American College of Surgeons survival calculator for biliary tract cancers: using machine learning to individualize predictions
- Lauren M Janczewski 1, Joseph Cotler 2, Xuan Zhu 2, Bryan Palis 2, Kelley Chan 2, Ryan P Merkow 3, Elizabeth B Habermann 4, Ronald J Weigel 5, Judy C Boughey 6
- Lauren M Janczewski 1, Joseph Cotler 2, Xuan Zhu 2
- 1American College of Surgeons Cancer Programs, Chicago, IL; Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL. Electronic address: https://www.twitter.com/LJanczewskiMD.
- 2American College of Surgeons Cancer Programs, Chicago, IL.
- 3Department of Surgery, University of Chicago Pritzker School of Medicine, Chicago, IL.
- 4American College of Surgeons Cancer Programs, Chicago, IL; Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, MN.
- 5American College of Surgeons Cancer Programs, Chicago, IL; Department of Surgery, University of Iowa, Iowa City, IA.
- 6American College of Surgeons Cancer Programs, Chicago, IL; Department of Surgery, Mayo Clinic, Rochester, MN.
- 0American College of Surgeons Cancer Programs, Chicago, IL; Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL. Electronic address: https://www.twitter.com/LJanczewskiMD.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Machine learning created a Biliary Tract Cancer Survival Calculator for personalized prognosis. This tool, using patient, tumor, and treatment data, offers more accurate survival estimates than stage alone.
Area Of Science
- Oncology
- Medical Informatics
- Machine Learning in Healthcare
Background
- Cancer prognosis commonly relies on tumor stage, but survival is multifactorial.
- Biliary tract cancers (bile duct, gallbladder) have complex survival dynamics.
- Developing advanced prognostic tools is crucial for personalized cancer care.
Purpose Of The Study
- To develop a prototype Biliary Tract Cancer Survival Calculator using machine learning.
- To generate personalized survival estimates based on diverse patient, tumor, and treatment factors.
- To improve upon traditional staging-based prognosis for biliary tract malignancies.
Main Methods
- Utilized the National Cancer Database (2010-2017) for 62,877 biliary tract cancer patients.
- Employed random forest algorithms to identify key prognostic variables.
- Developed a survival model using extreme gradient boosting with survival embeddings for 3-year survival prediction.
Main Results
- Identified metastatic disease, age, and lack of surgery as key factors for worse survival.
- The machine learning model demonstrated superior performance (c-index: 0.74) compared to a stage-only model (c-index: 0.64).
- The calculator integrates patient demographics, tumor characteristics (stage, grade, site), and treatment modalities (surgery, chemotherapy, radiation).
Conclusions
- The Biliary Tract Cancer Survival Calculator is a highly accurate prognostic tool.
- It provides individualized, real-time survival estimates.
- This tool enhances personalized treatment planning and patient counseling in oncology.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.

