gianmarco; Cárdenas-Jaén, Karina; Poropat, Goran; Korpela, Taija; Maisonneuve, Patrick; Aparicio, José R; Casellas, Juan Antonio; Arcidiacono, Paolo Giorgio; Mariani, Alberto; Stimac, Davor; Hauser, Goran; Udd, Marianne; Kylänpää, Leena; Rainio, Mia; Di Giulio, Emilio; Vanella, Giuseppe; Lohr, Johannes Matthias; Valente, Roberto; Arnelo, Urban; Fagerstrom, Niklas; De Pretis, Nicolò; Gabbrielli, Armando; Brozzi, Lorenzo; Capurso, Gabriele; de-Madaria, Enrique.
Afiliación
  • Archibugi L; Pancreato-Biliary Endoscopy and Endosonography Division, Pancreas Translational & Clinical Research Center, San Raffaele Scientific Institute IRCCS, Vita-Salute San Raffaele University, Milan, Italy. Electronic address: archibugi.livia@hsr.it.
  • Ciarfaglia G; Pancreato-Biliary Endoscopy and Endosonography Division, Pancreas Translational & Clinical Research Center, San Raffaele Scientific Institute IRCCS, Vita-Salute San Raffaele University, Milan, Italy.
  • Cárdenas-Jaén K; Gastroenterology Department, Alicante University General Hospital, Alicante Institute for Health and Biomedical Research (ISABIAL), Alicante, Spain.
  • Poropat G; University Hospital of Rijeka, Department of Gastroenterology, Rijeka, Croatia.
  • Korpela T; Helsinki University Hospital and University of Helsinki, Gastroenterological Surgery, Abdominal Center, Helsinki, Finland.
  • Maisonneuve P; Unit of Clinical Epidemiology, Division of Epidemiology and Biostatistics, IEO, European Institute of Oncology IRCCS, Milan, Italy.
  • Aparicio JR; Gastroenterology Department, Alicante University General Hospital, Alicante Institute for Health and Biomedical Research (ISABIAL), Alicante, Spain.
  • Casellas JA; Gastroenterology Department, Alicante University General Hospital, Alicante Institute for Health and Biomedical Research (ISABIAL), Alicante, Spain.
  • Arcidiacono PG; Pancreato-Biliary Endoscopy and Endosonography Division, Pancreas Translational & Clinical Research Center, San Raffaele Scientific Institute IRCCS, Vita-Salute San Raffaele University, Milan, Italy.
  • Mariani A; Pancreato-Biliary Endoscopy and Endosonography Division, Pancreas Translational & Clinical Research Center, San Raffaele Scientific Institute IRCCS, Vita-Salute San Raffaele University, Milan, Italy.
  • Stimac D; University Hospital of Rijeka, Department of Gastroenterology, Rijeka, Croatia.
  • Hauser G; University Hospital of Rijeka, Department of Gastroenterology, Rijeka, Croatia.
  • Udd M; Helsinki University Hospital and University of Helsinki, Gastroenterological Surgery, Abdominal Center, Helsinki, Finland.
  • Kylänpää L; Helsinki University Hospital and University of Helsinki, Gastroenterological Surgery, Abdominal Center, Helsinki, Finland.
  • Rainio M; Helsinki University Hospital and University of Helsinki, Gastroenterological Surgery, Abdominal Center, Helsinki, Finland.
  • Di Giulio E; Department of Gastroenterology, Sant'Andrea Hospital, University Sapienza, Rome, Italy.
  • Vanella G; Pancreato-Biliary Endoscopy and Endosonography Division, Pancreas Translational & Clinical Research Center, San Raffaele Scientific Institute IRCCS, Vita-Salute San Raffaele University, Milan, Italy; Department of Gastroenterology, Sant'Andrea Hospital, University Sapienza, Rome, Italy.
