Personal Profile

Medical Sciences

JetPack Latest NVIDIA JetPack Developer Tools Will Double Your Deep Learning Performance

NVIDIA Helping Developers Get Started with Deep Learning

NVIDIA Helping Developers Get Started with Deep Learning

From self-driving cars to medical diagnostics, deep learning powered artificial intelligence is impacting nearly every industry.

In 2015, NVIDIA’s Deep Learning Institute delivered more than 16,000 hours of training to help data scientists and developers master this burgeoning field of AI – and the need for deep learning training is rapidly growing.

In the next four months developers can take more than 80 instructor-led workshops and hands-on labs at one of the eight GPU Technology Conferences around the world – starting this week at GTC China.

“We want to share all our knowledge about deep learning with the world so others can create amazing things with it,” said Mark Ebersole, director of the institute.

Julie Bernauer, an NVIDIA Deep Learning Institute instructor, teaches a class on deep learning on GPUs.

Julie Bernauer, an NVIDIA Deep Learning Institute instructor, teaches a class on deep learning on GPUs.

The Deep Learning Institute has joined forces with three industry-leading organizations to train data scientists and developers interested in deep learning:

  • Teaming up with Coursera to create a series of courses on how deep learning is poised to transform healthcare
  • Collaborating with Microsoft on a hands-on workshop about how to use deep learning to create smarter robots
  • Partnering with Udacity to help developers learn how to build a self-driving car

Deep Learning to Unlock Mysteries of Parkinson’s Disease

Deep Learning to Unlock Mysteries of Parkinson’s Disease

Deep Learning to Unlock Mysteries of Parkinson’s Disease

Researchers at The Australian National University are using deep learning and NVIDIA technologies to better understand the progression of Parkinson’s disease.

Currently it is difficult to determine what type of Parkinson’s someone has or how quickly the condition will progress.
The study will be conducted over the next five years at the Canberra Hospital in Australia and will involve 120 people suffering from the disease and an equal number of non-sufferers as a controlled group.

“There are different types of Parkinson’s that can look similar at the point of onset, but they progress very differently,” says Dr Deborah Apthorp of the ANU Research School of Psychology. “We are hoping the information we collect will differentiate between these different conditions.”

Researchers Alex Smith (L) and Dr Deborah Anthrop (R) work with Parkinson’s disease sufferer Ken Hood (middle).

Researchers Alex Smith (L) and Dr Deborah Anthrop (R) work with Parkinson’s disease sufferer Ken Hood (middle).

Dr Apthorp said the research will measure brain imaging, eye tracking, visual perception and postural sway.

From the data collected during the study, the researchers will be using a GeForce GTX 1070 GPU and cuDNN to train their deep learning models to help find patterns that indicate degradation of motor function correlating with Parkinson’s.

The researchers plan to incorporate virtual reality into their work by having the sufferers’ wear head-mounted displays (HMDs), which will help them better understand how self-motion perception is altered in Parkinson’s disease, and use stimuli that mimics the visual scene during self-motion.

“Additionally, we would like to explore the use of eye tracking built into HMDs, which is a much lower cost alternative to a full research eye tracking system and reduces the amount of equipment into a highly portable and versatile single piece of equipment,” says researcher Alex Smith.

Advanced Real-Time Visualization for Robotic Heart Surgery

Advanced Real-Time Visualization for Robotic Heart Surgery

Advanced Real-Time Visualization for Robotic Heart Surgery

Researchers at the Harvard Biorobotics Laboratory are harnessing the power of GPUs to generate real-time volumetric renderings of patients’ hearts. The team has built a robotic system to autonomously steer commercially available cardiac catheters that can acquire ultrasound images from within the heart. They tested their system in the clinic and reported their results at the 2016 IEEE International Conference on Robotics and Automation (ICRA) in Stockholm, Sweden.

The team used an Intracardiac Echocardiography (ICE) catheter, which is equipped with an ultrasound transducer at the tip, to acquire 2D images from within a beating heart. Using NVIDIA GPUs, the team was able to reconstruct a 4D (3D + time) model of the heart from these ultrasound images.

Generating a 4D volume begins with co-registering ultrasound images that are acquired at different imaging angles but at the same phase of the cardiac cycle. The position and rotation of each image with respect to the world coordinate frame is measured using electromagnetic (EM) trackers that are attached to the catheter body. This point cloud is then discretized to lie on a 3D grid. Next, infilling is performed to fill the gaps between the slices, generating a dense volumetric representation of the heart. Finally, the volumes are displayed to the surgeon using volume rendering via raycasting, leveraging the CUDA – OpenGL interoperability. The team accelerated the volume reconstruction and rendering algorithms using two NVIDIA TITAN GPUs.

“ICE catheters are currently seldom used due to the difficulty in manual steering,” said principal investigator Prof. Robert D. Howe, Abbott and James Lawrence Professor of Engineering at Harvard University. “Our robotic system frees the clinicians of this burden, and presents them with a new method of real-time visualization that is safer and higher quality than the X-ray imaging that is used in the clinic. This is an enabling technology that can lead to new procedures that were not possible before, as well as improving the efficacy of the current ones.”

