This article explores advances in neural engineering research and is based on my interview with Dr. Metin Akay, Founding Chairman of the new Biomedical Engineering Department and the John S. Dunn professor of biomedical engineering at the University of Houston. His discipline unites the fields of engineering, computer science, physics, chemistry, and mathematics with cellular, molecular, cognitive and behavioral neurosciences.
One expected result of these fields converging is to lower health care costs. Another is to extend life, but as Dr. Akay put it, “While it’s very important to live longer, it’s much more important to have quality of life.”
The Aging Problem
When I was born 65 years ago, the world population was about 2.5 billion people. It’s nearly 7 billion today. At the same time, average lifespan has extended 30 years. The aging problem will worsen as post-WWII baby boomers retire at the rate of 10,000 per day, because soon there will be more old people over age 65 than young productive workers to support us, drive our economies and fund our governments.
These aging trends raise two types of questions. One relates to the increased demand and cost of medical care for the increasingly elderly population. The other relates to how prepared the younger generation is to fill the void.
“We are an aging society with almost 35 million Americans over age 65 today, and by 2040, we’ll have 70-75 million Americans above 65,” Akay said. “Of course we love our elderly and want them to live long and healthy. We also want them to live with us and be taken care of well, but we are all professional people and need to work.” Therefore, the advanced assistive technologies are needed to enable the elderly to live independently.
I’ll discuss that issue in a follow-up article on Aging and the Future Workforce.
Lowering Costs, Really?
When Akay said, “Technology is the driving force to reduce health care costs” I had to challenge that claim, because some people may argue that modern technology adds to the costs.
In the past, he said, when people needed heart surgery, it meant opening the chest and required a 10-15 day hospital stay (or more), with hospital rooms that are much more expensive than 5-star hotels. Today, thanks to less invasive surgery techniques, the patients can be safely discharged in just 2-3 days. And although surgery is always expensive, if we reduce rehabilitation time and the number of days in the hospital, we reduce health care costs significantly.
Coronary angioplasty. As we age, a waxy plaque can build up inside any artery in the body, including the coronary arteries that feed oxygen-rich blood to the heart. As the plaque builds up and hardens, it reduces the flow and can cause chest pain or angina. If the plaque ruptures, a blood clot can block the coronary artery, and that’s the most common cause of a heart attack.
Angioplasty is a minimally invasive procedure that can restore blood flow to the heart. It involves threading a thin, flexible catheter with a balloon stent at its tip through a blood vessel to the affected artery. Once in place, the balloon is inflated to compress the plaque against the artery wall and restore blood flow through the artery. In the near future, the procedure will be able to also direct medications to the specific location where it’s needed and even implant a sensor device with wireless radio communications to remotely monitor plaque buildup.
Epilepsy. It’s a chronic neurological disorder characterized by seizures and resulting from abnormal hypersynchronous neuronal activity in the brain. Although epileptic seizures can be usually controlled with medication, regrettably, currently available medications do not control some epilepsy. Thanks to federal funding in Obama’s Brain Mapping project, scientists can now access desperately needed resources to help them understand the mechanism and cause of this deliberating disease.
Telehealth in Developing Nations
Telehealth technology allows people to be diagnosed and triaged at a distance, even from home, thus further reducing the need for hospital visits. Technologies developed for Emerging Markets such as Bangalore, India and Sub-Saharan Africa must be more affordable than here in America. Applications using text messaging, for example, overcome the limitations of not having high-speed broadband networks, but other issues include size, weight, portability, cost, and ease of use.
It’s too difficult to take an ultrasound machine or multi-lead EKG machine to the outer regions of Mongolia, so Dr. Akay is excited about new smartphone and tablet based solutions that are smaller and cost much less.
Imagine a comprehensive, clinically relevant, well-patient checkup using only smartphone-based devices, which are easy to transport and relatively inexpensive and easy to use. That was the theme of an April article on the Checkup of the Future. The article featured the a digital stethoscope from ThinkLabs, the Withings blood pressure monitor, The $1 plastic EyeNetra attachment for testing visual acuity, CellScope for ear exams, Masimo iSpO2 pulse oximeter for tracking and trending blood oxygenation, the AliveCor portable EKG, and more. These smartphone attachments may prove especially useful in regions where today’s medical equipment is too large and expensive to get into rural clinics and villages.
Even with the advances shown in the video, Akay complained that medical imaging technology is still very expensive, but he is thinking on an entirely different scale – a Nano scale – where image sensors and electronics eventually get to the size of a single cell and can be used to explore deep in the brain and other areas of the body.
Referring to the endoscope pill, I asked Akay if miniaturization will eventually allow man-made electronic devices to shrink to the size of individual cells that then flow through the blood stream seeking out cancer cells and other problems, and depositing medications in precise locations. He agreed with that vision and said one of the challenges he still faces is imaging at that scale. Rather than just use imaging techniques to watch the activity of a cluster of neurons, for example, he also wants to sense proteins (much smaller than cells) as biomarkers of neuron activity.
