A Vision of Our Mobile Future

Smartphones represent an enormous opportunity for the creation of field-portable, compact and cost effective analytical instrumentation of the type that you would normally find only in laboratories or hospitals. Such applications have the potential to tackle the lack of analytical and diagnostic capability in certain parts of the world or in field settings. The absence of such services is not just because advanced instruments are very expensive; beyond the initial expense, there is also the requirement for solid infrastructure, which, in developing countries, is often lacking.

Big numbers and big challenges

No-one could have predicted the current status of mobile telecommunications 10 or 15 years ago. The numbers are simply staggering: fifteen billion cell phones have been sold and there are currently seven billion cell phone subscribers worldwide, more than 75 percent of whom are in developing countries, despite a lack of basic infrastructure – or even roads in some cases. In such countries, cell phones are the most advanced technology that you will find; phone towers, communication networks, and mobile power stations for charging cell phones appear to have found their way into every corner of the globe.

Cell phones are extremely cost effective. The sheer economy of scale and fight for market share have driven unprecedented strides in technological advancement and capability at amazingly low cost. Let me illustrate exactly how cheap this technology has become: if you were to somehow magically remove three zeros from either the number of cell phones sold or the number of subscribers (that is to say, replace billion with million), the cell phone in your pocket would cost you the same amount as a high end car.

The megapixel count of cell phones has been doubling every two years for the last 10 years (from 0.2 to 40 megapixels). So, if, like me, you’re a researcher who’s interested in developing portable high-end microscopes, the constant improvement in cell phone performance offers regular opportunities to push for more and more functionality. We can now routinely see single viruses and sub-100 nanometer florescent particles using cell phones. Admittedly, these cell phones are very high end, but they have enabled us to expand the boundaries of mobile imaging, sensing and diagnostics. Virus imaging is no simple task so it is a real milestone that proves the worth of our approach and the potential that the technology has in other areas, such as environmental monitoring and materials science.

One of the next steps is commercialization and deployment of existing instruments and designs, and it’s already happening to a degree. There are commercially available applications and hardware to convert cell phones into laboratory instruments. For example, I co-founded a company called Holomic LLC (www.holomic.com, see bottom of page: Introducing Holomic), which develops devices to image and quantify lateral flow immunochromatographic assays. Such cell phone-based systems can quantify analytes at concentrations in the parts-per-million or even parts-per-billion range, depending on the test of interest.

Once this and other devices gain regulatory approval, it’s not hard to imagine the rapid rise of “off-the-shelf” consumer products for a number of different applications from health monitoring to food analysis.

Ironically, one of the biggest barriers to the development of cell phone-based technologies is the very fast rate at which cell phones are evolving in terms of the hardware and software that they use. This is, of course, at the heart of the business model for providers and carriers. In diagnostic applications, however, stability is a major requirement. If we wanted to develop an application for, say, the Samsung Galaxy S4, we need to know that it would still be available in its current guise for at least the next five years. This time is required to develop, test, gain regulatory approval and market our application while users still have access to the relevant phone model. However, the Galaxy S range is likely to evolve significantly over just the next two years – the S5 is already on shelves – and there is no real end in sight to this marketing strategy.

There is an old saying that “every challenge is an opportunity.” New ventures could take advantage by taking control of the billions of used handsets and smart devices, communicating with the industry, discovering its needs, and offering a regulated supply chain to ensure that biomedical device manufacturers have access to the smartphones they need. Forget recycling – how about diagnostic upcycling? In this way, used phones become the hearts and brains of new portable analytical systems rather than add-on devices being made available to consumers.

Another solution to the problem could present itself, if whispers in certain circles about phone modularization come to fruition: imagine that rather than constantly changing phones, we could simply upgrade or change modules within an endoskeleton. Google has already staked out this potentially fertile field with Project Ara (www.projectara.com) – a forum that aims to bring together module developers with that exact aim. Clearly, from a diagnostic device point of view, this would be a very positive development.

Big data and big players

One of the reasons that mobile health tools are better than laboratory-based instruments that perform the same tasks is in the collection and use of data: mobile tools are inherently connected. The wireless connectivity of cell phones coupled with smart and secure servers means that rather than working with a single disconnected instrument or sensor, an entire network of instruments from all over the world can be accessed. Reference libraries would virtually self-assemble and databases would get richer and richer, enabling increasingly sophisticated analysis, such as the self-classification of images or signals, and automatic flagging of risk signatures.

