Imaging Informatics

The 2017 Liver Imaging Reporting and Data System (LI-RADS) included an algorithm to assess hepatocellular carcinoma (HCC) treated with local-regional therapy, but the method was never validated. A team of researchers took on the task, publishing their results April 30 in Radiology.

“If (virtual non-contrast) VNC images could provide comparable diagnostic performance in detecting biliary stone disease to (true non-contrast) TNC, we might replace TNC with VNC, and consequently, reduce the radiation dose,” wrote authors of a new study published in the European Journal of Radiology.

The supply of radiation oncologists hasn’t kept up with the global demand for radiation therapy. But could experts from across the world help create an AI algorithm capable of closing that gap?

HistoQC, created by bioengineering researchers at Case Western University, is an open-source quality-control tool that helps users flag low-quality images while preserving those that can help clinicians make accurate diagnoses.

A clinical decision support (CDS) tool designed to limit inappropriate high-cost imaging reduced targeted scans by 6%, reported authors of a study published in PLOS One. The results may be useful for creating a more efficient tool given the upcoming CDS mandate in 2020.

What happens when a software virus runs up against a physician trained to spot ailments? In a recent study, malware designed to create fake nodules on images successfully fooled radiologists into making incorrect diagnoses.

Interventional radiologists at UC-San Francisco have adopted the popular smart-speaker system Google Home (aka “Hey Google,” “OK Google”) for use in the surgical suite.

When it comes to digital medicine, digital pathology is very late to the game. But its time is coming. And the benefits could be many: Bolstering the capabilities, efficiency and reach of individual pathologists, cutting patient wait times, streamlining multidisciplinary team meetings (MDTs) and offering more data-rich decision-making. It could even obviate a shortage of pathologists. Where does it fit into your strategic plan?

Researchers looked at data from more than 2,500 free-text radiology reports of patients undergoing hepatocellular carcinoma surveillance to determine if an NLP approach could extract clinical data and predict downstream utilization of resources.

Documenting contrast allergies in the electronic health record (EHR) is central for safe imaging, but a new study found such records are incomplete, misleading and often ambiguous. A multidisciplinary approach may be needed to solve the problem.

Educating radiologists on ACR-recommended follow-up for incidental adnexal lesions and incorporating such guidelines into normal workflow significantly improved the rate of adherence, reported authors of a Feb. 26 study published in the American Journal of Roentgenology.

“Our results show that worklists organized by relative individual interpretation times can decrease the overall group interpretation time in a multireader setting,” wrote authors of a Feb. 26 study published in the the American Journal of Roentgenology.