Today, the task of shrinking devices to the size of cells almost seems an insurmountable task, but for reference consider the miniaturization of the tiny digital cameras on smartphones, with capabilities approaching that of SLR cameras. But it’s not just about making things much smaller, Akay says. It’s also about making them cheaper, and that depends also on market size.
Poor and developing countries offer especially interesting markets for new medical devices, and developers see the sheer size of these markets as justification for their work to lower costs, knowing that lower costs also allow their products to be more widely adopted globally.
Personalized & Preventative Medicine
Tests that prevent disease or detect it early can dramatically lower health care costs more than improved treatment. One example is an easy and affordable ($0.10 per test) solution that’s being developed with public health officials in China to test food spoilage and prevent food poisoning and other medical problems.
Dr. Akay noted that the accuracy of new tests and technologies is critical. “The issue is to ensure reliability, because it’s not good if the test fails 50% or less of the time.” To ensure that, he says bio-engineering is moving toward the molecular level, then integrating the molecular level info with those at the cellular and system levels (multiscale).
Thanks to genetic and biomedical research and the power of much larger computers, a single drop of blood or saliva can now be used to determine if someone has a high or low chance of suffering from a medical condition and can also help in prioritizing treatment programs with the highest chance of success.
Earlier I wrote about the recently announced breakthrough discovery by 16-year-old Jack Andraka, who developed an early stage diagnostic test that costs just $0.03, takes just 5 minutes, and can detect pancreatic cancer with 100% accuracy in clinical trials so far.
Big Data Analytics
When looking at cancer therapy and the large number of treatment options, it’s important to note that chemotherapy works well on some people but not as well on others, and while it can destroy cancer cells, it’s also quite harmful to cells in other parts of the body.
Just as technology at the molecular level can help by directing treatment and medication to precise locations in the body to minimize effects elsewhere, tech innovation in supercomputing, artificial intelligence, and Big Data analytics gives us new ways to personalize treatment based on immense databases from an individual’s genome and real-time sensor monitoring, and the world population having the same condition.
“Everyone has different genetic structures. Certain genetic pathways are active or not active in certain patients, so understanding the differences in these pathways help physicians determine the right drugs and therapies for the right person.” As we move toward personalized medicine, individualized drug formulations can be tailored to each patient, with advanced imaging monitoring to ensure that drug is metabolized appropriately.
Understanding Brain Functions
We barely discussed IBM’s Watson supercomputer and spent most of the time talking about how molecular sized technologies can help improve our understanding of brain functions, especially in areas deep inside where the hippocampus is located.
The hippocampus belongs to the limbic system and plays important roles in connecting emotions and senses, such as smell and sound, to memories, as well as in consolidating information from short-term memory to long-term memory and spatial navigation.
If compared to a computer system, the hippocampus might function like RAM (random access memory) while the cerebral hemisphere functions more like permanent storage on the hard disc. That analogy helps explain why Alzheimer’s disease, which damages cells in the hippocampus, causes short-term memory loss before long-term loss.
How Sleep Affects Brain Health
New research with mice shows that during restorative deep sleep, the flow of cerebrospinal fluid in the brain increases dramatically, washing away harmful waste proteins that build up between brain cells during waking hours. What apparently causes this is that brain cells actually shrink during sleep, allowing more space between them for this fluid. This extra space may also allow researchers with molecular sized sensors to study brain activity and learn more.
The results of this new research appear to offer the best explanation yet of why animals and people need sleep. It also seems to help explain the association between sleep disorders and brain diseases, including Alzheimer’s.
“Right now we can not only record the electrical activity of individual neurons, but with optogenetics, we can also image the behavior, shape and area of the cells and monitor the genetic biomarkers.”
The mind-machine interface
Today’s computers don’t learn and reason like the human brain does. They’re only fast at following pre-programmed instructions. They don’t really think, but that’s changing. IBM’s Watson supercomputer is famous for beating the best human players in the game Jeopardy, a feat that exploited its artificial intelligence learning ability, understanding of the English language, and Big Data analytics ability. Now IBM is applying Watson to medical diagnostics.
I mentioned Watson and provided brain-computer comparisons because of my interest in the future man-machine interface. By extrapolating Moore’s Law, which sees computer chips getting faster, cheaper and smaller exponentially, futurist Ray Kurzweil predicts that by 2023, a $1,000 computer will have the power of the human brain and by 2037 a $0.01 computer will have that power. I hypothesized that if a tiny embedded computer were built to the size of a cell, it might interface directly with a neuron, potentially giving each neuron the reasoning and analytical ability of a human brain, with the ability to then connect with the billions of other neurons in a computer-like mesh network.
Dr. Akay did not entirely agree with my vision, saying that, “We don’t know today exactly how many neurons we have in our brain. We’re guessing 1011 neurons (100 billion) in just 1.1 to 1.3 liters of volume, with the brain only weighing about 2 pounds.” He described energy consumption as easily as important as small size and suggested that packing so many neurons/nodes into a brain-size computer system using today’s technology would cause it to “explode” when powered up, but being a computer technologist, I disagreed.
