Influencing Factors for Postembolization Fever in Patients Undergoing Transarterial Chemoembolization Based on...

Area of Science:

• Oncology
• Medical Imaging
• Machine Learning

Background:

• Postembolization fever is a common complication of transarterial chemoembolization (TACE).
• Currently, no established framework exists for predicting postembolization fever.
• Predictive factors may encompass demographic, clinical, laboratory, and radiologic data.

Purpose of the Study:

• To develop and validate machine learning-based prediction models for postembolization fever following TACE.
• To identify key clinical, laboratory, and radiologic predictors of postembolization fever.

Main Methods:

• Retrospective review of data from 1495 patients who underwent TACE.
• Development of seven machine learning algorithms using SPSS and Python.
• Validation of prediction models for postembolization fever occurrence.

Main Results:

• An ensemble method demonstrated the best performance in predicting postembolization fever.
• Positive predictors included post-TACE AST, CRP, ALT, bilirubin, INR, platelets, pre-TACE AST, alpha-fetoprotein, platelets, lipiodol/doxorubicin amounts, and tumor size >5 cm.
• Negative predictors included post-TACE lymphocyte/monocyte counts, albumin, pre-TACE albumin/lymphocyte count, and likely hepatocellular carcinoma.

Conclusions:

• Machine learning models can effectively predict postembolization fever after TACE.
• Clinicians should monitor various patient data before and after TACE to assess fever risk.
• Awareness of TACE side effects and management strategies is crucial for healthcare professionals.