CME Radiology Grand Rounds

Radiology practices have benefited from the adoption of electronic-based information technology, especially with respect to practice efficiency. However, electronic tools, such as PACs, RIS, and speech recognition (along with their associated workflow), are still relatively immature and arguably support only “commodity-level” capability. Accordingly, these technologies can and have been exploited to commoditize and “outsource” radiology services. There is a critical need for a new generation of “meaningful innovation” in radiology IT that will allow radiology to maximize value to patients and other stakeholders throughout the enterprise. However, unless we are willing to dramatically re-engineer our own attitudes and practices, we will not only fail to effectively use these advanced electronic tools, we will facilitate the perceived devaluation of radiology and participate in its marginalization and commoditization. Radiologists must be “value innovators” who maximally leverage information technology to ensure their relevance and value to patient care through measureable improvements in quality, efficiency, and safety.

The world of medical imaging is changing. Tasks that were once the sole domain of meticulously-engineered hardware are now beginning to be accomplished in software, increasingly informed by diverse arrays of inexpensive auxiliary sensors.  Information once pursued through the laborious acquisition of carefully separated datasets is now being derived from newly integrated, and richly quantitative, data streams. It is a time of remarkable creative flux, with long-held assumptions about the underpinnings of imaging technology and practice beginning to be questioned in earnest. It is a perfect time, in short, to reimagine imaging.  This talk will survey some of the disruptive forces and disruptive innovations which currently promise to change the value proposition of biomedical imaging.  Taking specific examples from magnetic resonance imaging, with selected generalizations to other imaging modalities, the talk will explore three noteworthy trends:

1. an emerging paradigm of rapid, continuous, comprehensive imaging;
2. the use of artificial intelligence not only to interpret images but also to accelerate and otherwise modify image acquisition and reconstruction; and
3. early indications of a shift from large, carefully-calibrated scanners to flexible, self-calibrating, multi-sensor imaging devices.

IPF manifests in peripheral and basal lung regions. Accepted dogma relates scarred regions to peripheral deposition of fibrous tissue. Contrary evidence demonstrates a uniformly injured lung with collapsed alveoli surrounding dilated airways. Distribution of collapse can be related to gravitational gradients and airway closure in an inflamed and aging lung.

Big data like that found in Electronic Health Records (EHRs) and personal genotypes has the potential to revolutionize the care we provide to patients. For example, risk factors like demographic variables, environmental exposures, healthcare data, germline genetic variants (single nucleotide polymorphisms - SNPs) and intermediate phenotypes, like mammographic breast density, have a better chance of stratifying risk and targeting interventions than any of these data types alone. An intermediate phenotype, a measureable, but clinically unobservable phenotype falling somewhere between an illness genotype and phenotype is a conceptual framework that has the potential to increase the power of imaging.  Discovering and exploiting intermediate phenotypes observable only on imaging for diagnosis and risk stratifications is a paradigm shift that could profoundly impact the healthcare we provide. Our team has developed risk prediction tools that combine demographic risk factors, germline genetic variants, and mammographic features using novel machine learning methods. As one of the site PIs of the All of Us—Wisconsin initiative (https://allofus.wisc.edu/), our team is working to determine how imaging can most powerfully enable the transition to personalized medicine and precision health.

LIRADS is recently expanded to incorporate contrast enhanced ultrasound (CEUS).  LI-RADS 5 has an objective of 100% specificity for the confident diagnosis of HCC so as to allow treatment of HCC without the requirement for invasive biopsy.    Therefore its requisite features are stringent and include all three of the following:  nodule size greater than 10 mm, arterial phase hyperenhancement (APHE) and late and weak washout, occurring after 1 minute and most often after 3 minutes.    LR-M is an essential category of probably malignant tumors with malignant enhancement features, but without specificity for HCC.   LR-M categorization requires any of the following features on CEUS:  arterial phase rim enhancement, rapid washout in less than 60s, and marked washout, evident prior to 2minutes following contrast injection.   LR-M will detect virtually all nonhepatocellular carcinomas and some HCC that do not meet the requirements for LR-5.  Lack of specificity of LR-M features requires biopsy for confirmation of these cases.  

