Technology and Imaging: “Ask Not What Radiology Can Do, But What It Should Do!”

Each year, over 54,000 people assemble in Chicago on the Sunday following Thanksgiving to attend the Radiological Society of North America Annual Meeting (RSNA), which is consistently one of the largest medical meetings in the world.  The 103rd Annual Meeting recently concluded and was a great chance to reflect on 2017, while taking a peek at 2018 and beyond.  The hot topic this year was artificial intelligence (AI) and machine learning.  Though the idea of AI instills fear in many, it is clear that opportunities abound. Beyond any concerns, we must remain focused on how we utilize these technologies to meaningfully improve the delivery of healthcare for an individual as well as a broader population.   Machine learning will undoubtedly help us achieve these goals as data integration plays a bigger role in the practice of radiology.  With this in mind, there is no question that we must embrace these tools and be at the forefront of their development.

Dr. Elias Zerhouni, former Director of the National Institutes of Health, presented a stimulating lecture on the opening day, challenging radiologists to think differently.  When discussing the future of imaging innovation, Dr. Zerhouni refers to three principles:

  1. What will be is already here
  2. Imaging innovation is multidisciplinary
  3. Focus on what imaging can uniquely do
He noted that radiologists need to enable earlier disease detection, enable disease quantitative characterization, reduce error rates and optimize therapies at the individual patient level.  He closed his lecture by emphatically stating, “Ask not what Radiology can do but what it should do!”.

In the field of imaging, we have technologies and a wealth of data that are being underutilized.  One example is the use of SUVmax in PET/CT.  This semiquantitative metric selects the “hottest” single pixel while disregarding the remainder of the activity in a lesion.  SUVmax has become routine because of its ease of use, but it remains imprecise and unreliable.  When using FDG, other quantitative parameters exist such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG) that have been shown to provide important prognostic information, but these have yet to become a part of routine clinical practice.  Examples of underutilization of data in imaging are endless, but this is changing. 

Radiogenomics is another burgeoning area of investigation as we correlate imaging phenotypes to their genetic properties.  As we venture into these new territories, the benefits of a multidisciplinary approach cannot be underestimated, and it is extremely gratifying to see the growth of these collaborations at all levels from our international/national meetings to our local tumor boards.  According to a recent communication from the American Society for Radiation Oncology (ASTRO), “In 1992, the word “image” appeared in just 3 published article abstracts in the Red Journal.  By 2012, however, the word appeared 106 times, a 35-fold increase!” This will culminate with a special edition of the Red Journal in 2018 covering the impact of imaging in radiation oncology.

With regards to genitourinary imaging, we continue to see great excitement and progress regarding PET technologies and MRI.  PET radiopharmaceuticals have traditionally focused on prostate cancer patients with biochemical relapse, but we are now seeing a growing body of data with regards to imaging with PET at initial diagnosis in patients with intermediate or high-risk disease.  The questions regarding imaging are starting to shift away from not how we image, but when and why we do it.  We must dig deeper into the “why” to understand what impact imaging plays in the care of our patients, and I am confident we will find these answers as we have more and more multidisciplinary teams working together around the globe.

Written by: Philip Koo, MD, FACS, Division Chief of Diagnostic Imaging, Banner MD Anderson Cancer Center in Arizona, Gilbert, Arizona

Published Date: January 8th, 2018