- My News
- All News
- Most Popular
USARAD Holdings Inc., the leading US Joint Commission Accredited teleradiology and telemedicine company backed by Siemens Healthineers and several healthcare VC firms announces successful collaboration with its partner, Siemens Healthineers. The companies collaborated on the lung nodule detection part of the AI-Rad Companion Chest CT software which assists radiologists in detection of pulmonary nodules. This software was awarded the CE mark in Europe as well as 510k FDA approval and can start to be marketed as a medical product effective immediately.
Dr. Matthias Mitschke coordinated the project and worked together with Dr. Anna Jerebko and her team who oversaw the development of the lung nodule detection software. Anna and Matthias championed this collaboration which involved multiple steps. According to Siemens specifications, USARAD together with MDW developed specialized AI annotation software and radiology workflow for the project. USARAD then leveraged MDW technology to assist in obtaining over 10,000 deidentified computed tomography cases, radiology reports and annotations from its data partners. The image data as well as the clinical reports were carefully anonymized in compliance with HIPAA. Unique heterogeneous acquired CT data was obtained from over 100 imaging centers in multiple states throughout USA. Siemens Healthineers, USARAD and MDW then collaborated to oversee numerous radiologists who provided annotations in multiple phases which included providing input for an FDA submission.
Anna stated, "We are pleased to work together with USARAD on data collection for AI development. It is very important to 'stress test' medical imaging and computer aided detection algorithms on as much data as possible with a large variety of different acquisition protocols, scanner models and patient populations to ensure stable performance thus assessing its generalizability and for use in medical practice. The massive amount of chest CT scans were read and annotated by multiple experienced radiologists. In this manner, the performance of the AI algorithm could be benchmarked against the opinions of experts. It is a very common pitfall in AI development when an algorithm is trained and tested on a limited amount of data acquired from a single site possibly with scanner(s) from a single vendor and interpreted by only a few radiologists. This type of evaluation can lead to conclusions on algorithm results that are too optimistic and can manifest in a suboptimal performance in clinical practice. Stress-testing the AI algorithms for chest CT on a massive amount on clinical data has shown a significant improvement compared to existing standard CAD algorithms."
Previously USARAD and Siemens Healthineers have announced a world-wide distribution partnership with USARAD services to be offered via a global cloud-based sharing and collaboration platform as part of the teamplay digital health platform. This partnership is currently being expanded to include AI collaboration and the integration with teamplay environment is ongoing. USARAD is excited to deploy this solution and other algorithms in its daily practice upon FDA approval. At the same time, USARAD plans to deploy multiple additional algorithms which are not yet FDA approved in research setting using the teamplay environment.
It is anticipated that algorithm's outputs will become part of reporting to facilitate radiologists' workflow with an overall anticipated significant improvement of speed and efficiency. USARAD is also excited to continue collaborating with Siemens Healthineers on multiple new AI projects.
"It is clear that AI is transforming our industry in a radical way. By leveraging deep learning and artificial intelligence technologies USARAD is well positioned to continue its global leading position and will accelerate its quest to affordably bring high quality healthcare around the world," explained Michael Yuz, M.D. MBA, CEO and founder of USARAD Holdings Inc." Dr Yuz is considered to be both the leading expert physician and serial entrepreneur in the field of Artificial Intelligence. He has helped develop algorithms as well as has invested into health-related AI platforms since 2005. He added, "Our involvement with AI is quiet strategic, because not only are we able to provide assistance in developing these AI algorithms, but, when they gain required approval status, we are also able to immediately integrate these algorithms into the production, workflows and day-to-day practice."
About USARAD Holding Inc. including SecondOpinions.com®
USARAD Holdings, Inc. is a global telemedicine company consisting of two operating divisions - USARAD and SecondOpinions.com®. USARAD is a radiology services provider offering interpretations of a full range of sub-specialty and general radiology patient studies. With an existing advanced technology network connecting medical facilities to board-certified radiologists and support staff, USARAD offers radiology services in all 50 states as well as more than 15 countries. Via its unique Radiology-On-Demand® platform, USARAD is committed to providing timely and responsive image review 24/7. SecondOpinions.com® is a medical consultation and second opinions leader providing health care customers, physicians and consumers with expert opinions from all medical specialties and sub-specialties. For more information please visit http://usarad.com or email Elli Yuz at email@example.com
Medical diagnostic web (MDW.io) is first of its kind decentralized autonomous organization operating Blockchain based platform which utilizes smart-contract to facilitate complex B2B and consumer transactions between various stakeholders in radiology field. MDW.io designed to facilitate remote medical diagnostic consultations including radiologic and multispecialty second opinions, peer reviews, primary and preliminary teleradiology interpretations as well as synchronous and asynchronous telemedicine visits. MDW puts patients in forefront of their diagnostic health records by allowing patients and their diagnostic providers to upload findings in a secure, anonymized manner to be further evaluated by the community. For more information please visit http://mdw.io or email Michael Averbach at firstname.lastname@example.org.
Did you like this article?