DeMartini:  Many countries and healthcare systems are navigating transitions to value-over volume-based care practices. In general, value is defined as health outcomes achieved relative to associated costs. Value-based models emphasize and reward activities that measurably contribute to value by improving outcomes or decreasing costs, typically for specific diseases and for episodes-of-care. In radiology, activities that contribute value include using tests that have definitively been shown to improve patient outcomes, interpreting these tests with high accuracy, and generating reports that are clear and actionable. In breast imaging, there is a strong foundation in value-directed activities. These include using screening mammography that improves outcomes by decreasing breast cancer mortality, auditing diagnostic performance to measure and maintain accuracy, and generating reports using the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS). This talk will discuss key current and future considerations for breast imaging in leading value-based patient care.

Pitteri: Early detection biomarkers have the promise to improve cancer diagnostics and decrease patient mortality by identifying disease at an earlier and more curable stage.  In vitro biomarkers, for example measured in the blood or tissue, also have the potential to compliment in vivo imaging approaches for cancer diagnosis.  However, despite the promise of such biomarkers, there have been relatively few new biomarkers used in the clinic over the past several decades.  This talk will describe the different classes of molecules that are being used and explored as cancer biomarkers.  We will also discuss the process to develop a clinically used biomarker and what makes a good biomarker, as well as the associated challenges.  Finally, new opportunities, strategies, and approaches in biomarker development will be described.

Segall:  The four parts of the ABR Maintenance of Certification Program will be discussed. Recent changes to improve the program so that it is both more meaningful and less obtrusive into the lives of radiologists will be emphasized.

Donnelly:  The four parts of the ABR Maintenance of Certification Program will be discussed.   Recent changes to improve the program so that it is both more meaningful and less obtrusive into the lives of radiologists will be emphasized.

Conveying information via presentations is a critical skill used very frequently in academic medicine, particularly in the image-rich field of Radiology. This presentation on how to on give great talks will review the ideal structure of presentations (whether it be for education, research, or otherwise) as well as practical tips for deciding what content needs to be present, assembling effective slides, practicing and delivery.

Gathering storm clouds indicate continued turbulence in diagnostic radiology worldwide. The incidence of professional burnout among radiologists continues to increase, work dissatisfaction is driving more radiologists out of the field, practice stressors and workloads continue to increase at a staggering pace, and our work style and environments need improving. Recent research on burnout has suggested a two-pronged attack is necessary to adequately address physician burnout. First is an effective personal strategy for maintaining health and wellness, but that alone is not sufficient. The second attack has to be at the department, hospital, and health system levels to engage physician leaders, department chairs, and system administrators to work together to change the entrenched culture of medicine to a healthier one.

This presentation will discuss the factors driving radiologists” work dissatisfaction, the incidence and implications for health and career of professional burnout, and provide effective personal strategies to avoid burnout. These strategies will focus primarily on improving radiologists’ work-life balance in the six Domains of Balance. Finally, I will suggest a number of challenging interventions to department environments and leadership priorities designed to change the Culture of Radiology, and in so doing, save our specialty.

Imaging biomarkers have a substantial contribution in detecting, localizing, and determining the extent of disease. Identifying imaging biomarkers and risk factors of pre-symptomatic disease and development of predictive models for functional and structural alterations requires large, prospective epidemiological studies in unselected populations in which sub-clinical pathology can be assessed. Population imaging addresses this question by providing analysis of medical images at a large scale in controlled population cohorts.  

For instance, in neurodegenerative disease population imaging has allowed to identify a number of imaging biomarkers like regional brain volumes, distribution and quantification of white matter lesions, subclinical brain infarcts or microbleeds as well as the structural and microstructural integrity of the white matter associated with the development of mild cognitive impairment and full-blown dementia long before any symptoms were present. For example new MR imaging sequences allowing to increase the conspicuity of hemosiderin deposits showed a 25% prevalence of microbleeds in the elderly asymptomatic population. Microbleeds are thought to reflect vessel fragility, on the basis of either hypertensive vascular damage or amyloid angiopathy.

Moreover, common genetic variants could be found that contribute to explain variance in imaging phenotypes associated with dementia by a systematic analysis of the genome based on genome-wide association studies (GWAS). The power of GWA analyses has been recently demonstrated with the identification of susceptibility genes involved in dementia-associated imaging phenotypes like  intracranial volumes, or size of the hippocampus. These associations are highlighting new etiological pathways and are expected to improve the understanding of the molecular basis of neurodegenerative diseases.