Last year, more than 800 prostate MRI scans were done at The York/WellSpan and Dr. Steiner reads about 700 of those scans. A number that is expected to keep increasing due to the change in guidelines for PCa diagnosis published by the ACR (American College of Radiology).
Last year, 400 of those cases were biopsied.
Among biopsy pathology results, research* shows that complete agreement was seen in 56% of cases. Dr. Steiner find this fact quite concerning as, in general, the field is more likely to undergrade a pathology of prostate cancer.
The undergrading of prostate tumors comes primarily from the fact that these tumors are mainly heterogeneous and multifocal which makes prostate MRI interpretation more complex and a challenge to master. And even though he recognizes that PI-RADS is a great grading system, he thinks radiologists could benefit from getting some support. That's why he was quite excited about the potential of an AI solution applied to prostate imaging.
*Goodman et al. Frequency and determinants of disagreement and error in gleason scores: A population-based study of prostate cancer.
Dr. Steiner found that the AI of our software is very valuable, especially because it returns results, such as the prostate volume and the PSA density, within seconds of inputting the PSA value.
Besides that, the software analyzes the kinetic curves of enhancement, an important feature to Dr. Steiner as they help him assess tumor heterogeneity and rapid inflow and outflow patterns, making him even more alert for cancerous tissue.
Furthermore, Dr. Steiner also finds the bi-paramatric combination image that is provided by the software quite useful since it highlights suspicious areas. This makes it truly easy to focus on the areas that require his further attention.
Lastly, Dr. Steiner values a lot the standardized report that the software automatically generates because it is a structured, complete and nice depiction that referring physicians like to see.
To further test the performance and the utility of Quantib® Prostate, Dr. Steiner and his team looked at 25 sequential cases of clinically significant prostate cancer retrospectively. Those cases were initially read by 3 board-certified radiologists and underwent fusion biopsy afterwards.
During RSNA 2022, Dr. Steiner announced that Quantib® Prostate detected all lesions, thus demonstrating a sensitivity rate of 100%, and detected 0 false negative and 4 false positive readings, thus achieving a positive predictive value (PPV) of approximately 90%. He also stated that 9 of the detected lesions by Quantib® Prostate were previously under-called by radiologists. Those were later confirmed as cancer by biopsies.
Dr. Steiner found that Quantib® Prostate can make prostate MRI interpretation significantly more rapid and accurate. He believes it could easily reduce this interpretation time by 30% or even more, once radiologists get more comfortable with the technology.
Quantib® Prostate also helps him focus on secondary abnormalities that may not be immediately obvious. There actually have been multiple cases where the AI software highlighted a second, less prominent ROI that he overlooked in his first read, and after the biopsy, he has found that those were indeed cancer. A surprising added value, according to Dr. Steiner.
He also believes the tool is exceedingly valuable for teaching prostate MRI interpretation. He has noticed that the starting radiologists he trains can initially be slightly uncomfortable when reading prostate MRI, so it is of tremendous help for them to support their training.