AI-based assistants for magnetic resonance imaging

31 Dec 2019
Laura Sisman
Administrator / Office Personnel

The new applications in the Siemens Healthineers product family, AI-Rad Companion for the body region's brain and prostate, are designed to ease the burden of routine activities for radiologists and improve the quality of care.

Siemens Healthineers has introduced two software assistants based on artificial intelligence (AI) that are designed to free radiologists from the burden of performing routine activities during magnetic resonance imaging (MRI) examinations in the body region's brain and prostate. AI-Rad Companion Brain MR for Morphometry Analysis automatically segments the brain in MRI images, measures brain volume, and marks volume deviations in result tables used by neurologists for diagnosis and treatment.

AI-Rad Companion Prostate MR for Biopsy Support automatically segments the outer contour of the prostate on MRI images and enables radiologists to mark lesions, so that it’s easier for their colleagues in Urology to perform targeted prostate biopsies.

Both new applications can be used on MRI scanners from different manufacturers and are available on teamplay, the cloud-based healthcare platform from Siemens Healthineers.

“With the new AI-based assistants, we are expanding our diagnostic offering to help our customers increase efficiency and improve the quality of care. We firmly believe that AI will help physicians deal with their workload and benefit patients by helping achieve an improved, patient-focused decision-making process,” said Peter Koerte, Head of Digital Health at Siemens Healthineers. “We demonstrated this when we introduced the first application of the AI-Rad Companion focusing on CT for the chest body region. At RSNA 2019, we will present additional AI-Rad Companion applications for MRI examinations in the body regions, brain and prostate. Further applications for radiography and radio oncology will follow for the AI-Rad Companion family.”

AI-Rad Companion Brain MR for Morphometry Analysis

The AI-Rad Companion Brain MR for Morphometry Analysis supports brain volumetry, which is performed in clinical practice for example in cases of suspected dementia. This involves measuring the volume of gray matter (nerve cells), white matter (nerve cell connections), and cerebrospinal fluid in various segments of the brain and comparing the results to normal volumes. A tendency for the brain volume to shrink may point to diseases like Parkinson’s or Alzheimer’s and other dementia diseases. Previously, segmentation and comparison to the norm were performed manually or only semi-automatically.

Based on AI algorithms, the AI-Rad Companion Brain MR for Morphometry Analysis can automatically identify about 50 brain segments on MRI images, measure their volumes, and compare the results to data in a normative reference database for brain morphometry made available by the Alzheimer’s Disease Neuroimaging Initiative (ADNI). AI-Rad Companion Brain MR for Morphometry Analysis feeds the results into a report where deviations from the norm are automatically marked. This means that radiologists can provide the neurology department with objective, quantitative data that’s relevant to differential diagnosis and therapy management. Patient care can also be improved by generating reports faster and in a more standardized and evidence-based form.

According to Alzheimer’s Disease International, more than 50 million people around the world are living with some form of dementia. By 2030, that total could exceed 75 million and as many as 131.5 million by 2050 – which means that it’ll be increasingly important to identify dementia diseases at an early stage in order to slow their progress with the appropriate treatments, if possible.

AI-Rad Companion Prostate MR for Biopsy Support

Prostate cancer is the second-most common malignant cancer (after lung cancer) affecting males world-wide. Both the European Association of Urology (EAU) and the National Institute of Clinical Excellence (NICE), UK, provide evidence-based guidelines for diagnosing and treating prostate cancer, and in the past 12 months, they’ve incorporated the primary diagnosis of prostate cancer using MRI and MRI/ultrasound fusion biopsy in their guidelines.

Previously, urologists had to identify the suspected cancer areas using written reports and pictograms of the prostate provided by the radiologist based on the MRI examinations, and then try to take tissue samples accurately from the areas in question using ultrasound-guided biopsy. For the fusion procedure, the prostate had to be manually segmented in the MRI images, which can take up to five minutes per patient.

AI-Rad Companion Prostate MR for Biopsy Support automatically segments the outer contour of the prostate, which can cut the time needed for this routine activity down to just a few seconds. The radiologist then simply marks the suspect areas and hands the annotated MRI images to the urologist for fusion with the ultrasound images during the biopsy. Targeted, MRI-supported biopsies like this can make it easier for the urologist to detect significant prostate carcinomas and improve the quality of patient care.

The work of healthcare providers in recent years has become subject to significant time and cost pressures, and that goes for radiologists as well. In some conurbations, radiologists have to interpret a new image every three to four seconds throughout their workday. This means not only stress at the workplace but also that it is becoming increasingly harder to keep the quality of diagnostic decision-making consistently high. Siemens Healthineers aims to take some of the load off radiologists with its AI-based assistants.

The new applications in the product family AI-Rad Companion can be used on MRI scanners from different manufacturers. They are cloud-based and use the certified, secure teamplay infrastructure, which is compliant with the HIPAA (Health Information Portability and Accountability Act, USA) and the EU GDPR (General Data Protection Regulation). The software is seamlessly integrated into the existing clinical workflow and complies with DICOM (Digital Imaging and Communications in Medicine) standards. The clinical images and all supporting information can be made automatically available in the picture archiving and communication system (PACS) depending on the radiologists’ requirements.

Want more of the latest science news straight to your inbox? Become a SelectScience member for free today>>

Links

Tags