“Look before you leap.” This moral is from a fable called The Fox and the Goat. Its author was an ancient Greek storyteller named Aesop. Though his life is obscure, his legendary fame is based on his tales of animals that have charmed kids while also teaching practical lessons. The Fox and the Goat teaches you to gather information about the consequences of your decisions before you take action.
A diagnosis of prostate cancer (PCa) is a call to action. Until recently, most treatment decisions were based on information gained from a PSA blood test followed by TRUS biopsy. However, TRUS biopsy is subject to error. It can overdetect insignificant PCa (does not necessarily require urgent treatment), and it can miss significant disease (if not immediately treated the risk for recurrence increases). Thus, untold numbers of PCa patients jumped into surgery, radiation or Active Surveillance before they had a true look at their disease—and thus risked the consequences of mismatched treatment.
mpMRI offers a better look at PCa
Within the last decade, a revolution has occurred that gives doctors and patients a better look at PCa before leaping to a treatment decision. Multiparametric MRI (mpMRI) is a real-time 3D portrait that reveals the location, size, shape and extent of suspicious lesions (abnormal tissue) in the gland. It characterizes the lesions, distinguishing PCa from infection, inflammation, or BPH. As a result, up to 1/3 of men who have a high PSA test result can avoid a biopsy if mpMRI does not detect significant PCa. In fact, mpMRI “…can reduce overdiagnoses, with 40% fewer clinically indolent PCa and approximately 15% more clinically significant PCa cases detected.”[i] It offers the accurate information needed to assess the consequences of each treatment option, and therefore choose the one most likely to avoid treatment regret. At last, patients can truly look before they leap.
Artificial Intelligence improves image-based diagnosis
Our esteemed European colleagues at Radboud University Medical Center (Nijmegen, The Netherlands) have given us a peek around the corner in terms of the future of mpMRI. In a December, 2021 journal article, de Rooij et al. foresee a sharp increase in the number of patients who will undergo prostate mpMRI before biopsy, a demand that will increase the radiology workload for busy clinicians who administer and interpret the images. They describe the challenge as one of providing “good image quality and diagnostic accuracy while meeting the demands of the expected higher workload.”[ii]
One way to address it is to boost efficiency via shorter mpMRI scanning protocols. Studies have shown that there is very little loss of accuracy; however, these ideal studies were conducted by experts on state-of-the-art magnets at academic centers like Radboud. But out in the real world, there are less powerful magnets and less experienced radiologists—a situation that can shortchange the findings of the authoritative studies. Here’s where Artificial Intelligence (AI) can improve prostate cancer diagnosis, even when image quality is inferior or human expertise is lacking.
The authors write, “AI not only has the potential to improve the detection of clinically significant PCa, which is generally considered the most obvious benefit, but can also play a role in other steps in the diagnostic pathway, from MRI acquisition to generating the radiology report.” This optimism is justified by AI’s proven ability to:
- Enhance the speed of identifying abnormal tissue
- Provide diagnostic accuracy that is competitive with experienced readers
- Classify PCa as significant or insignificant, and even assign a PI-RADS score
- Increase inter-reader agreement
- Predict treatment outcomes for patients.
On the other hand, AI is always a work in progress. It takes time to assemble enormous sets of data (e.g., biopsy slides or mpMRI prostate scans), label or annotate each one, use the annotated data to “train” software recognition, and create algorithms (probability formulas) for AI calculations. When properly trained, AI’s speed at delivering accurate clinical information is astonishing; the results complement the interpretation of readers, and help reduce their workload. We are justifiably enthusiastic, while we also recognize that AI is still a young science with an unlimited future. According to a recent analysis,
…the artificial intelligence model based on medical data such as digital imaging and pathological images is effective in completing basic diagnosis of urinary system tumors, image segmentation of tumor infiltration areas or specific organs, gene mutation prediction and prognostic effect prediction, but most of the models for the requirement of clinical application still need to be improved. On the one hand, it is necessary to further improve the detection, classification, segmentation and other performance of the core algorithm. On the other hand, it is necessary to integrate more standardized medical databases to effectively improve the diagnostic accuracy of artificial intelligence models and make it play greater clinical value.[iii]
Artificial Intelligence will always be a work in progress, especially as ever-larger data sets are accumulated, but there’s no question it’s here to stay and grow. Proven AI tools are already available and in use—our own Center incorporates specific products to boost our diagnostic power—each day witnesses the ever-increasing development of AI for the benefit of doctors and patients everywhere.
NOTE: This content is solely for purposes of information and does not substitute for diagnostic or medical advice. Talk to your doctor if you are experiencing pelvic pain, or have any other health concerns or questions of a personal medical nature.
[i] de Rooij M, van Poppel H, Barentsz JO. Risk Stratification and Artificial Intelligence in Early Magnetic Resonance Imaging-based Detection of Prostate Cancer. Eur Urol Focus. 2021 Dec 15:S2405-4569(21)00305-9.
[iii] Liu K, Zhang M, Li H, Wang X et al. Research status and prospect of artificial intelligence technology in the diagnosis of urinary system tumors]. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2021 Dec 25;38(6):1219-1228.