Transitioning to Bayesian inference in medical diagnostics: priors, MCMC costs, and

[Zoey_S]

I'm a data scientist working on a project where we need to quantify uncertainty in our model's predictions for a new medical diagnostic tool, and my team is debating whether to adopt a Bayesian inference approach over our traditional frequentist methods. I understand the theoretical appeal of priors and posteriors, but I'm struggling with the practical implementation, especially choosing appropriate priors without introducing bias and the computational cost of MCMC for our large dataset. For practitioners who have integrated Bayesian methods into production pipelines, what were the biggest hurdles in moving from theory to application, and how did you validate that your chosen priors were reasonable and not unduly influencing the results? I'm also curious about the trade-offs between different sampling algorithms in terms of convergence speed and accuracy for high-dimensional problems.