RSNA 2018 Imaging Informatics Preview
By Erik L. Ridley, AuntMinnie staff writer
November 8, 2018

Our final destination on the Road to RSNA is imaging informatics and specifically PACS, cybersecurity, structured reporting, clinical decision support (CDS), radiation dose-monitoring software, analytics, and issues regarding patient access to radiology results. For previews of papers on artificial intelligence or advanced visualization topics, please also check out our Road to RSNA Artificial Intelligence and Advanced Visualization previews.

Researchers will come to RSNA 2018 ready to discuss ways to improve on the radiologist's primary product: the radiology report. A number of presentations in Chicago will highlight, for example, the potential for multimedia structured reports to help radiologists produce higher-quality reports and automate assessment of categories in BI-RADS and similar classification systems. One study found that radiologists will frequently utilize interactive multimedia reports if they have the opportunity.

Furthermore, interventional radiologists may have some useful wisdom to offer for creating reports that satisfy referring physicians, according to another presentation. In addition, researchers will explain why radiology reports are often too complicated for the average patient to comprehend. Speaking of patients, one study will show that there's room for improvement in terms of patients accessing their radiology results through online portals.

Clinical decision support remains a timely topic, especially in light of the recent final rule by the U.S. Centers for Medicare and Medicaid Services (CMS) to finalize its plans for CDS; effective January 1, 2020, ordering providers will be mandated to consult appropriate use criteria during ordering of advanced imaging studies for Medicare patients. At McCormick Place, researchers will discuss topics such as lessons learned from a CDS pilot initiative and how a CDS initiative was able to decrease the number of duplicative abdominal ultrasound studies.

Radiation dose monitoring remains a hot area. Presenters will highlight, for example, how standardized reporting templates can ensure compliance with an enterprise-wide dose reduction initiative and how an informatics platform can monitor both radiation dose and image quality.

In other applications for analytics, natural language processing techniques can facilitate correlation of radiology and pathology results, as well as help determine the need for CT contrast, according to two presentations. It can also search CT reports for mentions of incidental lung nodules, enabling analysis of compliance with Fleischner Society guidelines.

Cybersecurity will always be a big concern in today's world. However, financial penalties currently in place for healthcare data breaches in the U.S. may have actually led to an increased risk for loss of protected health information, according to a scheduled talk at the meeting. In addition, another presentation will review the potential -- and limitations -- for blockchain technology in radiology.

See below for our previews of select imaging informatics-related scientific papers and posters at RSNA 2018. Of course, these are just a sample of the imaging informatics content planned for the meeting. For more information on those talks, as well as other abstracts and refresher courses, view the RSNA 2018 meeting program.

