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

Abstract

Laser ablation (LA) is a minimally invasive cancer therapy that uses laser energy to selectively heat and destroy cancerous tissues while minimizing damage to surrounding healthy tissue. Accurate temperature predictions derived from mathematical models are instrumental in optimizing treatment outcomes. These models assist surgeons during both pre-operative planning and intraoperative guidance. However, their accuracy can be compromised by inherent model limitations, measurement errors, or parameter uncertainties. To address this challenge, we combined the Delayed Rejection Adaptive Metropolis (DRAM) algorithm, an advanced version of the Markov Chain Monte Carlo (MCMC) method, with bioheat equations to tune and quantify key parameters influencing temperature distribution. Our findings showed that laser standard distribution, tissue absorption coefficient, and thermal conductivity impact the temperature profile, with correlation coefficients of −0.64, −0.36, and 0.15, respectively. To validate the model, temperature data were obtained from an experimental LA setup using ex vivo porcine liver, monitored by fiber Bragg grating sensors. After tuning the model parameters, the simulation accurately predicted temperature distributions within 1.0 ± 0.5 °C. Additionally, the parameter distributions were available at each time point during ablation, offering valuable insights for real-time decision-making, particularly regarding laser energy delivery.