We’re at a pivotal moment to unlock a new, AI-accelerated era of discovery and medicine, says Kimberly Powell, NVIDIA’s vice president of healthcare.
Speaking today at the J.P. Morgan Healthcare conference, held virtually, Powell outlined how AI and accelerated computing are enabling scientists to take advantage of the boom in biomedical data to power faster research breakthroughs and better patient care.
Understanding disease and discovering therapies is our greatest human endeavor, she said — and the trillion-dollar drug discovery industry illustrates just how complex a challenge it is.
How AI Can Drive Down Drug Discovery Costs
The typical drug discovery process takes about a decade, costs $2 billion and suffers a 90 percent failure rate during clinical development. But the rise of digital data in healthcare in recent years presents an opportunity to improve those statistics with AI.
“We can produce today more biomedical data in about three months than the entire 300-year history of healthcare,” she said. “And so this is now becoming a problem that no human really can synthesize that level of data, and we need to call upon artificial intelligence.”
Powell called AI “the most powerful technology force of our time. It’s software that writes software that no humans can.”
But AI works best when it’s domain specific, combining data and algorithms tailored to a specific field like radiology, pathology or patient monitoring. The NVIDIA Clara application framework bridges this gap by providing researchers and clinicians the tools for GPU-accelerated AI in medical imaging, genomics, drug discovery and smart hospitals.
Downloads of NVIDIA Clara grew 5x last year, Powell shared, with developers taking up our new platforms for conversational AI and federated learning.
Healthcare Ecosystem Rallies Around AI
She noted that amid the COVID-19 pandemic, momentum around AI for healthcare has accelerated, with startups estimated to have raised well over $5 billion in 2020. More than 1,000 healthcare startups are in the NVIDIA Inception accelerator program, up 4x since 2017. And over 20,000 AI healthcare papers were submitted last year to PubMed, showing exponential growth over the past decade.
Leading research institutions like the University of California, San Francisco, are using NVIDIA GPUs to power their work in cryo-electron microscopy, a technique used to study the structure of molecules — such as the spike proteins on the COVID-19 virus — and accelerate drug and vaccine discovery.
And pharmaceutical companies, including GlaxoSmithKline, and major healthcare systems, like the U.K.’s National Health Service, will harness the Cambridge-1 supercomputer — an NVIDIA DGX SuperPOD system and the U.K.’s fastest AI supercomputer — to solve large-scale problems and improve patient care, diagnosis and delivery of critical medicines and vaccines.
Software-Defined Instruments Link AI Innovation and Medical Practice
Powell sees software-defined instruments — devices that can be regularly updated to reflect the latest scientific understanding and AI algorithms — as key to connecting the latest research breakthroughs with the practice of medicine.
“Artificial intelligence, like the practice of medicine, is constantly learning. We want to learn from the data, we want to learn from the changing environment,” Powell said.
By making medical instruments software-defined, tools like smart cameras for patient monitoring or AI-guided ultrasound systems can not only be developed in the first place, she said, but also retain their value and improve over time.
U.K.-based sequencing company Oxford Nanopore Technologies is a leader in software-defined instruments, deploying a new generation of DNA sequencing technology across an electronics-based platform. Its nanopore sequencing devices have been used in more than 50 countries to sequence and track new variants of the virus that causes COVID-19, as well as for large-scale genomic analyses to study the biology of cancer.
The company uses NVIDIA GPUs to power several of its instruments, from the handheld MinION Mk1C device to its ultra-high throughput PromethION, which can produce more than three human genomes’ worth of sequence data in a single run. To power the next generation of PromethION, Oxford Nanopore is adopting NVIDIA DGX Station, enabling its real-time sequencing technology to pair with rapid and highly accurate genomic analyses.
For years, the company has been using AI to improve the accuracy of basecalling, the process of determining the order of a molecule’s DNA bases from tiny electrical signals that pass through a nanoscale hole, or nanopore.
This technology “truly touches on the entire practice of medicine,” Powell said, whether COVID epidemiology or in human genetics and long read sequencing. “Through deep learning, their base calling model is able to reach an overall accuracy of 98.3 percent, and AI-driven single nucleotide variant calling gets them to 99.9 percent accuracy.”
Path Forward for AI-Powered Healthcare
AI-powered breakthroughs like these have grown in significance amid the pandemic, said Powell.
“The tremendous focus of AI on a single problem in 2020, like COVID-19, really showed us that with that tremendous focus, we can see every piece and part that can benefit from artificial intelligence,” she said. “What we’ve discovered over the last 12 months is only going to propel us further in the future. Everything we’ve learned is applicable for every future drug discovery program there is.”
Across fields as diverse as genome analysis, computational drug discovery and clinical diagnostics, healthcare heavyweights are making strides with GPU-accelerated AI. Hear more about it on Jan. 13 at 11 a.m. Pacific, when Powell joins a Washington Post Live conversation on AI in healthcare.
Subscribe to NVIDIA healthcare news here.