Dr. Chip Truwit attended the School of Foreign Service and School of Medicine at Georgetown University, completed his residency in Diagnostic Radiology at Brooke Army Medical Center and his fellowship in Neuroradiology at University of California, San Francisco. As the Chief Medical Officer of diagnostic imaging at Philips, he draws upon his history of healthcare innovation to lead the company. From his pioneering introduction of digital publishing to Radiology in the early 90s to his novel approach to intraoperative MRI-guided neurosurgery in the late 90s, Dr. Truwit has helped to bring in the age of near real-time medical publishing and intraoperative imaging.
Chip Truwit has 19 years of experience as Chief of Radiology, guiding Hennepin Healthcare’s Department of Radiology through transformation, through its ascent to leadership as the premiere Department of Radiology in the Twin Cities. Taking on new challenges, he is now the Chief Medical Officer of the Diagnostic Imaging Business Group at Philips. As a medical imaging industry veteran, he provides medical leadership and radiology expertise to clinical vision
Please Provide Our Readers With Insights About The Key Areas Of Innovation And Diagnostic Imaging Technology Advances That Helped Shape Today’s Medical Imaging Market.
I was a neuroradiologist for many years, and during that time, I was also privileged to serve as a department chair. So, I have seen things changing. The modern history of radiology consists of four pivotal points of inflection.
First,was the introduction of computed tomography (CT) itself. No one would argue that there have been many key innovations in imaging: MRI, PET, digital x-ray, point-of-care ultrasound, to name a few. That said, it is unquestionably the CT scanner that fundamentally changed radiology and medicine. The CT scanner undeniably is the most important modality in almost any department of radiology: because of CT, lives are saved every day.
Second,was the multi-detector CT scanner, whose speed reached such heights that we could essentially “freeze” heart motion. This opened up a dazzling array of diagnostic possibilities including not only 3D but 4D imaging. Third, the CT vendors brought forth what Hounsfield anticipated, dual energy CT. Despite the promise, however, adoption of dual energy was slow. Recently, with the latest technology advances and innovations in Spectral CT and software, radiologists were empowered to bring this solution to bear on patient care.
Finally, the advent of artificial intelligence is introducing yet another era in medical imaging. The application of AI in imaging is multifaceted, impacting many aspects of the patient journey: workflow enhancements both before and after the patient is seen, during image set-up and acquisition, image reconstruction, both reducing noise (de-noising) and dose reduction, and raising the bar on the quality of diagnostic interpretation.
Based On These Four Key Inflection Points, What Trends Do You See In Medical Imaging To Day Or On The Horizon For Radiologists In The Future?
One specific CT-related trend we’ll see is better ways to take advantage of image de-noising techniques to both improve image quality and decrease both radiation and contrast dose. Initially, this work included machine learning techniques: iterative reconstruction, model-based iterative reconstruction, both of which were successful, albeit incomplete. These techniques offered considerable advantages in particular, with respect to CT angiography. More recently, de-noising has assumed more of a deep learning character. With successful implementation, there seems to be little question that radiation doses will be dramatically reduced without significant image compromise, which I see as the second coming of CT–improved image quality with reduced dose.
Another trend we’ll see relates to AI and the explosion of data, not just the imaging data, but the entire digital patient from digital pathology and digital genomic information to digital analytics of workflow and performance. One application of AI will focus on analyzing this data to look for patterns across similar populations of patient data. From this use of AI, what will evolve is imaging profiles, genomic profiles, and pathology profiles that, when viewed together, may reveal more insight on patients that could potentially benefit from one type of chemotherapy or that will predict chances of worse outcomes consequent to a particular therapy.
Increasingly, we’ll see that AI will be able to detect more than the average human, by virtue of reviewing thousands, if not millions, of scans as part of the learning. AI will perform these image reviews in seconds, if not milliseconds, and, unlike humans, AI will not fatigue. Thus, for mammography, CT, MR, PET and others, image interpretation will be undertaken by AI software, both as a form of triage (i.e. - Which patients have intracranial hemorrhage? Which have pulmonary nodules, pneumothorax, or pulmonary embolism?), and as a security check against human performance (the night shift tele-radiologist, the resident radiologist, the potentially compromised radiologist or as a screen to identify incidental findings on CT or MR studies).
Finally, a third trend we’ll see is increasing recognition that the addition of Spectral CT imaging data affords more degrees of freedom to the AI picture. Thus, we can expect to see two additional features of the AI story consequent to the recent rapid evolution of Spectral CT. These include Spectral CT being able to reveal “applets” such as routine reconstruction of, display,and AI assessment of gall bladder images, for example, to ensure universal diagnosis of cholesterol gallstones, otherwise invisible on conventional CT. Similarly, AI is likely to render straightforward with one exam – first time right - the diagnosis of bone edema in the assessment of acute versus chronic compression fractures, differentiation of pancreatitis versus pancreatic necrosis, unsuspected myocardial infarction, bowel infarction, and others, largely due to spectral isolation of iodine in CT contrast. In all likelihood, Spectral CT will afford simple diagnoses that were often challenging by conventional CT, many of which typically required follow-up ultrasound or MR imaging.
What Are Major Pain Points Or Challenges That You See When It Comes To The Medical Imaging Space, And How Is Your Company Working To Mitigate Those?
The major pain points that are emerging are mainly from inefficiencies in the workflow and the failure to offer patients comprehensive service all the time. Philips is putting a lot of effort in this regard. One key example is ourIQon Elite Spectral CT which can easily integrate with workflow and deliver unparalleled diagnostic quality leading to fast procedures and precision diagnosis. The always-on design, without increased radiation dose, and simple workflow solutions are important innovative distinctions of IQon Spectral CT over other spectral scanners.