Google Cloud strikes imaging partnerships with Change Healthcare, Dicom
At the RSNA annual meeting in Chicago this week, several IT companies struck or extended deals with Google Cloud aimed at helping hospitals tackle the storage challenges posed by precision medicine, improve radiology workflows, bring machine learning to imaging analytics and more.
Change Healthcare, for one, has extended a strategic partnership with Google Cloud to develop new tools for radiologists and other imaging professionals, exploring how artificial intelligence can better be brought to bear on clinical workflows and analytics projects.
By working more closely with the Google Cloud platform, Change aims to create new technologies to connect care teams across location, building a scalable data infrastructure to enable more effective insights via machine-learning technology.
Officials say a cloud-based infrastructure gives health systems a more cost-effective and reliable way to manage and interpret their troves of imaging data.
"Change Healthcare is positioned to transform the value that imaging brings to healthcare providers," said Erkan Akyuz, Change's executive vice president and president for imaging, workflow and care solutions, in a statement. "By working with Google Cloud in this strategic collaboration, we are poised to accelerate that transformation."
In another collaboration, announced Monday, Dicom Systems said it is working as a technology partner with Google Cloud to launch a hybrid cloud VNA, de-identification and imaging data supply chain platform.
The Universal Cloud Archive Adaptor, which is available to test now on Google Cloud Launcher, can integrate with PACS, RIS and EMR systems and, with no migration costs or termination fees, is meant to make cloud-based imaging easier and less expensive for healthcare providers, officials said.
Customers pay only for what they use, according to DICOM Systems, which noted that Google Cloud Platform could offer substantial savings as health systems earn rebates on their operating costs with participation in a de-identified imaging data lake.
"The Universal Cloud Archive Adaptor is the point of origin in a data supply chain that will serve the medical imaging research and development market," said Dmitriy Tochilnik, president and CTO of Dicom Systems, in a statement. "We believe this is how the imaging community should crowdsource for health data: by bringing customers to the frontline where they can innovate alongside us."
Florent Saint-Clair, executive vice President of Dicom Systems, added that the tool is focused on a specific segment of the imaging data supply chain: enabling machine learning and building neural networks to help support the work of physicians.
"As an industry, we're just beginning to understand the building blocks necessary to design effective and reliable AI in imaging diagnostics," he said. "We could not have found a better, no-nonsense cloud provider than Google Cloud to deliver our vision."
In a blog post about RSNA, Gregory J. Moore, MD, vice president of healthcare at Google Cloud, said the company sees imaging as a key area for innovation, and pointed to an array of other technology partnerships aimed at addressing storage, workflow, information exchange and analytics challenges.
"Next to genomics, medical images are one of the fastest growing data sources in the healthcare space," said Moore. "At Google Cloud, we’re working with the research community, clinical community and the diagnostic imaging industry to help care providers be more accurate and effective in order to improve patient outcomes."
Beyond the Change Healthcare and DICOM Systems deals, he highlighted initiatives such as the one Kanteron Systems is pursuing with Google Cloud Platform. Kanteron, which specializes in clinical genomics, is leveraging Google's cloud-based AI to make tools for radiologists, pathologists, oncologists and surgeons to bring precision medicine data to the point of care.
Other Google partners – Ambra Health, lifeIMAGE, Nautilus Medical – are using cloud technology to expand and improve their medical image sharing networks, said Moore. And other companies, such as Zebra Medical Vision are using Google's TensorFlow open source library and Cloud Machine Learning Engine as they train neural networks on existing radiology scans and develop new models to help clinicians detect specific conditions.
And other customers, such as Montreal-based clinical AI developer Imagia are availing themselves of Google's machine learning APIs to create technologies to enable multi-institutional research in precision medicine and create radiomics biomarkers for personalized care. Imaging IT vendors Arterys and Client Outlook, meanwhile, are developing cloud-enabled tools to help automate some of radiologists' most repetitive tasks and enable easy image visualization, respectively, said Moore.