  • Lohr JM; HPD Disease Unit, Karolinska University Hospital, Stockholm, Sweden; Department of Clinical Science, Intervention, and Technology (CLINTEC), Karolinska Institutet, Stockholm, Sweden.
  • Valente R; Department of Clinical Science, Intervention, and Technology (CLINTEC), Karolinska Institutet, Stockholm, Sweden; Department of Surgical Oncology, Anschutz Medical Campus, University of Colorado, Denver, USA.
  • Arnelo U; Department of Clinical Science, Intervention, and Technology (CLINTEC), Karolinska Institutet, Stockholm, Sweden.
  • Fagerstrom N; HPD Disease Unit, Karolinska University Hospital, Stockholm, Sweden.
  • De Pretis N; Gastroenterology Unit, Department of Medicine, Pancreas Center, University of Verona, Verona, Italy.
  • Gabbrielli A; Gastroenterology Unit, Department of Medicine, Pancreas Center, University of Verona, Verona, Italy.
  • Brozzi L; Gastroenterology Unit, Department of Medicine, Pancreas Center, University of Verona, Verona, Italy.
  • Capurso G; Pancreato-Biliary Endoscopy and Endosonography Division, Pancreas Translational & Clinical Research Center, San Raffaele Scientific Institute IRCCS, Vita-Salute San Raffaele University, Milan, Italy.
  • de-Madaria E; Gastroenterology Department, Alicante University General Hospital, Alicante Institute for Health and Biomedical Research (ISABIAL), Alicante, Spain.
Dig Liver Dis ; 55(3): 387-393, 2023 Mar.
Article en En | MEDLINE | ID: mdl-36344369
BACKGROUND: Predicting Post-Endoscopic Retrograde Cholangiopancreatography (ERCP) pancreatitis (PEP) risk can be determinant in reducing its incidence and managing patients appropriately, however studies conducted thus far have identified single-risk factors with standard statistical approaches and limited accuracy. AIM: To build and evaluate performances of machine learning (ML) models to predict PEP probability and identify relevant features. METHODS: A proof-of-concept study was performed on ML application on an international, multicenter, prospective cohort of ERCP patients. Data were split in training and test set, models used were gradient boosting (GB) and logistic regression (LR). A 10-split random cross-validation (CV) was applied on the training set to optimize parameters to obtain the best mean Area Under Curve (AUC). The model was re-trained on the whole training set with the best parameters and applied on test set. Shapley-Additive-exPlanation (SHAP) approach was applied to break down the model and clarify features impact. RESULTS: One thousand one hundred and fifty patients were included, 6.1% developed PEP. GB model outperformed LR with AUC in CV of 0.7 vs 0.585 (p-value=0.012). GB AUC in test was 0.671. Most relevant features for PEP prediction were: bilirubin, age, body mass index, procedure time, previous sphincterotomy, alcohol units/day, cannulation attempts, gender, gallstones, use of Ringer's solution and periprocedural NSAIDs. CONCLUSION: In PEP prediction, GB significantly outperformed LR model and identified new clinical features relevant for the risk, most being pre-procedural.
Asunto(s)
Palabras clave
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE Asunto principal: Pancreatitis / Colangiopancreatografia Retrógrada Endoscópica Tipo de estudio: Clinical_trials / Etiology_studies / Prognostic_studies / Risk_factors_studies Límite: Humans Idioma: En Revista: Dig Liver Dis Asunto de la revista: GASTROENTEROLOGIA Año: 2023 Tipo del documento: Article Pais de publicación: Países Bajos
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE Asunto principal: Pancreatitis / Colangiopancreatografia Retrógrada Endoscópica Tipo de estudio: Clinical_trials / Etiology_studies / Prognostic_studies / Risk_factors_studies Límite: Humans Idioma: En Revista: Dig Liver Dis Asunto de la revista: GASTROENTEROLOGIA Año: 2023 Tipo del documento: Article Pais de publicación: Países Bajos