JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Prediction of 12-month recurrence of pancreatic cancer using machine learning and prognostic factors

  • 0Department of Health Information Management, Student Research Committee, School of Health Management and Information Sciences Branch, Iran University of Medical Sciences, Tehran, Iran. raoof.n1370@gmail.com.
Bmc Medical Informatics and Decision Making +

|

November 15, 2024

|

November 15, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

Machine learning models, specifically random forest and support vector machine, accurately predict pancreatic cancer recurrence. Key factors include lymph node metastasis, tumor size, and grade, aiding clinical decisions.

Area Of Science

  • Oncology
  • Medical Informatics
  • Machine Learning

Background

  • Pancreatic cancer (PC) is a lethal disease with high recurrence rates, particularly in patients not receiving adjuvant therapy.
  • Early prediction of PC recurrence is crucial for improving patient prognosis and survival outcomes.
  • Machine learning (ML) techniques show promise in enhancing predictive accuracy across various medical applications.

Purpose Of The Study

  • To develop and evaluate a machine learning-based prediction model for pancreatic cancer recurrence.
  • To identify key clinical and pathological factors associated with PC recurrence.
  • To leverage ML for improved clinical decision-making in PC management.

Main Methods

  • A retrospective analysis of 585 pancreatic cancer patient cases from January 2019 to November 2023 across three clinical centers.
  • Implementation and comparison of ten distinct ensemble and non-ensemble machine learning algorithms for predictive modeling.
  • Evaluation of model performance using metrics such as Area Under the Receiver Operating Characteristic Curve (AU-ROC).

Main Results

  • Random forest and support vector machine algorithms achieved high performance, with an approximate AU-ROC of 0.9 for predicting PC recurrence.
  • Significant factors influencing PC recurrence identified include lymph node metastasis, tumor size, tumor grade, and prior radiotherapy and chemotherapy.
  • The developed models demonstrated strong predictive efficiency for PC recurrence.

Conclusions

  • Random forest and support vector machine algorithms exhibit high performance and clinical utility for predicting pancreatic cancer recurrence.
  • These ML models can assist clinicians in making informed therapeutic and diagnostic decisions.
  • The study highlights the potential of ML in enhancing the management of pancreatic cancer patients.

Keywords: