Sharper images and CDS

Mary Stevens, editor, CMIO magazine

New techniques aimed at reducing radiation exposure and curbing duplicate tests are coming online, and the movement to measure cumulative medical radiation exposure in EHRs is gaining steam. Clinical decision support systems might seem like the logical place for dose/image conscientiousness to take root. That might stand to reason, but it isn’t always the case, according to a study in the April edition of the Journal of the American College of Radiology.

The call for clinical decision support (CDS) for imaging orders of all kinds has grown in response to concerns of overutilization and inappropriate imaging. Associations and providers in the U.S. are lobbying Medicare to consider electronic CDS, the study authors noted. In addition, expanded CDS is part of CMS’ recently released guidelines for accountable care organizations.

However, implementation of Canadian Association of Radiology (CAR)-based decision support in a pediatric hospital showed little relevance of guidelines to clinical cases and even poorer compliance, with physicians admitting that they “blatantly cheated” to avoid using the CDS software program, according to the study. The impact on duplicate orders was minimal—while the program alerted physicians to 367 identical studies performed within the previous 30 days, physicians heeded the program’s advice in just 40 cases. Physicians branded the CDS too generic and not relevant to their complex or high-risk patients.

Clearly, one CDS system won’t fit all imaging needs any more than one type of CDS can fit all non-imaging modes of clinical information. As the study authors stated, evaluation of any CDS implementation is a critical step.

Another round of visualization milestones saw cardiac SPECT cameras “undergoing a profound change in design for the first time in 50 years.” So stated researchers at Emory University recently in the Journal of Nuclear Medicine. Innovative design and advanced algorithms give newer SPECT cameras the ability to acquire images of the heart in less time and with better resolution than conventional SPECT cameras.

In another recent study, artificial intelligence (AI)-enabled neural networks demonstrated strong sensitivity and specificity in the detection of liver cancer using PET/CT. Interpreting radiologists who considered the AI network findings benefitted from improved accuracy in their interpretations, suggesting a future role for the systems in cancer detection, according to the study published in the March issue of Radiology.

Sharper images and more accurate interpretations will have a role alongside CDS in reducing duplicate imaging and lowering patients’ long-term radiation exposure. Where else does advanced visualization fit into the CDS picture? Tell me at mstevens@trimedmedia.com.

 

Mary Stevens, editor