Bayesian method

I want perform a diagnostic test accuracy meta-analysis and specifically to construct an HSROC curve or perform a bivariate analysis of sensitivity and specificity.

According to the article introdusing the MetaAnalyst software in BMC, I gather that a Bayes approach would be most appropriate. However, I find difficulty in setting the starting values in order to perform the analysis (image attached), and consequently an error message occurs.

I want to ask some questions:

* Is the Bayes approach the correct approach in order to construct an HSROC curve? If not, what is the correct method?
* How can we input the correct values to run this specific analysis? (are they automatically implementes through MCMC?)
* Are there any other specific instructions in order to perform a bivariate analysis of sensitivity and specificity with the MetaAnalyst?

Thank you