This blog is totally independent, unpaid and has only three major objectives.
The first is to inform readers of news and happenings in the e-Health domain, both here in Australia and world-wide.
The second is to provide commentary on e-Health in Australia and to foster improvement where I can.
The third is to encourage discussion of the matters raised in the blog so hopefully readers can get a balanced view of what is really happening and what successes are being achieved.
Quote Of The Year
Quotes Of The Year - Paul Shetler - "Its not Your Health Record it's a Government Record Of Your Health Information"
H. L. Mencken - "For every complex problem there is an answer that is clear, simple, and wrong."
Friday, April 15, 2016
This Is A Nice Example Of Cross Disciplinary Effort To Try And Make A Difference.
A hedge fund trading desk seems an unlikely starting point for a medical technology breakthrough to detect heart disease.
Yet with the help of Australian-founded data scientist start-up Kaggle, American quantitative traders Tencia Lee and Qi Liu have designed a world-first algorithm to analyse heart scans.
The revolutionary machine appears to match the accuracy of highly trained cardiologists.
The algorithm, once commercialised, has the potential to save cardiologists 20 minutes per examination and lower the cost to a patient by hundreds of dollars.
Doctor Andrew Arai, advanced cardiovascular imaging group senior investigator at the US National Institute of Health, says a preliminary assessment showed the computer measurements were "at least as similar" to checks undertaken by a cardiologist.
"This is the first quantitative MRI [magnetic resonance imaging] measurement that may be feasible to do in large numbers automatically," he says.
"I was very surprised because I never trusted automated tools."
Just three months
The medical industry has been trying to achieve this goal for 15 years. It took the American maths geniuses just three months.
Lee and Liu combined to win a $US125,000 ($166,045) prize among 993 data scientist contestants in a problem-solving competition hosted by Kaggle.
Silicon Valley-based Kaggle, a data scientist platform with thousands of community members who compete for prizes in online data challenges sponsored by businesses, was founded by former Reserve Bank of Australia and Treasury economist Anthony Goldbloom in Melbourne six years ago.
Goldbloom says the availability of huge data sets and the recent advent of a chip known as a graphical processing unit (GPU) that can process algorithms quickly has significantly enhanced so-called machine learning and artificial intelligence (AI) in healthcare and other areas.
Lee, a 28-year-old maths graduate, says she drew on quantitative skills learned from her six years working at a Los Angeles hedge fund to solve the medical problem.
Artificial intelligence could be used by doctors to diagnose heart failure following a breakthrough trial hosted by Silicon Valley crowdsourcing company Kaggle, the American National Institute of Health and consulting giant Booz Allen Hamilton.
Kaggle, which was founded in Melbourne by Anthony Goldbloom and relocated to San Francisco five years ago, uses the internet and social media to operate data analysis competitions for tens of thousands of scientists worldwide to solve problems.
It recently hosted a competition to develop algorithms that could diagnose heart failure from MRI scans.
The winning algorithm managed to diagnose the disease in the same way as human cardiologists, for the first time offering the potential to slash the time and cost of diagnosis.
“We now have an algorithm that can diagnose heart failure instantly,’’ Mr Goldbloom told The Australian. “It is cool that Kaggle is a company taking this technology and applying it to real problems to make a real difference. The NIH will take possession of this and eventually it could go into the big MRI machines that will be developed by the likes of GE and Siemens.”
The winning algorithm was built by US hedge fund analysts Qi Liu and Tencia Lee. While they had no medical experience, by using a technique known as deep learning they created an algorithm that had an accuracy rate on par with the human eye.
They were competing in Booz Allen Hamilton and Kaggle’s second annual Data Science Bowl, which was supported by the American National Heart, Lung and Blood Institute. The NHLBI provided more than 1000 MRI images from a broad sample set, including individuals of different ages and genders, and 1392 algorithms were submitted for consideration in the competition.
“We gave the data science community a challenge of unprecedented complexity and importance for the second annual Data Science Bowl,” said Josh Sullivan, senior vice-president at Booz Allen Hamilton.