MSKai, a developer of Artificial Intelligence-based radiological assist software for the musculoskeletal system, announces the release of the company’s Lumbar Spine MRI segmentation and measurement algorithm in conjunction with submission of a United States Provisional Patent application for “Systems and Methods for Improving Patient Outcomes for Musculoskeletal Care.”
“Artificial Intelligence is rapidly integrating into healthcare,” said Adam Bruggeman, MD, chief medical officer of MSKai. “The ability to accurately and consistently identify and measure spine anatomy will lead to further efficiencies for patients and providers in the practice of spine surgery.”
The current software is a radiological assist system for the evaluation of Lumbar Spine pathologies providing spine care practicitioners the ability to identify, measure, and classify spinal anatomy and pathologies from a quantitative and qualitative standpoint within seconds. The distinctive machine learning algorithm provides anatomical segmentation, anatomical measurement, pathology classification, and range threshold alerts.
The initial validation for MSKai’s Lumbar Spine MRI segmentation and measurement is derived from 413 patient MRI studies. When comparing ground truth data from six independent medical and biomechanical professionals, the measurement accuracy and segmentation accuracy range from 76% to 90%, depending on the focus. The current segmentation and measurement algorithm continues to improve with consistency and sets the framework for the evolution of machine learning-based radiological assist systems being integrated into clinical practice.