Smart Reporting integrates LLM Technology built on AWS

Munich, 24th November 2023. In an era where medical knowledge evolves at an unprecedented pace, healthcare professionals are facing an ever-increasing workload. To address this challenge, Smart Reporting, a pioneer in synoptic medical reporting, has built an innovative solution integrating large language model (LLM) technology to suggest impressions in its reporting software SmartReports. Leveraging Amazon Bedrock’s generative AI capabilities, this solution is set to provide healthcare professionals with an enhanced reporting experience, ultimately improving patient outcomes. The innovative features are coming to the clinical workflow through popular report editors, including SmartReports, the synoptic reporting software from Smart Reporting. An outlook to the future solution will be shown at RSNA (Chicago, November 26-29). The new feature will be available in the first half of 2024 on the US market.

Radiology professionals often face two key challenges: variable and incomplete results from manually drafted reports, and non-machine-readable, unstructured findings that cannot be used for proactive analysis and improved treatment planning. With Amazon Bedrock, SmartReports will be able to offer:

• A significant reduction in words spoken or typed during report creation through automatic impression suggestion.
  
  
• Enhanced research and organizational efficiency with structured, machine-readable findings, resulting in 60% higher data quality for clinical decision-making and 150% improved report accuracy and detail, improving communication with referrers and other stakeholders.^{[1, 2]}

By prioritizing faster input and better output, better care can ultimately be achieved through optimized treatment planning, which has been previously demonstrated by a 4.3% reduction in patient mortality using structured reporting in pathology3. The new features are deeply integrated into the Smart Reporting portfolio: In the near future, US customers of SmartReports simply need to update their SmartReports installation to start using AI assistance (from the first half of 2024). An LLM is then used to process the information and provide AI-generated results for radiologist review, which can be easily referenced, edited, and finalized, all while safeguarding patient privacy.

Smart Reporting leads the way to greater efficiency in the healthcare system

“We are always looking for ways to make radiology reporting faster and more efficient,” said Prof. Wieland Sommer, founder and CEO of Smart Reporting. “We have been open to technology and vendors and are working agnostically to broadly integrate LLMs into our software. By building on Amazon Bedrock, we are able to leverage their state-of-the-art AI technology to enhance our radiology reporting solution. With automated impression suggestion, we will provide our customers with faster, better, and smarter reports that ultimately benefit patients.”

Smart Reporting will focus on evolving to receive and exploit the outputs of AI that advance report quality and reduce clinician workload. Future exploration into guideline-based decision support systems will progress streamlined reporting and referrals, while planned incorporation of consistency checks will help to ensure integrity of reporting. Smart Reporting’s adaptive solution upgrades clinical communication by keeping clinician needs and workflows at the forefront.

“I’m thrilled to see Smart Reporting bringing generative AI technology, powered by Amazon Bedrock, to the world of clinical reporting. Healthcare professionals can benefit from seamless and efficient AI-enhanced reporting experience, spend less time on documentation and focus more on delivery of person-centered healthcare with improved outcomes. “ – Dr. Razvan Ionasec, Tech Leader for Healthcare, Amazon Web Services EMEA SARL.

"Integrating cutting-edge AI into our reporting paradigm has been nothing short of revolutionary. It's a perfect synergy of speed and precision, letting us keep our gaze fixed on the heart of our profession: unparalleled patient care," shares Dr. Dominik Nörenberg, Associate Chair of Clinical Operations, University Medical Center Mannheim.

The outlook on LLM-powered automated impression suggestion will be demonstrated at both the Smart Reporting (South Hall A, Level 3, Booth 4953) and AWS (South Hall A, Level 3, Booth 4724) booths during RSNA in Chicago.

1 Blum SFU, Hertzschuch D, Langer E, et al. Routine Use of Structured Reporting in Whole-body Trauma CT Facilitates Quality Improvement. Rofo. 2023;195(6):521-528. doi:10.1055/a-2031-2364

2 Nörenberg D, Sommer WH, Thasler W, et al. Structured Reporting of Rectal Magnetic Resonance Imaging in Suspected Primary Rectal Cancer: Potential Benefits for Surgical Planning and Interdisciplinary Communication. Invest Radiol. 2017;52(4):232-239. doi:10.1097/RLI.0000000000000336

3 Sluijter CE, van Workum F, Wiggers T, et al. Improvement of Care in Patients With Colorectal Cancer: Influence of the Introduction of Standardized Structured Reporting for Pathology. JCO Clinical Cancer Informatics. 2019;(3):1-12. doi:10.1200/CCI.18.00104

About Smart Reporting

Smart Reporting was founded in 2014 by Wieland Sommer, Professor of Radiology with a Master's in Public Health (Harvard, US). The company was established in Munich with the vision of harnessing the value of medical data for better decision-making and treatment. Today, it has an interdisciplinary team of 70 doctors, data scientists and software engineers who develop software from Germany, Austria, Switzerland and Brazil based on their detailed understanding of clinical workflows. The company's multilingual software for structured diagnostics in radiology and pathology is used by more than 15 000 physicians in over 90 countries.

Press contact:

Thomas Doriath

Mail: t.doriath.ext@smart-reporting.com

Mobile: +49 176 24244283