Early and Late Fusion Machine Learning on Multi-Frequency Electrical Impedance Data to Improve Radiofrequency...

Area of Science:

• Biomedical Engineering
• Medical Physics
• Oncology

Background:

• Radiofrequency ablation (RFA) is a widely used tumor treatment method.
• Real-time monitoring of RFA, especially across diverse tissue types, remains a significant research challenge.
• Accurate depth estimation is crucial for effective RFA treatment and patient safety.

Purpose of the Study:

• To develop and evaluate a novel method for real-time RFA depth estimation.
• To utilize multi-frequency electrical impedance data and data fusion techniques.
• To incorporate non-linear machine learning (ML) models for improved monitoring.

Main Methods:

• Collected multi-frequency complex electrical impedance measurements from tissues.
• Implemented data fusion schemes to integrate impedance data.
• Employed non-linear machine learning models for depth estimation.
• Assessed the performance of fusion schemes in reducing estimation errors.

Main Results:

• Data fusion schemes significantly reduced the spread of residuals in depth estimation.
• The mean of the residuals for depth estimation was substantially decreased by the fusion schemes.
• The proposed method demonstrated improved accuracy in real-time RFA monitoring.

Conclusions:

• Data fusion is a powerful tool for enhancing ML-based RFA monitoring.
• The integration of multi-frequency impedance data improves real-time RFA depth estimation.
• This approach holds promise for advancing the precision and safety of RFA tumor treatments.