There is a nonlinear threshold beyond which machine learning becomes very powerful – something that Google has taught the entire world. To breach that threshold requires progress on both the technology side and in terms of deployment. By bringing analysis to the masses at a fraction of the cost and by stabilizing the technology, the output of big data (and the analytics and machine learning that will result) will benefit not only the users, but also those who collate information for large-scale studies to discover wider patterns and trends. The new opportunities presented by such large amounts of networked analytical data and the potential size of the overall impact is hard to exactly predict right now. But perhaps a simplistic musical analogy is in the difference between only being able to access your own CD collection in the 1990s to having the ability to listen to almost any song ever recorded today…

Google, Apple, Samsung (and others) are all building collaborations in the medical diagnostics area and working on products and business models behind closed doors. They have the cash and the muscle to make waves, and the outcome may look like the phone market: who’s winning or losing at any given time will depend on the user interface, the relevance of the data captured, the level of integration into the consumer’s life, and the “coolness” factor. If you remember when the iPhone was introduced, it was a different kind of phone and a different way of interfacing with a computing device – and that lit the fuse for an explosion of innovation.

I anticipate a fragmented market, which means that we will see leapfrog advances from these giant companies driven by the desire to be the first to present the next ‘big thing’. The next couple of decades will be a frantic struggle to be increasingly involved in the consumer’s daily life and routine – to the point of monitoring the bodily fluids as well as biochemical and physical signals that we leak over the course of the day – and making insightful and actionable ‘sense’ out of the resulting data.

Where is health care in all this?

Our technological future should not strive to replace health care professionals. Rather, it should improve their performance by providing better, faster diagnostics and more in-depth patient data. Mobile diagnostics will simply bring in new complementary tools for the medicine of the future, driving us closer towards preventative health care.

The idea that we could replace people with gadgets and algorithms is a dangerous and misleading one that goes against the fundamental and centuries-old philosophy of medicine, which is all about “feeling empathy for the patient”. We have to be very aware of our continued need for the human touch. I certainly don’t want to live in a world where we replace doctors and other healthcare providers “entirely” with robots, no matter how advanced artificial intelligence and machine learning becomes. But surely such a view is not in conflict with the fact that health care delivery can be significantly improved with technology and new instruments that assist professionals with their medical practice.

Certainly, regulatory agencies will be strict with new diagnostic devices. Where there have been attempts to skirt around the rules, the US Food and Drug Administration (FDA) has been quick to point out the requirements. I don’t think the FDA is going to fight against change, rather they will continue to set and monitor appropriate safety and performance standards. In the case of Holomic’s platform, which is actually a diagnostic reader for many kinds of tests, the approval process for the cell phone-based system as a whole will be much shorter because similar (non-cell phone-based) systems already exist, which allows us to go down the 510(k) route of proving equivalent performance to a validated bench-top instrument.

Getting on board

The automation of signal reading is a no-brainer; it makes tests more robust, improving accuracy, sensitivity and repeatability. The cell phone provides everything needed for automated reading of a signal or image: an advanced camera for imaging, powerful processing capabilities for computational tasks, and a high-resolution screen to display data, all within a compact package. Even though not every application will make full use of all these abilities of the cell phone, anyone interested in developing field-portable devices who fails to utilize these advantages will quickly fall behind. Or at the very least they will find it extremely costly to improve specifications at the same rate as cell phone technology, which is simply not sustainable in the long run: how can a small biotech company compete with Samsung or Apple on those terms? Instead of competing with emerging consumer devices, we must accept them and leverage their power for our own applications.

Using the power of consumer electronics to bring the advanced functions normally found in a hospital or laboratory into field settings empowers applications in a whole range of areas, from environmental monitoring to material science to health care in developing countries. It also helps build research capacity in developing countries. Insufficient infrastructure and/or funding can make it impossible to buy and/or maintain expensive laboratory instrumentation or perform some research; however the innovation landscape generated by the coupling of consumer electronics with diagnostic tools changes the dynamic. Through democratization of measurement toolsets using mobile phones and other ubiquitous and cost-effective devices and interfaces, researchers in developing countries will be capable of generating high-quality scientific output, matching that of their colleagues in developed countries. Not only that; mobile analytics will also have a big impact on the democratization of science in general. Right now, the research world is highly polarized in terms of output: there is a close correlation between a country’s GDP and the number of papers published.

In education, the same holds true. But the recycling of cell phones or their components to make innovative, high-end analytical devices will boost science and engineering education. Hands-on education experience is very important, especially for science, technology and engineering fields; it enables skills in solving problems, the testing of hypotheses, and prompts students to ask the right questions. In developing countries, where even basic instrumentation is lacking, education suffers. And, in fact, it’s unlikely, even in developed countries, that we would happily use a $50,000–$100,000 microscope to show a kid what a HIV virus looks like; however, now we can use a phone that costs less than $500 to do the same thing. That’s a game changer.

The term “citizen science” is a little fuzzy – but it certainly hints at another facet of the current direction of innovation. Acquisition of high-quality data from large numbers of cell phones or other consumer electronics devices all over the world will enable us to discover patterns and trends that would be impossible to find otherwise.

To conclude, various benefits of mobile phone-based diagnostics, for example, improved implementation of health care and more widespread environmental monitoring, are immediately obvious. The slow-burning transformation in the behavior of researchers and educators in resource poor countries is less obvious – but it too is almost inevitable.