Open source is powering a revolution in medicine and health care in multiple ways. Open source software and methods make large-scale collaborative research projects feasible, multiplying the brainpower applied to a project, expanding the data pool, and creating transparency and accountability. This is a huge win for the advancement of new treatments and cures, and cutting the costs of research. Open source practice and records software cut the costs of running medical practices, and puts practitioners in charge instead of software vendors.
Slideshow: Bio-Health Aids for the Digital Age
Slideshow: 11 Open Source Programming Tools on the Rise
This is a marked contrast to the traditional secretive, highly-competitive methods of research and medical product development. It's expensive to bring new drugs and devices to market. Research and testing can take years, and FDA approval is expensive, bureaucratic, and time-consuming. But keeping everything in-house promises big profits for the winners. So the old ways persist, but at a high cost to people who can't afford expensive patented treatments, in side effects and defects that are not discovered until after a new medicine or device is released into the market, and in entire categories of diseases that are not studied because the profit potential isn't big enough.
It's not that one way is good, and the other is bad. There is room for both. However, the great power of open source is you never know where breakthroughs will come, which is the whole point of letting everyone play.
Raven the surgical robot
The University of Washington Biorobotics Lab and the University of California, Santa Cruz Human Bionics Laboratory are jointly developing Raven the Surgical Robot. Raven is an amazing machine that provides a 3D real-time view inside the human body, performs surgical procedures, and operates remotely over a computer network. Raven uses a high-speed graphics processor, similar to the graphics processors in computer gaming systems, to stream a continual series of ultrasound images. The original Raven came out in 2005. Raven II was released early this year (2012) in a smaller, more dextrous version. Seven models were built and sent to seven different biorobotics labs (University of Washington, UC Berkeley, UC Santa Cruz, UCLA, Johns Hopkins, University of Nebraska, and Harvard) for collaborative study and development, including networked experiments over the Internet. Raven runs on a real-time Linux kernel and the Robot Operating System, and is designed to provide a common platform for collaborative research in robot assisted surgery.
There is great potential for robot-assisted surgeries (RAS). Two common surgeries, prostate removal and hysterectomies, have been routinely performed using RAS for several years. They use smaller incisions, and are less invasive, so there is less pain and patients heal faster. Giving surgeons a more accurate view inside the body should result in more precise surgeries with less damage, and someday the ability to perform surgeries that are now considered too risky because they're too close to vital structures such as arteries, nerves, or brain tissue. Heart surgery is exceptionally challenging because the heart is in constant motion. Traditional heart surgery involves hooking up the patient to a heart-lung machine and stopping the heart. Raven could make "beating heart" surgery a reality by compensating for the movement of a beating heart.
Raven could also enable remote-control surgery over the Internet or other networks. Before Raven, surgery robots were very expensive and proprietary, and impossible to modify or build research on. The open Raven platform should result in significant progress. This YouTube video gives a brief demonstration of Raven II in action.
Personal health care revolution
Want to be a citizen scientist? Join a study? Organize a study? Probe your own genome for health risks, learn surprising facts about your ancestry, and share your findings with other citizen scientists? Thanks to DIYgenomics you can. Some of the currently running studies are "Aging #1: Telomere length and telomerase activation therapy", "Aging #2: Risk reduction for common aging conditions through monitoring and intervention", "MTHFR / Vitamin B deficiency and linkage with homocysteine levels", and "Knowledge generation through self-experimentation". The various studies rely on open source-based services such as Google Docs and Genomera, which provides a free platform for hosting open health studies.
Many studies incorporate personal DNA tests from retail DNA labs such as 23andme, Decodeme, and Navigenics. DIYgenomics even has a mobile app (iPhone and Android) for comparing profiles from all three side-by-side.
