Advanced Computing Holds Promise for Health Care, Ethical Hurdles Remain

Advanced computing technologies like artificial intelligence and quantum computing are poised to revolutionize the healthcare industry, offering unprecedented capabilities. Federal IT leaders at the GovCIO Media & Research Health IT Summit discussed the full transformative potential of these tools, and how agencies can navigate the ethical, security and practical challenges.

Killer Application

Agencies are migrating to quantum-resistant systems based on the latest post-quantum cryptography standards set by the National Institute of Standards and Technology. The National Science Foundation’s Deputy Division Director for the Division of Information and Intelligent Systems Wendy Nilsen said the quantum world is exciting but theoretical. A “killer application” that will revolutionize healthcare through quantum computing is still needed.

“We need that area where we can figure out how to get [quantum] applied and then look at NIST’s risk management framework,” said Nilsen. “We have to think about balancing all of these things: privacy, security and biases in the data.”

Anuj Kapadia, section head for Advanced Computing in Health Sciences at Oak Ridge National Lab, added that data security is essential when moving data from one point to another. He also highlighted the integrity of data sets that also need to be protected, ensuring AI models are trained on the best data possible.

“If you have an AI model that’s been trained on certain data, you do not want someone to be able to reverse engineer that AI model and learn about the underlying data set,” said Kapadia.

Securing Patient Data

The security of patient data remains a challenge, especially as commercial data from wearable technologies is used and collected. Technologies like smart watches keep patients updated about their health data and easily share it with healthcare professionals. These technologies also use cryptography to ensure data is secure from user to professional.

Richard Persinger, general manager of Federal Civilian & APG at HP, said the future of healthcare will allow patients to take control of their health in a new way. Persinger described a future where patients could draw blood samples and analyze them in their homes with the right technology. The question then becomes how the data is protected and who oversees it.

“We may come to that point where that happens, and then do you have to save that data the same way you would with a copy of your will,” said Persigner. “If you’re retaining that data, how is it leveraged from a clinician? [We have to ask] how is your health looked at holistically?”

Improving Customer Experience

Researchers and government officials are creating policies to improve customer experience nearly a year after President Biden’s executive order on digital experience. Nilsen said user-facing applications need the same amount of attention as the technology itself.

Creating a user interface that people can easily understand is crucial to data and device security. Medical devices that needed an update in the past were shipped in for repair or were replaced. Today’s technology allows users to download software onto their medical devices easily and from home.

“I have a system where it still says: ‘Go find a SIM card and download this to your SIM card externally, put it back in the system and upload. This is not the right century for this, but we used to do that,” said Nilsen. “Keeping the user as a central factor, whether the user is a clinician, an administrator or a patient, is important.”

Nilsen said the user experience needs to extend to the caregivers, like the administrators and clinicians. She added that without user-friendly systems for workers, it becomes easy to keep a password sticky note rather than following security procedures. Nilsen advocated for working with the flaws of human memory rather than against it.

“There’s some research that looked at could you have people memorize a little song and put those notes back in and use that as a security code,” said Nilsen. “We don’t want to make systems so complicated that you do something [tedious] to make it work.”