What would you do if you
could know about a health care concern way before
real problems occurred? See professional sooner? Save up for a potential
surgery in advance? Well, Intel alongside Philips,
a health care company, are paving the way to just that. I’m Stephanie Essin. Find out how we’re getting
closer in this IDZ Weekly. [MUSIC PLAYING] Using Intel Xeon Scalable
processors and the OpenVINO toolkit, Intel and Philips
tested two health care use cases for deep learning
inference models– one on X-rays of bones for
bone-age-prediction modeling, and the other on CT scans of
lungs for lung segmentation. In these tests,
Intel and Philips achieved a speed
improvement of 188 times for the bone-age-prediction
model and a 38 time speed improvement for the
lung-segmentation model over the baseline measurements. AI techniques like object
detection and segmentation can help radiologists identify
issues faster and more accurately, which can translate
to better prioritization of cases, better outcomes
for more patients, and reduce costs for hospitals. So how are AI techniques
like this possible? It all comes down to using
CPUs instead of GPUs. Until recently, GPUs were really
the only prominent hardware solution for accelerating
deep learning. But they have inherent
memory constraints that data scientists have
had to work around when building models in the past. CPUs– in this case Intel
Xeon Scalable processors– don’t have those same
memory constraints and can accelerate
complex hybrid workloads. And for medical
imaging, this is vital, as models are typically
memory-intensive. For a large subset
of AI workloads, Intel Xeon Scalable
processors can better meet data scientists’ needs
than GPU-based systems. As Philips found in
the two recent tests, this enables the company to
offer AI solutions at a much lower cost to its customers. Curious to learn how the
OpenVINO toolkit and Intel Xeon Scalable processors
could enhance deep learning in your projects? Check out the links
to discover how. And thanks for
watching IDZ Weekly. See you next Monday.

Accelerating Deep Learning for Medical Imaging | IDZ Weekly | Intel Software
Tagged on:                                                                                                                                                         

Leave a Reply

Your email address will not be published.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