Intelligent ensemble learning-enhanced finite element modeling for precision thermal ablation in cancer therapy

Area of Science:

• Computational modeling in oncology
• Minimally invasive thermal therapies
• Biomedical engineering applications

Background:

• Microwave ablation (MWA) is a key minimally invasive treatment for liver tumors.
• Accurate prediction of MWA ablation zones is challenging due to tissue variability and complex thermal dynamics.
• Current methods struggle with precise prediction of ablation outcomes.

Purpose of the Study:

• To develop a hybrid computational framework integrating finite element modeling (FEM) and supervised machine learning (ML).
• To enhance the prediction and optimization of MWA-induced tissue ablation zones.
• To improve accuracy and reduce reliance on iterative laboratory experiments for treatment planning.

Main Methods:

• Conducted ex vivo porcine liver experiments with varying power and duration settings.
• Utilized FEM to simulate coupled electromagnetic and thermal processes.
• Trained a Random Forest regression model on FEM data to optimize antenna placement and predict ablation geometry.

Main Results:

• The integrated FEM-ML framework accurately predicted ablation dimensions, aligning well with experimental data.
• Optimized antenna placement improved temperature predictions, enabling precise estimation of lesion size, shape, and volume.
• The approach facilitates rapid, patient-specific treatment planning and minimizes collateral tissue damage.

Conclusions:

• Combining FEM simulations with supervised learning offers a scalable, data-driven approach for precision MWA.
• This framework enhances predictive reliability and accelerates treatment planning for thermal therapies.
• The method holds potential for improved clinical outcomes in minimally invasive oncological interventions.