The price of DNA testing has dropped dramatically, for example a mere $207 plus a bit of spit gets you a partial genotyping from 23andme. If you want a complete profile that will cost a few thousand dollars (which is not provided by 23andme), which is a remarkable drop from the six-figure price tags of just a few years ago. The 23andme profile looks for selected SNPs, which are single-nucleotide polymorphisms (SNPs). You can think of SNPs as individual variations in your genome sequence, the differences that make you you and not someone else. Then 23andme generates a report from these results that shows your risk factors for 237 diseases and conditions, whether you are a carrier of any of these, and potential responses to certain drugs. You also get information on your ancestry. Sadly, it is unlikely that you will learn you were descended from exotic royalty. But you may learn you have certain traits or risk factors, and then investigate further. You might get genetic counseling, make some lifestyle changes, get certain health screenings, and share your experiences and findings. It's a revolution in personal health care because we can get all kinds of amazing information about ourselves at a low cost, and without gatekeepers.
DIY citizen science will not replace clinical trials or scientific research, but it could contribute useful data. It's not a new concept; for example archeology, paleontology, and astronomy all have a tradition of incorporating amateur contributions. The founder of DIYgenomics, Melanie Swan, researched how these citizen studies need to be conducted so that they will be taken seriously by scientists, and be published in scientific journals. Her paper Crowdsourced Health Research Studies: An Important Emerging Complement to Clinical Trials in the Public Health Research Ecosystem goes into detail on the value of citizen science.
Open source drug discovery for malaria
Tropical diseases take a big toll in illness and death every year, but commercial medicine is not very interested in them because the profit potential is so low; the people affected by them are mainly in poor countries. About 1% of newly-developed drugs are for tropical diseases, but malaria, respiratory infections, HIV/AIDS, diarrheal diseases, and tuberculosis are leading killers in lower-income tropical and sub-tropical countries.
Open Source Drug Discovery for Malaria, OSDD, was launched in 2011 as a worldwide hub for malaria drug research. The first participants were the Todd Research Group at the University of Sydney and the Medicines for Malaria Venture. The Medicines for Malaria Venture (MMV) is a global organization with partnerships with universities, research labs, and pharmaceutical companies. OSDD makes good use of open source technologies: Twitter, Google+, MediaWiki, and The Synaptic Leap, which hosts a number of open source biomedical research communities, is built with the Drupal content management system.
Dr. Matthew Todd, head of the Todd Research Group, already has a major open source drug success story, and that is the inexpensive synthesis of enantiopure praziquantel, the drug used to treat schistosomiasis (Bilharzia). This is a debilitating disease that affects about a half of a billion people worldwide. Enantiopure praziquantel is now on the World Health Organization's list of essential medicines, and is distributed by WHO and the Gates Foundation. Take a look at the list for children -- it's an eye-opener. So are the leading causes of death, sorted by income. The diseases of poverty and the diseases of affluence are very different.
Silver bullet for cancer
There are always dark rumors about cures that are suppressed because sickness is so profitable. Big pharma, the government, men in black -- whoever is doing the suppressing is very effective, because cancer is a leading killer worldwide. (Tobacco use is the leading preventable cause of cancer, just in case anyone still harbors any doubts.) The beauty of open source is it doesn't matter if it's true or not, because once information is released into the wild it can't be hidden. A great illustration of this is the research launched by Dr. Jay Bradner at the Dana-Farber Cancer Institute, which to me, seems like a genuine breakthrough. There are hundreds of different cancers, and treatments are specific to each type of cancer. Obviously this is inherently difficult and expensive, so what if there is a "silver bullet" that cures all cancers? The quest for a cancer silver bullet is not new, but Dr. Bradner may have found it by asking "How does cancer know it's cancer?" A cancer cell has all the genes necessary to become any other kind of cell, so as it reproduces, how does it know to keep reproducing as a cancer cell?
