Sign up FAST! Login

Data & Healthcare. Humans are not good when 500 variables affect a disease. We can handle three to five to seven, maybe.


http://venturebeat.com/2014/05/24/valley-heavyweight-vinod-khosla-says-replacing-doctors-with-data-crunchers-is-good-medicine/

Vinod Khosla: Doctors can't beat big data, machines - San Francisco Chronicle

“Biological research will be important, but it feels like data science will do more for medicine than all the biological sciences combined,” the Silicon Valley heavyweight said at the Stanford University School of Medicine’s Big Data in Biomedicine Conference Friday. “I may be wrong on the specifics, but I think I will be directionally right,” Khosla said.

Khosla has for a long time believed that machines armed with mountains of data will (and should) make most of the clinical decisions in the future, eliminating the need for most doctors. Humans, he believes, just can’t handle enough data to understand and prevent illness.

“Humans are not good when 500 variables affect a disease. We can handle three to five to seven, maybe,” he said. “We are guided too much by opinions, not by statistical science.”

Khosla may be “directionally” right that Big Data will end up being a big deal for healthcare, it’s one of the industries that has the most to gain from data science. That’s a big part of the reason the federal government is mandating the use of electronic health records — so that mountains of patient and treatment data can be captured in digital form, and studied.

We imagine huge databases pulling health and non-health information from a thousands of sources, predicting when disease will strike, and forecasting the efficacy of treatments.

The scary thing about Big Data is that it’s cold science. Big Data might predict the incidence of diabetes by learning that it correlates with some random marker like purse thefts or banana imports. Big Data doesn’t care, as long as the prediction works.

A human doctor is more likely to try to predict diabetes by tying it to factors that bear a natural, causal, relation to the disease, like obesity or junk food intake.

In all likelihood both of these approaches — the cold science and the warm — will end up being useful in making people healthier in the years to come. And doctors will more than likely be the ones to formulate the questions we need to ask of the data, and the ones to figure out how to put the answers to work at ground level.

Kohsla’s comments hit an old sore spot for doctors, who have been grousing for years about “cookie cutter medicine” driven by new health IS systems. Any doctor will tell you that practicing medicine is both science and art.

Naturally, some of the doctors in the audience at Stanford didn’t care much for Khosla’s comments. “I don’t agree with 80 percent of your remarks,” said one.

http://www.sfchronicle.com/business/article/Vinod-Khosla-Doctors-can-t-beat-big-data-5501778.php

Here's a little DIY job:

human technology implants

Tim Cannon becomes the first human to implant technology into his body. He calls himself a "biohacker" after implanting a big electronic chip beneath the skin of his arm.

Using a device which they dubbed as Circadia 1.0, built by a Pittsburgh-based firmGrindhouse Wetware where Cannon works as a software developer, he worked with a team of piercing and tattoo specialists to have the battery-powered device implanted beneath his skin. The procedure was done with no anesthesia and used only ice to decrease the pain. He tried to get a doctor for a surgery but was refused so he ended up with this team.

The device is designed to capture data from Cannon's body and send it to any Android mobile device. While there are existing wearable devices that can serve the same purpose, the implant is still better as it allows the user to control the data collected by the device.

Steve Haworth conducted the surgery for Cannon. He specializes on 3D tattoos and the metal Mohawk. He was convinced to do the surgery after Cannon assured him that they can use ice as anesthesia.

Cannon considered the procedure a dream come true as he wanted to be a cyborg when he was young. "Ever since I was a kid, I've been telling people that I want to be a robot," Cannon told The Verge. "These days, that doesn't seem so impossible anymore."

Cannon's team is now working on further improving the device. They are also working on a pulse monitoring device and have made the implant smaller than what Cannon had to make it more user-friendly.

"We have been working on the Circadia Chip for 18 months, needing only a fraction of the costs that big companies would use for this," Cannon told Motherboard. "The same will go for our next projects and an artificial heart is a goal for us for the next decade."

Cannon also mentioned that they are planning to sell Circadia 1.0 for about $500 in the next few months. Haworth said the surgery would cost $200.

Here is a video of the Motherboard documentation of Tim Cannon procedure.

And then check this one out: http://pandawhale.com/post/29080/i-constantly-monitor-my-blood-sugar-level-ive-collected-more-than-565k-rows-of-data-about-my-blood-sugar-i-am-big-data 

Stashed in: Big Data!, @vkhosla, Big Data, Healthcare, Correlation is not causation.

To save this post, select a stash from drop-down menu or type in a new one:

This is well said:

The scary thing about Big Data is that it’s cold science. Big Data might predict the incidence of diabetes by learning that it correlates with some random marker like purse thefts or banana imports. Big Data doesn’t care, as long as the prediction works.

A human doctor is more likely to try to predict diabetes by tying it to factors that bear a natural, causal, relation to the disease, like obesity or junk food intake.

In all likelihood both of these approaches — the cold science and the warm — will end up being useful in making people healthier in the years to come. And doctors will more than likely be the ones to formulate the questions we need to ask of the data, and the ones to figure out how to put the answers to work at ground level.

Is it safe to say that doctors look for causations whereas data scientists look for correlations?

You May Also Like: