IBM Unveils a ‘Brain-Like’ Chip With 4,000 Processor Cores. The TrueNorth chip mimics 1 million neurons and 256 million synapses that IBM calls “spiking neurons.”
...the chip can encode data as patterns of pulses, which is similar to one of the many ways neuroscientists think the brain stores information.
IBM Research: Neurosynaptic chips provides more information on the low power system architecture and potential applications:
This is similar to Qualcomm's Brain-Inspired Computing effort.
I had been reading about "internet of things" (IoT) connectivity standards groups (links below) when I came across Bridget Moorman's post Connectivity Standards Adoption Over Time.
I made an off-the-cuff comment because of the contrast I saw between the way other industries approach standards vs. the medical device industry. Bridget and Tim kindly responded with informative feedback.
Standards are hard, so I guess it's not easy for anyone.
The reality of healthcare interoperabilty is going mainstream: Apple Launches HealthKit To Share Vital Stats With Nike, Mayo Clinic:
The problem Apple will run into: No one agrees on how to measure even very simple health metrics, ...
TL;DR: iOS 7.1 to iOS 8.0 API Differences (search for 'HealthKit' on the page).
UPDATE (8/18/14); Apple - iOS 8 - Health
The Food and Drug Administration Safety and Innovation Act (FDASIA) required the FDA to develop:
a report that contains a proposed strategy and recommendations on an appropriate, risk-based regulatory framework pertaining to health information technology, including mobile medical applications, that promotes innovation, protects patient safety, and avoids regulatory duplication.
Here's the report: FDASIA Health IT Report (warning: PDF).
It looks like EMR/EHR vendors (administrative and health management functionality) don't have to worry about FDA regulatory oversight. The medical device category (of course) does:
FDA would focus its oversight on medical device functionality because, in general, these functions, such as computer aided detection software and remote display or notification of real-time alarms from bedside monitors, present greater risks to patient safety than health IT with administrative or health management functionality.
This is a good read: Healthcare Innovation Day 2014: Igniting an Interoperable Healthcare System (warning: PDF).
Healthcare is the one industry that’s been the slowest to adopt the intelligent methods we have in most other parts of our lives. How did the communications revolution that transformed industries such as banking, entertainment and telecom somehow leave healthcare behind?
Great question!
Here's the 'Call to Action' list:
There is a lot to do...
The Internet Of Things: The Real Money Is The Internet, Not The Things (my highlight):
The trick will be whether hardware companies will push hard enough for standardization so they can capitalize on services revenue. Companies that see themselves as pure hardware manufacturers are likely doomed, but those that see beyond the "things" to instead focus on the services built on the "Internet," the future is very bright.
Is this the new reality for the medical device industry as well? If data and interoperability are the future, maybe it should be!
What do these two things have in common? Apple exploring cars, medical devices to reignite growth. That's right, Apple!
... Apple is exploring ways to measure noise "turbulence" as it applies to blood flow. The company wants to develop software and sensors that can predict heart attacks by identifying the sound blood makes as it tries to move through an artery clogged with plaque, ...
The first 5 pages of search result on this topic turn up the above quote and Elon Musk. And of course the media hype kicks into full gear: Apple's iWatch could save your life.
Fortunately, sanity does prevail: No, An Apple Device Won't Tell You If You're Having A Heart Attack. Even Eric Topol weighs in:
The Internet of Things (IoT) is being propelled by the dramatic reduction in size, power consumption, and cost of networking and computing capability. Many of the devices listed in the Wolfram Connected Devices Project are health related. "Things" like weight scales and thermometers can make measurements from many objects, but when that object is a human body, they effectively become medical devices. The sensors that come standard on smart phones also fit into this category. All of these network-connected devices that record data from humans make up the Internet of Medical Devices (IoMD).
In most ways the invasion of technology in Healthcare is no different than how mobile digital capability is changing that way we all live. For Healthcare though, the potential benefits of applying these technologies to solve both the cost problem and to improve patient safety and outcomes are tremendous.
The number of technologies and innovations (health tracking apps and devices, home monitoring, medication management, etc.) that are contributing to these goals are too numerous to count. From a medical device perspective there are four primary areas of concern that need to be addressed as the IoMD grows.
1. Interoperability
As I've written about many times (e.g. Interoperability: Arrested Progress), health data interoperability is a key factor in realizing both cost reductions and improved patient outcomes. Unfortunately, medical data is notoriously complex, which makes effective communication between systems very difficult.