The answer is the same for every cell in the body: each cell has a mechanism for remembering what it's supposed to be, so heart cells keep on growing as heart cells, and skin cells, and fat cells, and so on. So Dr. Bradner's team acquired a batch of midline carcinoma cells to experiment on. Midline cancer is rare enough that it's not an attractive target as a moneymaker. They figured out how to give the cancer cells amnesia by inserting a new molecule they had developed called JQ1, and the cancer cells forgot they were cancer and became normal cells.
Naturally this was very exciting, but it was just the beginning. Following the open source principal of "release early, release often" they mailed samples of compounds based on JQ1 and other compounds to other labs, published a detailed paper, and now 70+ labs and pharmaceutical companies worldwide are working on this. These are not silver bullets for all cancers, but they are promising for an entire category of cancers that includes midline carcinoma, acute myeloid leukemias, and multiple myeloma.
From research to a pill for people is always a long process. Some researchers just want to do research. Dr. Bradner is working with Tensha Therapeutics to bring new cancer treatments to market. Tensha already has two drugs in clinical trials from this new research, a topical drug for lymphoma of the skin and an oral medication for multiple myeloma. Dr. Bradner gives credit to open source methods in this inspiring Ted talk.
"This string of letters and numbers and symbols and parentheses that can be texted, I suppose, or Twittered worldwide, is the chemical identity of our pro compound. It's the information that we most need from pharmaceutical companies, the information on how these early prototype drugs might work. Yet this information is largely a secret. And so we seek really to download from the amazing successes of the computer science industry two principles: that of open source and that of crowdsourcing to quickly, responsibly accelerate the delivery of targeted therapeutics to patients with cancer."
Covering open source medicine would fill a book, so here are some references to a few more worthy projects.
A* Trisano helps the Centers for Disease Control monitor outbreaks of infectious disease, environmental hazards, and bioterrorism attacks.
A* Dr. Insup Lee of the University of Pennsylvania is working with the FDA on the development of a reference model-based approach to software conformance checking for producing highly-reliable software for medical devices, using a drug infusion pump for the prototype.
A* Medsphere gives healthcare administrators, clinicians, and developers a place to meet and collaborate on the administrative side of medicine.
A* The OpenEMR electronic health records and medical practice management software is a fully integrated electronic health, records, practice management, scheduling, and electronic billing application.
A* PLoS Medicine is a peer-reviewed, international, open-access journal that publishes original content and analysis.
A* Brainstorm is a free, open source cross-platform Matlab toolkit for Magnetoencephalography (MEG) and Electroencephalography (EEG) data visualization and processing.
A* The Veterans Administration is the largest healthcare organization in the US, and it runs on a single medical records network powered by the open source Veterans Health Information Systems and Technology Architecture (VistA). Over 300,000 medical staff use VistA every day while treating over 6 million patients.
A* Midland Memorial Hospital in Midland, Texas is one example of a successful VistA implementation. They spent $6.3 million, which included 680 workstations, plus bar code scanners and printers, and Red Hat Linux. If they had gone with proprietary software it would have cost $18-$20 million just for the software.
A* The Indian Health Service (IHS) uses a fork of VistA called Resource and Patient Management System (RPMS), which is deployed at more than 600 medical facilities. The main difference is the VA treats military veterans in hospitals, while IHS treats all people of all ages in small clinics.
At a glance: Open source in health care
How much penetration has open source made in medicine? It's always hard to measure because few OSS projects track usage. The 2008 study An Empirical Investigation into the Adoption of Open Source Software in Hospitals by Gilberto Munoz-Cornejo, Carolyn B. Seaman, and A. GA1/4nes Koru (University of Maryland, Baltimore County), using a combination of surveys and interviews, came up with these figures:
A* 23% of hospitals have adopted open source software
A* 90% use in-house IT staff
A* 76% of IT staff have five years or more experience
A* 67% of OSS users have IT budgets of 3% or less
A* Most popular software is general-purpose software like MySQL, Linux, Apache, PHP and Perl (LAMP stack) and Firefox