Phillip J. Koo: Hi, my name is Phillip Koo, and welcome back to our exclusive coverage of the 13th International Uro-Oncology Conference here in beautiful Sao Paulo, Brazil. We're very fortunate to have with us today, Dr. Baroni, who's Professor of Radiology at Albert Einstein in Sao Paulo. Thank you for joining us.

Ronaldo Baroni: My pleasure.

Phillip J. Koo: So you're giving a lecture discussing the future perspectives on prostate imaging. Can you give us some of the highlights from your lecture?

Ronaldo Baroni: Sure. First of all, thank you for having me here. It's a pleasure to participate this Uro-Oncology meeting. I would say that prostate MRI is already established as the standard modality for prostate detection prior to biopsy or in order also to avoid negative biopsy, so that's already established. I would say that metaanalysis already confirmed that. I would say that future perspectives includes the more rational and widespread use of prostate MRI. And I would say that we still have a lot of efforts to do worldwide, especially for developing countries to work on a cheaper and more widely available type of imaging for prostate cancer. We still rely on a multiparametric MRI. That's the ideal thing. The BI-RADS steering committee stands that you have to do prostate multiparametric MRI, but there are some studies and very, very important studies saying that for instance, you don't need an endorectal coil for that.

And you also don't need a three Tesla scanner for that. And maybe you don't need contrast paramagnetic gadolinium based contrast for that. So you can rely on a 1.5 Tesla scanner without endorectal coil and maybe without contrast for early detection of prostate cancer or to rule out prostate cancer in a scenario that there is a clinical suspicion for that. So this is one for future perspective to put that more widely available worldwide for patients to do an imaging study and not go from elevated PSA directly to a transrectal biopsy.

Phillip J. Koo: So it's interesting because one of the criticisms of MR has always been sort of that [inaudible 00:02:18] consistency and reliability. And I think, what are your perspectives on how we've progressed over the past 10 years and how we take that step forward? Because I think those concerns still exist, especially with our urology colleagues.

Ronaldo Baroni: There's a tool a lot of work to do in order to increase inter observer variability in terms of reading cases and using BI-RADS classification for that. There was a major achievement BI-RADS, have to rely on that because it's a Likert scale, but you have more standardized ways of performing the exam, reading the exam and reporting the exam. So I would say that was a major achievement, but there is still a lot to do in order to make this more consistent among radiologists and the communication between radiologists and urologists have to increase as well. But I would say that the efforts we have to do now is not only to use MRI, but to train radiologists and practically train and do some hands on workshops on that. We do that a lot in Brazil, the Brazilian College of Radiology and all other societies do a lot of hands on workshops on that.

So there are more radiologists, but let's say adequately trained on how to read MRIs and also to do, let's say the followup of the exam itself. So it's our duty as radiologists, not only to do their MRI and read the MRI, but we have to follow up the case and you have to have your own database and you have to compare your database of results. So what's the positivity for clinically significant cancers for BI-RADS-1, BI-RADS-2, BI-RADS-3, BI-RADS-4, BI-RADS-5 in [inaudible 00:04:16] as compared to the literature. So that's something we have to do all the time and you have to do more tumor boards, maybe online tumor boards. So more radiologists can participate together with urologists, oncologists, and pathologists, so that will make everyone gain experience on the method.

Phillip J. Koo: So that's a great point. And that multidisciplinary approach is something I think we all recognize, really helps everyone be better. I think one of the challenges at a place like Albert Einstein, very academic medical center, something like that is feasible. How is radiology practice more in the community when it comes to prostate MR? And are there opportunities for them to embrace that type of model?

Ronaldo Baroni: So of course there are radiology meetings that the radiologists should attend to become more familiar with the methods. Let's say more experienced radiologists should work on establishing the adequate protocols together with the technologists. So our protocol is, let's say 50% of the exam itself an adequate protocol for prostate MRI to train the technologist, to train the radiologist, let's say to further establish a more effective communication with pathologists, oncologists, and urologists. But I would say a major goal that we have to work together is to let's incorporate MRI as a multidisciplinary tool to predict the possibility of a patient having a significant prostate cancer. So it's not an isolated tool, it's not an isolated exam.

So most of the new literature that is combining MRI with BI-RADS classification, for instance, with PSA density, that's an easy thing to do. And you increase the accuracy of the method in terms of having a positive value or a negative value. When you combine MRI, let's say with a simple test, which is PSA and PSA density based on prostate volume. That's one way we should go and we can further increase this. Let's say this multidisciplinary or multifactor evaluation, because you can put that in a nomogram, including patient age, family history for cancer and other backgrounds that can be incorporated in a more predictive normal ground than MRI itself alone.

Phillip J. Koo: I agree 100%. It's really exciting, because I think we're only scratching the surface when it comes to the data we're actually harnessing from these tests. Well, thank you so much. I know it's great to hear a lot of the challenges that you're seeing are the challenges that we're seeing globally. And it's great to see Brazil, Albert Einstein in yourselves helping answer a lot of these questions. So thank you for joining us.

Ronaldo Baroni: My pleasure. One final point. I want to reach if I have time for that.

Phillip J. Koo: Sure. Please.

Ronaldo Baroni: Is the idea of the artificial intelligence? Is there a threat or is there an opportunity? And I would say definitely an opportunity. That's the final point of my talk on Friday on new perspectives incorporating artificial intelligence in this nomogram as well or in this algorithm as well because it's needing too, it's something that will help radiologists be more assertive, more accurate in reading prostate MRI and more time effective because it takes some steps that are now done by the radiologist that can be done by the machine. And that increases detectability of cancer. And that decreases the reading time of our exam. So if you want to put MRI as a more widely available tool, you have to not only in the future, but nowadays start thinking about artificial intelligence and computer assisted diagnosis to be a step by step incorporated into the daily workflow of prostate cancer detection.

Phillip J. Koo: I agree. It improves patient care and increases access. So I think those are great points. Well, thank you so much for joining us.

Ronaldo Baroni: It was my pleasure. Thank you very much.