Improving soft tissue laser ablation outcomes: A Markov chain Monte Carlo-based approach

Area of Science:

• Biomedical Engineering
• Medical Physics
• Oncology

Background:

• Laser ablation (LA) is a minimally invasive cancer treatment requiring precise temperature control.
• Mathematical models are vital for predicting temperature during LA, but accuracy is limited by parameter uncertainties.
• Optimizing LA treatment outcomes necessitates accurate pre-operative planning and intraoperative guidance.

Purpose of the Study:

• To enhance the accuracy of temperature predictions in laser ablation therapy.
• To quantify the impact of key parameters on temperature distribution during LA.
• To provide real-time insights for optimizing laser energy delivery.

Main Methods:

• Combined the Delayed Rejection Adaptive Metropolis (DRAM) algorithm with bioheat equations.
• Tuned and quantified parameters influencing temperature distribution, including laser distribution, absorption coefficient, and thermal conductivity.
• Validated the model using experimental LA data from ex vivo porcine liver monitored by fiber Bragg grating sensors.

Main Results:

• Identified laser standard distribution (r=-0.64), tissue absorption coefficient (r=-0.36), and thermal conductivity (r=0.15) as key parameters affecting temperature.
• Achieved accurate temperature prediction within 1.0 ± 0.5 °C after model parameter tuning.
• Obtained parameter distributions at each time point, enabling real-time decision-making.

Conclusions:

• The integrated DRAM-bioheat model significantly improves temperature prediction accuracy in laser ablation.
• Parameter quantification provides valuable insights for optimizing laser energy delivery during surgery.
• This approach enhances the potential for safer and more effective minimally invasive cancer treatments.