How GPU Accelerated Computing Opens Possibilities For Healthcare Industry

NVIDIA’s pioneering technology helps researchers and doctors find solution for many health challenges plaguing us now

Across the globe doctors, researchers and medical scientists are looking for ways and tools that will help them in better diagnosis and treatment of diseases – not just the old ones, but also the newer ones that we are constantly being plagued with.

Today the researchers and the doctors have a chance to make major progress in finding solution to population’s biggest health challenges – neurological disorders, cancer, blood disorders, and even diagnosing new diseases. The doctors and the scientists are already sitting on a large pile of data, which becomes more comprehensible when they use the right tools. High performance computing brings to them that ability.

NVIDIA today is working extensively in HPC to create technology tools that will aid the doctors in not just working on finding answers to new health issues, but also aid them in treating patients suffering from serious ailments. One of the key technologies is GPU accelerated computing.

GPU in Healthcare

The technology, which NVIDIA pioneered in 2007, uses a graphics processing unit (GPU) along with a CPU to accelerate scientific, analytics, engineering, consumer, and enterprise applications. In this the GPU runs compute-intensive parts of an application while CPU runs the regular codes of the application. The result is that the overall application runs faster and you get your results faster. Today GPU accelerators are being used to power datacenters that run from mission critical organizations to universities, enterprises and even small and medium businesses.

But for NVIDIA, healthcare industry is one where GP accelerated computing can offer an amazing number of capabilities that help the healthcare researchers and doctors find solutions to many present day health challenges. And the company is evolving its GPU accelerated computing on a regular basis with its NVIDIA Qaudro and NVIDIA Tesla GPUs. These can not only tackle the most intensive visual computing task, but also power sophisticated medical imaging devices and visualization on applications. Today, doctors can use the capabilities offered by the GPUs to deliver lower patient dose. Medical researchers can use the powerful features offered by NVIDIA’s GPUs to develop simulations that help in finding solutions to major ailments, 3D visualization systems, mobile medicine, and accelerated research results.

The uses

There are three major areas in healthcare industry where NVIDIA’s GPU accelerated computing is being used today and quite successfully. Here they are:

The robotic surgeon

Surgery is a very human intensive medical procedure where machines are still unable to perform at their optimal without a surgeon’s intervention. But now with GPU accelerated computing surgery can become more safe and accurate procedure – more accessible and less expensive.

One such example is what the doctors at Children’s National Medical Centre in Washington DC created. Dr Peter Kim associate surgeon-in-chief at the CNMC and lead researcher developed STAR (Smart Tissue Autonomous Robot). Kim and his team used different technologies including NVIDIA’s GeForce GTX TITAN GPU to develop STAR. It is an arm-like device that uses a 3D plenoptic camera and near-infrared vision and marking that locates a tissue that needs to be manipulated with sub-millimeter accuracy. The researchers also incorporated laparoscopic suturing tools with added sensing capabilities so STAR can “feel” tension and pressure when in contact with tissue. Additionally, they programmed the robot with a consensus of how surgeons would best perform a complex surgical task.

To test STAR’s capability, the team made it perform anastomosis (the suturing together of two tubular structures) on bowel segments in a pig. The surgery was not only successful, but was also the first completely autonomous robotic anastomosis. Kim was proud of the fact that the movements of STAR were so consistent and precise that the outcome was better than that done by experienced surgeons.

How GPUs helped? GPUs play a critical role in STAR. They’re used to speed up the calculations of data coming from the plenoptic camera, which captures information about light emanating from the scene. This provides STAR with positional awareness and the ability to track target tissues in real time.

The virtual therapist

Researchers at Oxford University showed that virtual reality could serve as a safe space for patients who suffer from various phobias that make their day-to-day life difficult. By being in a simulated environment they can face their fears and learn to fight them.

In a first ever experiment using VR to treat psychotic experience of phobia patients, the researchers created two VR simulations powered by NVIDIA GPUs. The first was a crowded elevator and a train car on the London Underground.

The team’s study put 30 patients with persecutory delusions into a virtual experience in the London Underground or an elevator. Each subject entered the virtual world seven times for up to five minutes. With each trial, more virtual people were added to the virtual space, gradually exposing the patients to more intimidating situations with up to 28 avatars. With just half an hour to explore the virtual world, their subjects felt a dramatic reduction in their feelings of paranoia.

In the long term, the team would like to incorporate more realistic facial animations and up to 50 avatars in a virtual scene. They also hope to build more personal avatars that look and dress like the subjects. Riding virtual train cars filled with people can help patients with paranoia reduce feeling under threat. With time, the researchers see VR therapy becoming mainstream in mental health clinics, hospitals and the home.

How GPUs helped? VR is computationally intensive, even more so with many avatars. According to the computer scientist with the project, without a GPU, the simulations wouldn’t have run fast enough. The study was conducted with a GeForce GTX GPUs and an NVIS nVisor SX111 head-mounted display. Currently, the team is upgrading to a TITAN X GPU and Oculus headset so they can include more avatars in their virtual world and achieve higher frame rates.

Deep learning and a cure

To find the best drug treatments for the paediatric ICU at the Children’s Hospital in Los Angeles, the hospital’s data scientist David Ledbetter has turned to GPU-powered deep learning to digest big data — a decade’s worth of health records. The aim is to optimize patient outcomes. The team wanted to focus on the patients and get them the best care based on the reports that their health data reveal.

Increasingly powerful GPUs allow scientists to apply deep learning, a fast-developing branch of artificial intelligence, to teach computers to learn to see patterns in giant datasets. To train its deep learning models, Ledbetter’s team created nearly 13,000 so-called “patient snapshots” from the records of patients who passed through the paediatric ICU at Children’s Hospital Los Angeles.

The snapshots show interactions between a patient’s vital state, heart rate, blood pressure and the treatments they were given. Ledbetter and his team fed these into two separate neural network models, using TITAN X GPUs to complete the training in a matter of hours. This aided their understanding of the key relationships between the patient’s vitals and the interventions performed in the unit. With a convolutional neural network, they were able to predict the probability of survival, and with a recurrent neural network, they could predict physiology through time.

Input source: NVIDIA Blog on GPU Accelerated Computing.

Click to comment

Leave a Reply

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

To Top