Implanting a bio sensor under a person's skin to monitor blood glucose and oxygen levels is a concept that is almost here. Powering such a device externally and reading the results is a technology borrowed from RFID chips and can work with a microprocessor with ample ROM and RAM. What is missing is the sensors themselves. These sensors must operate indefinitely without requiring consumable materials. One possibility is BioMatrix.
Two forces that support this concept are medical (obviously) and technical, from RFID technology and sensors on a chip.
Governmental forces impede development in two ways: First, a "tax the rich" policy reduces the incentive for entrepreneurs to take risks on new products, since if they succeed, financial reward will be reduced. Second, increasing regulations add to the complexity of any new product, especially in medicine.
Legal forces also reduce the incentive for new medical devices since anything new in medicine is a target for lawsuits.
The Delphi method would be appropriate here with experts chosen from all related fields, especially including silicon chip designers and experts in legal medicine. Closed collaboration allows experts from competing organizations to participate without trade secret issues.
Thursday, January 27, 2011
Friday, January 21, 2011
Neural Connectome
The following video may be found at http://www.ted.com/speakers/sebastian_seung.html
When I first started my quest for material on neural networks, I was afraid that there would be little available on a computer science perspective. I am pleased to find that there is an abundance of material!
This video shows the technological challenges of pursuing the brain. The medical motivations are obvious, since if this technology can be perfected, our ability to treat and diagnose diseases and other malfunctions of the nervous system will be greatly enhanced.
Although financial considerations exist in any research endeavor, the primary limitation in this area is lack of really brilliant people. There are many approaches to applying computer science techniques to biological neural networks, so there is plenty of room for many players.
The two points from the video that I would like to discuss are the ethical implications of this approach to defining who we are, and the technological requirements to achieve this sort of mapping for human beings.
When I first started my quest for material on neural networks, I was afraid that there would be little available on a computer science perspective. I am pleased to find that there is an abundance of material!
This video shows the technological challenges of pursuing the brain. The medical motivations are obvious, since if this technology can be perfected, our ability to treat and diagnose diseases and other malfunctions of the nervous system will be greatly enhanced.
Although financial considerations exist in any research endeavor, the primary limitation in this area is lack of really brilliant people. There are many approaches to applying computer science techniques to biological neural networks, so there is plenty of room for many players.
The two points from the video that I would like to discuss are the ethical implications of this approach to defining who we are, and the technological requirements to achieve this sort of mapping for human beings.
Saturday, January 15, 2011
Friday, January 14, 2011
Horizon Reports, 2009 & 2010
Does the heads up display in a military jet fighter count as Augmented Reality? In the 1950s, a "pipper" was projected on the inside of the windscreen as a computed gun sight to aid in dog fights with other aircraft. The computations took into account the relative velocities of the aircraft, projected bullet trajectories, etc. Not long after that, radar displays were also projected onto the windscreen along with an artificial horizon. The 2010 Horizon Report claims that the first examples of AR occurred in the late 1960s.
In a modern fighter like the F-22 a great deal of study has been directed to how to display the needed information without overloading the pilot and without requiring him to shift his eyes down to the instrument panel during a tense situation. Modern researchers into augmented reality should be aware of this body of information so they do not reinvent the wheel.
Games drive consumer versions of augmented reality, but there are obviously many other applications. Of particular interest to me are surgical techniques augmented in real time with super high resolution ultrasound. As an engineer, I know that ultrasound has a lot of potential for improvements in resolution and discrimination. If this information can be presented in 3D and real time to the surgeon, many new surgical procedures become possible.
My two favorite topics in the 2009 report, Smart Objects and Semantic-Aware applications, did not make it into the 2010 report. Smart Objects may suffer from the "solution in search of a problem" problem, while Semantic-Aware Applications may require more difficult development than was anticipated. In both cases, time-to-adoption horizon was estimated at four to five years. Our ability to predict that far into the future is questionable.
In a modern fighter like the F-22 a great deal of study has been directed to how to display the needed information without overloading the pilot and without requiring him to shift his eyes down to the instrument panel during a tense situation. Modern researchers into augmented reality should be aware of this body of information so they do not reinvent the wheel.
Games drive consumer versions of augmented reality, but there are obviously many other applications. Of particular interest to me are surgical techniques augmented in real time with super high resolution ultrasound. As an engineer, I know that ultrasound has a lot of potential for improvements in resolution and discrimination. If this information can be presented in 3D and real time to the surgeon, many new surgical procedures become possible.
My two favorite topics in the 2009 report, Smart Objects and Semantic-Aware applications, did not make it into the 2010 report. Smart Objects may suffer from the "solution in search of a problem" problem, while Semantic-Aware Applications may require more difficult development than was anticipated. In both cases, time-to-adoption horizon was estimated at four to five years. Our ability to predict that far into the future is questionable.
Subscribe to:
Posts (Atom)