The U.S. FDA has greenlighted two new modules for Siemens Healthineers’ AI-Rad Companion platform, giving radiologists new artificial intelligence (AI) tools to assist in interpreting magnetic resonance imaging (MRI) studies of the brain and prostate. The AI-Rad Companion Brain MR for Morphometry Analysis and AI-Rad Companion Prostate MR for Biopsy Support join earlier intelligent software modules for Chest CT.
The AI-based brain MR software assistant provides quantitative information on brain tissue volumes – including the nerve cells or gray matter (nerve cells), nerve cell connections, known as while matter, and cerebrospinal fluid – in various segments of the brain and comparing them to what are considered normal volumes. Evidence of reduced brain volumetry may be a sign of Parkinson’s disease, Alzheimer’s disease or other forms of dementia.
Using AI algorithms, the module automatically identifies about 30 different brain segments in MRI images and compares them to a reference database of normal images for brain morphometry from the Alzheimer’s Disease Neuroimaging Initiative. The results are then entered in a report where deviations from the norm are automatically highlighted.
The Prostate MR for Biopsy Support module automatically segments the prostate and delineates the outer shape of the organ in seconds vs. minutes with manual segmentation. The radiologist marks any suspect areas and sends the annotated MR images to the urologist, where they are melded with ultrasound images during biopsies to help detect malignancies and improve patient care.
Both modules are designed to be compatible with across multiple vendors’ MRI machines.
“These new AI-Rad Companion applications for MR exams in the brain and prostate regions will help physicians manage their workloads and achieve a patient-focused decisionmaking process to increase efficiency and improve the quality of care,” said Peter Shen, vice president of innovation and digital business at Siemens Healthcare North America, part of Erlangen, Germany-based Siemens Healthineers AG.
Siemens developed and trained the algorithms for the new AI-Rad Companion modules with the aid of a large, in-house supercomputer, which includes a repository of close to a billion curated studies, Shen told BioWorld. “What that means is not only do we have image sets that we can utilize for AI algorithm training and utilize different methodologies like deep learning or what not, but because they’re curated we also have the clinical outcomes that are associated with those particular images,” he said. In the case of prostate, that information can aid in both biopsy and treatment planning.
Shen said that having the clinical history associated with the algorithm training, as well as the images themselves, was helpful in getting FDA clearance quickly for the new offerings.
Expanding AI platform
The brain and prostate modules are the latest additions to the AI-Rad Companion family. The FDA cleared the AI-Rad Companion Chest CT engine, as well as pulmonary and cardiac modules, in September 2019. The final component of Chest CT – musculoskeletal/spine – received FDA clearance in March. As with the two MR modules, the Chest CT modules are vendor-agnostic.
The initial focus on chest CT and expansion to brain and prostate AI is indicative of what Siemens has in store for future development on the platform.
“We’re really trying to be this multimodality, multiorgan, vendor-agnostic AI diagnostic tool for our customers,” Shen said. Future applications could include modalities beyond CT and MR, and multiple organs or regions of the body, including the extremities, he added.
Siemens sees an opportunity to combine the MR AI applications with other of its MR solutions that are focused on brain and prostate, including ones that are sold in conjunction with a therapy-type solution, Shen said.
The entire AI-Rad Companion family is currently CE-marked. “We’ve had strong adoption and success of the solution not only in the U.S. market but in the markets in Europe and Asia as well,” he said.