Google’s recent claim that it has achieved quantum computing supremacy – apart from being contested by rival tech giants such as IBM – is still quite a ways off in terms of real-world applications.
The healthcare industry, for example, will still need to build a new set of applications to take advantage of quantum, and it still won’t account for the cost of the hardware and the operating costs to cool the systems and keep them operational.
If healthcare costs were not already high enough, these capabilities don’t help in keeping cost burdens low, and because quantum introduces all kinds of potential security risks, data privacy for healthcare patients could be compromised even further.
However, possible applications for artificial intelligence and machine learning to help with data analysis could prove critical further down the road.
Quantum computing could provide unprecedented power and speed of processing as well as novel and fundamentally different algorithmic search and data homogenization strategies.
“The exponential computing speedup offered by quantum computers will enable machine learning algorithms to rapidly identify patterns in healthcare data collected from millions of participating patients,” Mario Milicevic, an IEEE member and staff communication systems engineer at MaxLinear, told Healthcare IT News.
He explained medical imaging and pathology would likely be the first to benefit, as quantum computers could be used to train machine learning algorithms with more classifiers to identify diseases in a fraction of the time that it takes today.
Milicevic noted quantum computers could also accelerate DNA sequencing, which would enable the more effective cancer treatment through personalized medicine.
A central challenge that remains is collecting and curating healthcare data uniformly across a multitude of sources in such a way that it can be processed by quantum algorithms.
Nick Hatt, senior developer at digital health company Redox, cautioned that it’s going to be important to not buy into the hype too much at such an early stage.
“No one should be putting a down payment on a quantum computer today,” he said. “The methods used today in AI/ML are well understood and run reasonably fast on conventional computers.”
Hatt explained that what healthcare CIOs need to worry about is cryptography.
“Essentially all of the ways we secure our health data — from APIs that transmit it, to the actual storage on disk,” he said. “The data is at risk of being completely and utterly broken.”
From a clinical healthcare perspective alone, the quantum computing technology could lead to “dramatic” accelerations in speed and performance.
“MRIs were basically invented because of our acquired understanding of quantum physics, and getting a true quantum computer will allow us to truly understand the nature of all matter, which means everything from better medicine with less side effects to better diagnostics,” Roger Grimes, data-driven defense evangelist at KnowBe4, told HealthcareITNews.
With increased computing available, clinicians could easily review CT scans over time and quickly identify changes and anomalies. Similarly, precision medicine can be accelerated.
Targeted chemotherapy protocols can be identified more quickly, and with more customization, with quantum computing’s enhanced data processing abilities.
“All of the above apply to oncology specifically as well,” noted Dr. Doug Walled, an IEEE member and an attending physician in diagnostic radiology and nuclear medicine.
He explained roving machine learning algorithms crawling across disparate systems could adapt unlike data, and much more rapidly change the treatment landscape for various types of cancer.
The kind of massive processing power and intelligence quantum computing will bring could also change the landscape for AI-based healthcare applications, because clarity will be available much more rapidly.
One tenant of quantum computing is that two “objects” may seem unrelated and, with quantum applied, are realized to be somehow related.
“Extrapolate that idea to healthcare and AI and you can imagine that when AI brings together information and extrapolates parallels in the data that then, science will connect previously unconnectable dots,” Mark LaRow, CEO of patient matching services provider Verato, told Healthcare IT News. “Apply this to clinical trials in fields like oncology and pretty soon we cure undiagnosable cancers.”
LaRow cautioned one challenge to the adoption and full use of the technology’s potential is the limited, incomplete, or inconsistent data sets required to train and be available for mass computational consumption and AI learning.
For example, a chronically ill patient may see more than seven clinicians, and these clinicians document differently, they copy notes from other providers, use shorthand, and think differently.
“This creating inconsistency and incompleteness across the medical record,” he said. “Sophisticated solutions like AI and quantum computing will benefit from complete medical records paired with supplementary non-medical information.”
Although it may be years – even a decade or more – before quantum computing becomes a standard part of the healthcare business, LaRow noted these “wickedly futuristic technologies” have, in the last five years, evolved to seem tangible.
“Ultimately, I believe, that these technologies will become so reliable that it will be deemed unethical for a clinician NOT to consult with a powerful AI informed computing system to double check a diagnosis and to recommend a treatment regimen,” he said.
Nathan Eddy is a healthcare and technology freelancer based in Berlin.
Email the writer: [email protected]
Twitter: @dropdeaded209
Source: Read Full Article