Providing real-time procedure guidance requires the use of efficient algorithms combined with a high-performance computing platform. Images are acquired at up to 60 frames per second from the ultrasound machine. Generating volumetric renderings from these images in real-time is only possible using GPUs.

Scientists Gather at University of Delaware for OpenACC Hackathon

Scientists Gather at University of Delaware for OpenACC Hackathon

Scientists Gather at University of Delaware for OpenACC Hackathon

Oak Ridge National Lab, NVIDIA and PGI launched the OpenACC Hackathon initiative last year to help scientists accelerate applications on GPUs. OpenACC was selected as a primary tool since it offers acceleration without significant programming effort and works great with existing application codes.

University of Delaware (UDEL) hosted a five-day Hackathon last week. Selected teams of scientific application developers hacked away under close mentorship of GPU experts and others with extensive experience in programming with OpenACC.

Watch the recap video by Sunita Chandrasekaran, professor at UDEL and organizer of the Hackathon.

The Hackathon at UDEL brought together teams from NASA Langley Research Center, Oak Ridge National Laboratory, National Cancer Institute, National Institutes of Health (NIH), Brookhaven National Laboratory and the UDEL College of Engineering to accelerate their codes across medical imaging, chemical and bio-molecular engineering, computational fluid dynamics, quantum physics and computer science.

Scientists Gather at University of Delaware for OpenACC Hackathon

Scientists Gather at University of Delaware for OpenACC Hackathon

Codes represented at the Hackathon:

  • FUN3D: Large-scale computational fluid dynamics solver for complex aerodynamic flows seen in a broad range of aerospace applications.
  • Lattice QCD: Numerical simulation to solve the high-dimensional non-linear problems in strong interactions for nuclear and particle physics.
  • Application that determines RNA structure using data from small angle x-ray scattering experiments.
  • Application that identifies malicious software based on graph analysis
  • Kinetic Model Builder application that builds models based on chemical engineering principles.
  • DARTS: Open-source dataflow based runtime system to support heterogeneous architectures.

The highlight from the Hackathon was by the team from the National Cancer Institute who achieved a 13x performance increase during the UDEL Hackathon.

Check out the 2016 Hackathons page if you want to be a mentor or to register your team.

GPU-Accelerated Model Reveals Details of Nuclear Fission

GPU-Accelerated Model Reveals Details of Nuclear Fission

GPU-Accelerated Model Reveals Details of Nuclear Fission

Scientists from University of Washington, Warsaw University of Technology in Poland, Pacific Northwest National Laboratory, and Los Alamos National Laboratory, have developed a model that provides a detailed look at what happens during the last stages of the fission process.

According to their research paper, nuclear fission has almost reached the venerable age of 80 years and yet we still lack an understanding in terms of a fully quantum microscopic approach.

Using the new model, the scientists determined that fission fragments remain connected far longer than expected before the daughter nuclei split apart and the predicted kinetic energy agrees with results from experimental observations. This discovery indicates that complex calculations of real-time fission dynamics without physical restrictions are feasible and opens a pathway to a theoretical microscopic framework with abundant predictive power.

Snapshots of the total density profile of the 240Pu fission process.

Snapshots of the total density profile of the 240Pu fission process.

Evaluating the theory amounted to solving about 56,000 complex coupled nonlinear, time-dependent, three-dimensional partial differential equations for a 240Pu nucleus using a highly efficient parallelized GPU code. The calculations required nearly 2,000 NVIDIA GPUs on the Titan supercomputer at Oak Ridge National Lab.

By accurately modeling fission dynamics, the work will impact research areas such as future reactor fuel compositions, nuclear forensics, and studies of nuclear reactions.



Popular Pages
Popular posts
Interested
About me

My name is Sayed Ahmadreza Razian and I am a graduate of the master degree in Artificial intelligence .
Click here to CV Resume page

Related topics such as image processing, machine vision, virtual reality, machine learning, data mining, and monitoring systems are my research interests, and I intend to pursue a PhD in one of these fields.

جهت نمایش صفحه معرفی و رزومه کلیک کنید

My Scientific expertise
  • Image processing
  • Machine vision
  • Machine learning
  • Pattern recognition
  • Data mining - Big Data
  • CUDA Programming
  • Game and Virtual reality

Download Nokte as Free


Coming Soon....

Greatest hits

The fear of death is the most unjustified of all fears, for there’s no risk of accident for someone who’s dead.

Albert Einstein

Waiting hurts. Forgetting hurts. But not knowing which decision to take can sometimes be the most painful.

Paulo Coelho

Gravitation is not responsible for people falling in love.

Albert Einstein

One day you will wake up and there won’t be any more time to do the things you’ve always wanted. Do it now.

Paulo Coelho

Anyone who has never made a mistake has never tried anything new.

Albert Einstein

Imagination is more important than knowledge.

Albert Einstein

It’s the possibility of having a dream come true that makes life interesting.

Paulo Coelho

You are what you believe yourself to be.

Paulo Coelho


Site by images
Statistics
  • 6,279
  • 19,112
  • 62,644
  • 18,332
Recent News Posts