The Explosion of AI in the Imaging Space

This post has been republished via RSS; it originally appeared at: New blog articles in Microsoft Tech Community.

RSNA (Radiological Society of North America) took place recently in Chicago with attendance over 50,000 including 700+ exhibitors.  It is the world’s largest medical image event New this year was the AI Showcase” made up of 130 organizations showing various applications of artificial intelligence.  The term “AI” had greater mind space with a presence in nearly every booth in signage or in demo.  What is driving this explosion of focus on AI in the imaging space? 

 

Image quality itself is contributing to improvements in AI accuracy.  Higher resolution images based on improving camera quality and improved capture techniques were highlighted around RSNA – especially in MRI (Magnetic Resonance Imaging) and mammography devices.  Storage and movement of large images is becoming cheaper by the day with economies of cloud storage and options like “edge” computing bringing cloud benefits to on premise devices.  Faster and more reliable 5G cellular networks will have an impact on image exchange and mobility. 

 

Advanced viewing technologies enable high resolution image reads over any speed networks.  This is important in lowering costs while increasing image sharing and collaborative analysis using streaming technology and compute power of the edge device, like a smartphone, to view high resolution images stored in the cloud in near real time.  Another high potential viewing technology is mixed reality.  Using the latest headsets and holographic technology like Microsoft HoloLens 2.0, there were opportunities to view and manipulate 3d images of tumors to gain deeper insights to the image and aid in diagnosis and visualization of the problem  GE replicated a mammogram image by printing acrylic images of a breast highlighting with color ink the exact location of the tumor.  Then they printed the same image in a spongy plastic substance that enabled the patient to gain a sense of the tumor density, size, and location.   

 

It was promising to see DICOM images being integrated into the care collaboration workstreams using modern workplace technologies like Microsoft TEAMS or Dynamics 365.  This enables clinical care teams to review, comment, chat, launch a video conference, annotate etc. in a secure and productive manner.  Numerous vendors were highlighting this important capability. 

 

Lastly, the promising future of quantum computing to solve real world challenges and support the application of both AI techniques and computer performance. To see an example of this, make sure to watch what is progressing at Case Western with quantum computing. 

 

We see the convergence of many factors that are promising to healthcare. Specifically in the area of medical images: better image creation, easier and cheaper image storage management via cloud PACS and VNA’s,  anytime/anywhere viewing on edge devices like smartphones, enhanced 3d viewing with augmented reality and holographics, workstream integration, and application of AI techniques including machine learning to analyze to lead to better, faster, cheaper diagnosis.  

 

- Kevin Dolan, National Director, US Health & Life Sciences 

Follow me on LinkedIn 

Leave a Reply

Your email address will not be published. Required fields are marked *

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.