Another significant barrier to interoperability are EHR interfaces. The issue is that each EMR vendor has a propitiatory interface for consuming device data and associating it with a patient record. Without a direct device interface, data has to be manually transcribed into the record which is expensive and error prone.
This is a particular problem in the ambulatory EMR market because there are literally hundreds of vendors. Even if they all used a standard like HL7 there is still interfacing work that has to be done for each one. It is prohibitively expensive for any device company to develop and maintain that many interfaces.
2. Patient Safety
Because proper handling and presentation of medical data pose a safety risk the FDA has recently stepped in:
Most medical software applications, like the ones you might download to your Apple or Android phone, will not be affected by these regulations.
3. Privacy
Even though PHI (Protected Health Information) is protected by Health Information Privacy (HIPAA) laws most people consider their health a very private issue. Privacy concerns are a significant psychological barrier that must be overcome before sharing of medical data becomes commonplace.
4. Security
Reports like the one described in The Internet Of Things Has Been Hacked, And It's Turning Nasty are not encouraging:
Pinging one device brought up a login screen that said: Welcome To Your Fridge. She typed in a default password—something like “admin” or "adminadmin," Knight said—and suddenly had access to the heart of someone's kitchen.
The IoMD is not immune from this. Hacking Insulin Pumps And Other Medical Devices From Black Hat was big news last year. <TongueInCheek>If we're not careful search engines like Shodan will soon be discovering pace makers near real hearts!</TongueInCheek>
Final Notes
UPDATE (1/25/14): Also see: Digital Health In 2014: The Imperative Of Connectivity
UPDATE (2/19/14): I like “Medical Internet of Things” (mIOT) too: Keeping medical device designs relevant in a big data world migrating to outcomes-driven payment models.
I agree with Udi Dahan's Thoughts on a career in software development premise that developers will eventually be faced with a management "opportunity" at some point in their career.
It's too bad that this is usually a "fork in the road" decision. I don't think companies are necessarily trying to pigeonhole developers, but they certainly have specific roles (with associated job descriptions) they are trying to fill. It makes sense though. For large software projects, being a manager (and probably a scrum master) is a full time job. Put another way, if you try to split your time between being a contributor and a manager, you'll probably do both jobs poorly.
My advise is to make this type of decision with your eyes open. If you have a management opportunity and it's something you're interested in, take it. Treat it like the career change that it is. Get the additional training and improve your skills just like you would be doing if you were learning a new technology. Management isn't easy. It takes time and work get good at it.
Also, good companies do not pigeonhole technical managers. You will probably have the ability to switch back to a development or architect role as business needs and priorities change in the future. This could mean moving to a different company, but both you and your current employer know this. Switching from management back to a technical track will require yet another skills learning curve and a mindset change.
The way to create a Zigzag career is to go with the flow.
Whether you experience success, failure, or somewhere in between, each one of these transformations involves significant personal and professional growth. Also, no matter what you do always stay at the top of your game: Continuous Learning: 14 Ways to Stay at the Top of Your Profession.
Bringing artificial intelligence to mobile computing is a significant challenge. That's the goal of Qualcomm's new Zeroth Processors.
Mimicking the human nervous system and brain to allow computers to learn about their environment and modify their behavior based on this information has long been the goal of artificial neural networks. Whatever computing model is used to achieve this capability the real problem is one of scale. The human brain is estimated to have 100 billion neurons -- with 100 trillion connections. That is at least 1,000 times the number of stars in our galaxy.
These computational models can be implemented in software (e.g. Grok), but the ability to scale to the levels required for even simple human-like interactions is severely limited by conventional computing platforms. The Zeroth Neural Processing Unit (NPU) is a hardware implementation of the brain's spiking neural networks (SNN) method of information transmission. Integrating the NPU into computing platforms at the chip level would begin to address the computational and power requirements for these types of applications.
The goals of the Zeroth* platform are:
Achieving "human-like interaction and behavior" is an ambitious goal, but it seems like this is a good first step.
UPDATE (25-Oct-13): Good overview here: Chips 'Inspired' By The Brain Could Be Computing's Next Big Thing.
UPDATE (1-Jan-14): CES 2014: Intel launches RealSense brand, aims to interface with your brain in the long run
___________
* The name Zeroth comes from the science fiction Three Laws of Robotics. The First law was that "A robot may not harm a human being."
Asimov once added a "Zeroth Law"—so named to continue the pattern where lower-numbered laws supersede the higher-numbered laws—stating that a robot must not harm humanity.
We'll have to wait and see, but let's hope so!