Scientific and Educational Presentations
Structured reporting system enhances report quality
Sunday, November 25 | 12:30 p.m.-1:00 p.m. | IN009-EC-SUA | Lakeside, IN Community, Custom Application Computer Demonstration
In this poster presentation, researchers from Texas will detail how their multimedia structured reporting software can help radiologists produce better reports.
Interventional radiologists offer wisdom on reporting
Sunday, November 25 | 12:30 p.m.-1:00 p.m. | IN140-ED-SUA1 | Lakeside, IN Community, Station 1
Interventional radiologists have a unique perspective on radiology reporting that could benefit diagnostic radiologists, according to this poster presentation.
CDS initiative cuts duplicative abdominal ultrasounds
Sunday, November 25 | 12:30 p.m.-1:00 p.m. | QI102-ED-SUA6 | Lakeside, QR Community, Station 6
Researchers from Baltimore will share in this poster presentation how a clinical decision-support (CDS) initiative led to fewer unnecessary abdominal ultrasound exams.
Blockchain has both potential and shortcomings in radiology
Sunday, November 25 | 1:00 p.m.-1:30 p.m. | IN141-ED-SUB1 | Lakeside, IN Community, Station 1
Blockchain technology shows promise for a number of radiology applications, but it also has important limitations, according to this poster presentation.
Initiative confirms documentation of dose reduction
Monday, November 26 | 12:45 p.m.-1:15 p.m. | QI008-EB-MOB | Lakeside, QR Community, Hardcopy Backboard
In this poster presentation, researchers from MedStar Health will discuss how automated evaluation of CT reports has helped ensure documentation of an enterprise-wide dose reduction program.
Follow-up system helps ensure workup of incidental findings
Tuesday, November 27 | 9:40 a.m.-9:50 a.m. | RC327-04 | Room N229
In this talk, researchers from Boston will describe how they used RIS and electronic medical record searches to track follow-up on radiologist recommendations for incidental findings.
NLP can find incidental lung nodules on CT reports
Tuesday, November 27 | 11:50 a.m.-12:00 p.m. | RC327-14 | Room N229
A natural language processing (NLP) software tool can identify incidental lung nodules included in CT reports, enabling the analysis of compliance with Fleischner Society guidelines, according to researchers from New York City.
Patients infrequently view radiology reports on portals
Tuesday, November 27 | 12:15 p.m.-12:45 p.m. | IN214-SD-TUA3 | Lakeside, IN Community, Station 3
In this poster presentation, researchers from Alabama will discuss how only a minority of patients who logged on to their institution's online patient portal did so to read radiology reports.
Radiologists value interactive multimedia reports
Tuesday, November 27 | 12:45 p.m.-1:15 p.m. | IN219-SD-TUB2 | Lakeside, IN Community, Station 2
When given the opportunity, radiologists will commonly create interactive multimedia radiology reports that include hyperlinks to key imaging findings, according to this poster presentation.
Point-of-care photos reduce wrong-patient errors
Tuesday, November 27 | 3:00 p.m.-3:10 p.m. | SSJ13-01 | Room N230B
Photographs taken of patients at the point of care while they receive portable radiographs can reduce wrong-patient errors and even provide clinical context to aid in interpretation, according to this presentation.
Informatics platform monitors dose, image quality
Tuesday, November 27 | 3:00 p.m.-3:10 p.m. | SSJ21-01 | Room N229
Researchers from Duke University will describe how their informatics platform can monitor radiation dose and image quality from CT, mammography, radiography, and fluoroscopy exams.
Data breach penalties in U.S. have unintended effect
Tuesday, November 27 | 3:30 p.m.-3:40 p.m. | SSJ13-04 | Room N230B
In this talk, a multinational team of researchers will make the case that financial penalties currently imposed on healthcare organizations in the U.S. for data breaches have actually made IT systems more vulnerable.
Imaging CDS pilot reveals factors for success
Wednesday, November 28 | 11:50 a.m.-12:00 p.m. | SSK11-09 | Room S104A
Pennsylvania researchers learned some valuable lessons when they implemented imaging clinical decision-support (CDS) software, which they'll share in this Wednesday talk.
Software automates RADS assessment categorization
Thursday, November 29 | 10:30 a.m.-10:40 a.m. | SSQ11-01 | Room S103AB
In this presentation, researchers will describe how a multimedia structured reporting software application can automate the determination of assessment categories for BI-RADS and similar schemas.
Software standardizes proprietary imaging annotations
Thursday, November 29 | 11:10 a.m.-11:20 a.m. | SSQ11-05 | Room S103AB
Imaging annotations stored in a proprietary fashion can be converted into a standardized format, enabling interoperability and automated lesion matching across sequential imaging studies, according to this presentation.
Machine-learning software predicts need for CT contrast
Thursday, November 29 | 11:20 a.m.-11:30 a.m. | SSQ11-06 | Room S103AB
In this talk, researchers from California will present their machine learning-based method for automatically determining if contrast is needed for a CT exam.
Radiation dose analytics improves patient safety
Thursday, November 29 | 11:20 a.m.-11:30 a.m. | SSQ18-06 | Room N228
Retrospective quantitative assessment of radiation dose on CT studies can yield valuable insights for improving consistency and patient safety, researchers from Duke University have found.
Software enables use of quantitative imaging biomarkers
Thursday, November 29 | 11:30 a.m.-11:40 a.m. | SSQ11-07 | Room S103AB
A group from Stanford University will showcase a software plug-in that can make it easier for radiologists to assess quantitative imaging biomarkers when evaluating patient response to cancer treatment.
Natural language processing yields rad-path correlation
Thursday, November 29 | 11:40 a.m.-11:50 a.m. | SSQ11-08 | Room S103AB
Researchers from California will reveal how natural language processing technology can facilitate the correlation between radiologic and pathologic findings.
Patients often struggle to comprehend radiology reports
Thursday, November 29 | 12:45 p.m.-1:15 p.m. | IN235-SD-THB1 | Lakeside, IN Community, Station 1
Neuroradiology CT and MRI reports are written at a level too high for the average patient to understand, according to this poster presentation.