Putting 100 billion neurons into computer terms, that’s 100 GB (gigabytes), and for perspective, Apple already packs almost 2/3 of that memory (64GB) into its iPhone 5S, which is 0.18 liters in size and weighs 3.95 ounces (12-15% of the size & weight of a human brain).
I shared an image comparing a 1970’s era IBM mainframe computer and an Apple iPhone 4. The $3.5 million mainframe consumed lots of electric power, requiring special air conditioning and even liquid cooling of the process itself, but the much smaller and cheaper iPhone can execute machine instructions more than 5,000 times faster and gets up to 10 hours of talk time and 250 hours of standby time with just a rechargeable battery. The iPhone also has three types of network connections: cellular (for metropolitan area networks), Wi-Fi (for home & office networks), and Bluetooth (for short-distance networks, such as connecting to medical sensor devices).
As we continued to discuss connecting computers and neurons, it was clear that Dr. Akay does not see small computers as totally replacing brain function with existing technology in the near future but augmenting brain function instead. That could include bypassing damaged neurons in the hippocampus or spine, using cognitive implants to artificially connect short & long-term memories or helping someone with severe spinal cord injuries to walk again.
The Biggest Challenges
Even bigger than the challenge of shrinking size, reducing costs, connecting to neurons, or monitoring proteins & biomarkers are challenges in material science, Akay said. That’s so the body’s immune system doesn’t treat the device as a foreign object that must be destroyed or contained. So rather than rely on traditional semiconductors and electrical contacts, the trick is to develop biochemical sensors and neurotransmitters that are compatible with the body to monitor neural activities and neurotransmitters in the body.
IBM, a material science leader, last year demonstrated an organic data storage device that theoretically could cram 455 exabytes of data (1 exabyte is 1018 bytes or 1 million gigabytes) onto each gram of the double-stranded DNA molecule. If that sort of technology becomes mainstream, Kurzweil may be correct in predicting that by mid-century, a $1,000 computer could have the power of the human race. (‘Just dreaming.)
Another big issue Dr. Akay mentioned is providing power. Where neurologists can now control the misfiring of rogue neurons that cause epilepsy by implanting small electrodes in the brain cluster of misbehaving neurons, they still need a power source and must resolve battery-life issues. Today they do this by connecting the electrodes to a power source worn inside or outside the body. Researchers hope in the future to use other more natural power sources, such as relying on movement, warmth, or chemical reactions to generate power internally.
The power issue also influences the use of wireless communications. When restricted to very small amounts of transmit power, the signals won’t go very far, so a wearable device is needed to act as a gateway between the implanted sensor(s) and remote monitoring systems. Another way to conserve power is to only transmit occasionally or as needed, as opposed to transmitting continuously.
The importance I see in these exciting neuroscience and neuroengineering achievements is improving our understanding of the man-machine interface, solving the problem of implanting a sensor device in the body without the body rejecting it, discovering better ways to power the device, personalizing medicine by depositing the right medications at the exact spot so they don’t hurt other parts of the body. All of those things can help reduce the costs of health care here and in emerging nations while also helping people live longer. But that brings up an important issue of also helping them live better as they age. We never got around to discussing that in much detail. Maybe next time.
About Dr. Metin Akay
Prof. Metin Akay is the founding chair of the new Biomedical Engineering Department and the John S. Dunn professor of biomedical engineering at the University of Houston. He received his B.S. and M.S. in Electrical Engineering from the Bogazici University, Istanbul, Turkey and a Ph.D. degree from Rutgers University.
Dr. Akay has played key roles in promoting biomedical education in the world by writing and editing several books, editing several special issues of prestigious journals, including the Proceedings of the IEEE, and giving more than a hundred keynote, plenary and invited talks at international conferences, symposiums and workshops regarding emerging technologies in biomedical engineering.
He is the founding editor-in-chief of the Biomedical Engineering Book Series published by the Wiley and IEEE Press and the Wiley Encyclopedia of Biomedical Engineering. He is also the editor of the Neural Engineering Handbook published by Wiley/IEEE Press and the first steering committee chair of the IEEE Trans on Computational Biology and Bioinformatics.
“Dr. Phillip Alvelda, Program Manager in DARPA’s Biological Technologies Office (BTO), discusses the potential of next-generation neural interfaces to improve quality of life for people and revolutionize how we engage with machines. The talk was part of a two-day event held by BTO to bring together leading-edge technologists, start-ups, industry, and academic researchers to look at how advances in engineering and information sciences can be used to drive biology for technological advantage.”
For obvious security reasons, Alvelda’s talk avoids any mention of possible military applications of a brain-computer interface, but imagine if soldiers on one side of a battlefield, or HQ commanders, could see and hear what those on the front lines see. Imagine super-human vision and sound, whether that’s from an implantable brain interface or external peripheral devices like a smart contact lens or hearing aid. Beyond night vision, soldiers might be able to see non-visible light and energy, like infrared, ultraviolet, or even radio waves.