🔥 Find me at MedicalDevicesGroup.net 🔥
4 min reading time
“If you’re developing a device and it doesn’t integrate some kind of intelligence, you are contributing to the problem,” began Spyro Mousses PhD at his 10x Medical Device Conference talk in May.
His nine-minute talk and slides at https://medgroup.biz/Spyro
Dr. Mousses explained we need to transform your data into something useful, which is less likely if it doesn’t feed into an expert system or cognitive computing / machine learning system. It’s so impractical to consider data from each system independently that we don’t even collect it.
But “the little taste we’ve gotten from using genomic data to position cancer drugs, to using integrated image analysis, to look at cardiology, has taught us there is a massive amount of intelligence to individualize treatments.”
And machine learning algorithms are changing so rapidly that “everything you’re doing today in terms of software analytics is going to be obsolete in just a few years.” By ‘a few years’ Spyro means 2025 or sooner.
Dr. Anthony Chang, a 10x faculty alumnus and Chief Intelligence Office at Children’s Hospital of Orange County, is an early adopter. He wants to process echocardiograms, EKGs, genomic data, device-streaming data, so he can predict hemotherapy-induced cardiomyopathy from data.
The data sources not only stream from devices. Machine learning are beginning to integrate image data and decision support systems.
Spyro’s company, Systems Imagination, is working with Intel, Dell, and others to translate disparate data into unified multi-scalar models: Medical records, radiology images, pathology images, and molecular analyses including protein, DNA, and RNA.
So today’s your wake-up call. Integrate your data into machine learning systems or get left behind. Reaching Dr. Mousses might be a good place to start.
The call for 10x speakers closes at month’s end. Visit https://medgroup.biz/callforspeakers to apply.
Make it a great week.
Professor Dr Hiten R.H. Patel
Naveen Agarwal, Ph.D.
David G. Vita
UGOCHUKWU NWAEME, CSMLS(MLT)
Marked as spam