SimonMed, headquartered in Scottsdale, Arizona, is one of the largest providers of outpatient medical imaging and radiology practices in the United States. It covers 10 states with more than 150 locations and more than 200 subspecialty trained radiologists.
SimonMed offers the full modality of diagnostic scans including 3T MRI, CT, Ultrasound, 3D Mammography, PET/CT, Nuclear Medicine, DEXA, X-rays and others.
Answering important questions
The healthcare organization looked to AI tools to improve patient care and improve diagnostic tools.
“For example, we are now using AI ‘triage’ programs like AI to identify potential fractures or pneumonia,” said Dr. John Simon, CEO of SimonMed. “These cases are marked STAT and move up the radiology read queue so they can generally be read in less than 10 minutes. This ability to triage fundamentally resets expectations of turnaround for positive cases and will continue to other studies.
“In terms of improved diagnostic tools, 3D breast AI provides a second reading for our 3D mammography program, ensuring excellent sensitivity and specificity from our team,” he continued. “Another way AI is improving our diagnostic tools is identifying findings that are not necessarily visible to the radiologist, such as mapping asymmetric brain volume loss or disrupted white matter tracts or calculating blood flow in coronary arteries.”
Before implementing the AI tools, the common questions SimonMed employees asked were:
-
How can we ensure that positive cases are identified and that clinical providers and patients are notified more quickly?
-
How can we improve our sensitivity and specificity in our reporting?
-
Are there additional quantitative or pattern data that are not visible to the radiologist, but are present on the imaging?
-
How can we compare more accurately with previous exams?
Guarantee access
“Our approach with AI is to integrate it into radiologists’ workflow, but ensure they have access to the powerful technology that underpins these programs,” Simon explains. “Some programs offer simple image overlays, but the trend with more advanced programs is to provide a platform for the specific program.
“We use these programs to ensure that the AI findings are optimized for each patient,” he continued. “We are also introducing this technology to referring healthcare providers. For example, volumetric measurements of lung nodules are generally more sensitive than 2D measurements in evaluating change, but many of the standardized reporting protocols focus on 2D, while we have known for years that volumetric is generally more accurate.”
Properly integrating this new knowledge into clinical care will require collaboration between teams, he added.
Success statistics
SimonMed conducted a study on the results so far. Thanks to the AI tools, radiology reports have become approximately 82% faster than measurements without the automation, Simon reported. The performance analysis portion of the study to evaluate quality used 1,442 patients across 14 SimonMed centers to determine that the technology has a sensitivity range of 96.9% to 100% per bone.
Other provider organizations looking to work with similar AI should look at the technology for both triage and diagnosis, Simon advised.
“The requirements will be slightly different and with different KPIs,” he concluded. “In terms of improved diagnoses, the tools can be truly remarkable, so it is important to be open-minded and curious as this is a rapidly evolving field.”
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.