Proving the Business Case for the Internet of Things

Philips uses machine learning to help radiologists analyse images

Steve Rogerson
December 6, 2016



At last week’s Radiological Society of North America’s annual meeting in Chicago, Philips introduced IntelliSpace Portal 9.0, a visual analysis and quantification platform with machine learning capabilities.
 
Featuring a suite of multi-modality functions and expanded neurological tools, version 9.0 helps radiologists detect, diagnose and follow-up on treatment of diseases, while using machine learning capabilities to support the physician. Tools support the growing group of patients with brain injuries and neurological disorders such as dementia, strokes, amyotrophic lateral sclerosis and multiple sclerosis.
 
In addition to enhancements in areas such as CT brain perfusion and MR T2 perfusion, it provides longitudinal brain imaging, an application for neuro reading to support the evaluation of neurological disorders over time so clinicians can monitor disease progression. Another feature is the inclusion of the NeuroQuant measurement application from CorTech Labs, which lets clinicians quantify brain volume loss.
 
"As the rates of dementia and neurodegeneration rise, neuroradiologists need advanced tools to help referring physicians treat these challenging cases," said Leo Wolansky, acting chief of neuroradiology at University Hospitals Cleveland Medical Center. "IntelliSpace Portal 9.0 offers a comprehensive set of robust tools so neuroradiologists can quantify disease expediently."
 
With more than 70 applications and enhancements of many of the core applications, it is a single platform for visual analysis and quantification that spans clinical domains within radiology, including neurology, oncology and cardiology. The platform offers improved advanced 3D rendering and refined STL export that enable clinicians to print 3D models with high levels of detail and resolution to help understand the anatomy. It also includes enhancements to a host of core applications from MR cardiac analysis to CT Tavi planning, and system enhancements.
 
"Radiology has a unique ability to influence and improve outcomes, and intelligent tools enable us to empower radiologists with the right information," said Yair Briman, senior vice president for Philips. "With advances in machine learning, IntelliSpace Portal 9.0 will now be able to continually learn the usage patterns of users to enhance the important daily functions of a radiologist such as pre-processing of images, encouraging faster and more streamlined diagnosis."
 
The platform includes multimodality clinical applications that can be accessed from any point of the hospital network. It can also integrate with typical PACS and hospital information systems to allow for information to be shared broadly, helping drive collaboration across the network. With enterprise scalability, neurologists can access the power of analysis virtually anywhere across their organisation while maintaining consistent applications and user preferences.
 
MR cardiac enhancements facilitate visual scoring in various examination contexts. The package enables functional volumetric analysis. Enhancements to the package include whole heart STL export for 3D printing, and enhanced workflow and batch tools.
 
The spectral diagnostic suite of clinical applications has been optimised for the viewing and analysis of spectral data sets from the IQon spectral CT scanner. Users can access needed applications when and where they need, virtually anywhere in the enterprise. The tools help give a comprehensive overview of each patient.
 
Philips also announced the commercial availability of its IntelliSpace universal data manager, a scalable, secure, interoperable data management system that supports healthcare enterprises in organising large data sets, including millions of images and other data from multiple sources, and quickly delivering them to virtually any clinician throughout their health network.
 
And the Dutch firm announced Illumeo imaging and informatics technology with adaptive intelligence that redefines and enhances how radiologists work with medical images. The intelligent software combines contextual awareness capabilities with data analytics to augment the work of the radiologist. Its built-in intelligence records the radiologists' preferences and adapts the user interface to assists the clinician by offering tool sets and measurements driven by the understanding of the clinical context. It aims to enable faster diagnoses, to drive well-informed care decisions and improved patient care.
 
Also at the meeting, Boston Children's Hospital and GE Healthcare announced a collaboration to develop and commercialise digital products to advance the diagnosis and treatment of specific childhood diseases, starting with diseases that affect the brain. The first project seeks to improve diagnostic accuracy in paediatric brain scans by providing real-time contextual information at the time and place the radiologist needs it.