Underutilization of Reporting Automation in Radiology: Missed Opportunity?

Innovation, such as the large language model ChatGPT passing Radiology Board Exams, continues to reshape industries. However, preliminary findings suggest Radiologists are not taking full advantage of reporting automation tools that exist today. Automation tools transfer measurements and findings from imaging modalities into a reporting package, resulting in fundamental improvements in reporting efficiency, accuracy, and quality. It seems an easy argument that we should capture “easy” benefits today [MSR White Paper].

To better understand the current state of reporting automation, I would appreciate the few seconds to take this short, 3-question survey. Reporting Automation Survey

By raising awareness of the benefits and addressing concerns or barriers to adoption, reporting automation tools can be promoted to improve the content and clarity of more radiology reports. To this end, common Reporting Automation topics are discussed below.

Structured Reporting

A good starting point for a discussion regarding Reporting Automation is Structured Reporting since both focus on improving Radiology Reporting. Structured Reporting is an approach to creating reports in a predefined format with standardized terminology and checklists. Many institutions are adopting Structured Reporting to improve report content, clarity, and efficiency. Studies show the benefits of Structured Reporting in improving clinical outcomes, including reducing missed diagnoses [Link].

Though Reporting Automation efforts, such as automatic measurement transfer, are additive in achieving similar improvement goals, Reporting Automation is often overlooked as part of a Structured Reporting effort. The reasons for neglecting this significant opportunity are not clear. More than Structured Reporting, Reporting Automation can yield a Return on Investment (ROI) of just a few months! Additionally, Reporting Automation implementation resources are often outsourced. It seems an organization that implements Structured Reporting should implement Reporting Automation tools.

DICOM SR

The most common reporting automation tool in radiology is measurement transfer from the imaging modality to the reporting package, with ultrasound and DEXA being the most common modalities. Measurements are transferred from the imaging modality to the reporting system using DICOM Structured Reporting (DICOM SR). Companies such as Imorgon have developed products that consume the DICOM SR file and transfer it into the reporting package, typically Nuance Powerscribe. Some institutions reduce tedious and error-prone numeric dictation by transferring thousands of measurements daily. Documented studies talk about median reporting time reductions of 40% measurement intensive reports such as OB [Dr Horii SIIM Presentation].

Ultrasound Worksheets

Another reporting automation strategy is electronic sonographer worksheets and templates. Hundreds of publications and articles guide what constitutes a good ultrasound report. Many studies report positive outcomes using standardized ultrasound report templates, including a reduction in the use of CT [Link]. With tools like Imorgon’s, ALL sonographers can provide exceptionally clear, concise, and consistent observations directly into the reporting package. Newer devices with touch-selectable checkboxes on tablets running electronic worksheets can further streamline the reporting process.

Reporting automation is ideally implemented with both automated measurement transfer and electronic worksheet observations. Depending on the exam, more sonographer observations than measurements are transferred.

Reporting Automation Future

Looking to the future, there are a couple of Reporting Automation directions. The newest area of attention has been the Large Language Models and the potential impact on impression generation. It will be exciting to see these developments. Report Automation tools, such as those developed by Imorgon, complement these AI tools by providing consistent and standardized input. Another initiative is the development of Common Data Elements [Link], which aims to standardize and increase the clinical content of transferred information, facilitate data extraction, and improve decision support tools. Implementing a standardized report content process today can enable content-generating AI to be more consistent, leading to the development of better tools.

In conclusion, adopting Reporting Automation tools in Radiology improves efficiency and accuracy in reporting, leading to better patient care today and in the future.

Again, don’t forget to participate in our survey to help understand the current state of reporting automation in radiology. Reporting Automation Survey