Google Cloud expands role in supporting radiologists

The Google Cloud is grabbing more of a footprint in radiology, with two vendors relying on it to support new applications for radiologists.

In a new partnership, Minneapolis-based Flywheel is working on a new technology initiative that will use Google Cloud’s Application Programming Interface (API) with Flywheel’s platform to provide clinical researchers with advanced cloud technology for medical imaging research.

Another vendor, International Medical Solutions, is announcing a solution at the Healthcare Information and Management Systems Society conference in Orlando that will enable radiologists and other clinicians to use machine learning modeling to triage studies focusing on medical images that need immediate attention.

Both applications draw on the accessibility and coordination capabilities of the cloud to help integrate the work of radiologists within the clinical process.

Radiology-CROP.jpg
A technician looks at scanned imagery in the control room of the diagnostic imaging area at the Hong Kong Integrated Oncology Centre in Hong Kong, China, on Tuesday, Nov. 3, 2015. Equipped with biopsy facilities, body scanners, and quiet 'VIP' chemotherapy rooms, the Hong Kong Integrated Oncology Centre is the first of a string of such facilities that TE Asia Healthcare Partners, a portfolio company funded by TPG Capital, is planning in Asia. Photographer: Xaume Olleros/Bloomberg

Flywheel executives say the use of the Google Cloud API will enable clinical researchers using its biomedical research platform to efficiently capture multi-modality images and data. The approach will boost productivity of data classification and enable collaboration with peers to manage metadata.

The partnership with Google will enable Flywheel to incorporate Google Cloud’s BigQuery, a fully managed enterprise data warehouse for large-scale data analytics, and to enable scalable analysis of medical imaging metadata, biomarker data and tabular data, including genomics. Also, Google Cloud’s AutoML Vision API, a flexible machine learning service, is used to create a comprehensive learning workflow for artificial intelligence models using medical imaging data.

The platform offers a true multi-modality solution, including support for DICOM, HL7 and FHIR standards, and addresses the limitations of traditional PACS and VNAs to meet the needs of modern clinical research.

“With our partners at Google Cloud, we are delivering an innovative research platform to meet the needs of imaging center directors, principal investigators and clinical research institutions seeking to establish a scalable infrastructure for machine learning, advanced imaging research and secure collaboration to advance science and, ultimately, precision medicine,” says Flywheel CEO Travis Richardson.

With its announcement, International Medical Solutions is unveiling a prototype at HIMSS that focuses on two common use cases. One uses the Google Cloud ML Engine to provide an indicator for prioritization. That could help radiologists who might have several hundred chest cases to review, by using machine learning to indicate the cases that are most likely to have potential pathologies.

In the second use case, the Google Cloud ML Engine will be used to help provide better decision support, particularly helpful in situations in which there is no urgency involved with a diagnosis.

"In our experience, healthcare professionals are hyper-focused on using flexible solutions that give them the ability to improve their workflow, reduce costs and provide an accurate diagnosis for their patients,” says Vittorio Accomazzi, CTO of International Medical Solutions. “We believe our new platform using Google Cloud ML Engine will transform the way medical images are read and interpreted for many years to come."

For reprint and licensing requests for this article, click here.