emtelligent, a developer of clinical-grade natural language processing (NLP) software for healthcare organizations, announces the continued expansion of its emtelliSuite family of apps.
The emtelliSuite collection of apps is designed to improve patient safety and care quality by unlocking hard-to-reach medical data. emtelliSuite leverages emtelligent’s emtelliPro NLP engine, a proprietary and purpose-built, enterprise-scale platform that enables actionable clinical insights from unstructured medical text.
Attendees of RSNA 2022, the Radiological Society of North America’s annual global forum held Nov. 27 to Dec. 1 in Chicago, were able to try the emtelliPro and emtelliSuite apps at emtelligent’s booth 4449 in the South Exhibit Hall at McCormick Place.
emtelliSuite apps and the emtelliPro medical NLP platform improve efficiency and effectiveness for care providers, researchers, and administrators. Along with its existing apps, including smart searching, report reading, patient summary tools, some of the radiology-specific apps now included in emtelliSuite are:
- AI Concordance – Classifies radiologist reports as positive or negative for specific pathologies, for comparison with image analysis AI model output to quantify and track the accuracy of AI models over time
- Case Auditor – Identifies case results as positive or negative for their stated indication, to track referring physician accuracy for assistance in implementing decision support and ensuring high-quality care
- Critical Results Auditor – Identify cases by diagnosis where the radiologist has not included a statement of critical results communication to ensure adherence with departmental guidelines
- Follow-Up Detector – Automatically flag actionable recommendations or incidental findings from within radiology reports to ensure timely adherence and follow-up
- Indication Auditor – Find reports with low-quality indications to have evidence for providing feedback to referring clinicians to improve quality of care
- MRI Screener – Find patients with medical devices or other conditions that may be ineligible for MRI procedures before they get to the MRI suite to prevent un-filled timeslots and ensure high equipment utilization rates
- Trainee Scoreboard – Automatically tracks the number of actual diagnoses (rather than just exam counts) seen or made by a trainee to 1) improve educational outcomes and 2) ensure that knowledge “blind spots” are avoided.
AI Concordance Reporting
As vendors are now widely selling AI models with promises of high accuracy, hospitals, clinics, and AI vendors are discovering the need to track AI model performance. Differences in imaging equipment, image protocols, and patient factors can mean a significant gap between as-advertised and real-world performance of AI models. Tracking performance of these models in production settings is important to ensure value and patient safety.
emtelliPro can automatically classify radiologist reports as being positive or negative for specific pathologies, and these results can be compared to AI model output to produce concordance reports. These can be run regularly or on-demand to accurately assess AI model performance in a hospital or clinic’s real-world environment while helping to ensure patients are getting the care they need and creating added value for the medical imaging department.
“Radiologists struggling with high volumes of work and limited resources need help in quickly finding relevant clinical information for their patients,” said Dr. Tim O’Connell, emtelligent CEO and co-founder and a practicing radiologist in British Columbia. “emtelliPro and emtelliSuite give them the technology tools to improve patient safety and quality of care. We’re excited to have RSNA attendees try our solutions at the show.”
[Source(s): emtelligent, PR Newswire]