Aaron Morris

My coworkers at the University of Michigan and I decided to do just that.

Instead of taking an outside-in approach to diagnostics, we're taking an inside-out approach.

We are creating implantable sites that have similarities to other sites in the body and will improve our vision by giving us real-time access to molecular and cellular information about disease tissues.

These insights will enable us to predict the onset of disease and even identify therapies likely to work in an individual patient.

So, what does this inside-out approach look like?

Step one is to engineer new tissues just under the skin.

These tissues have similarities to other inaccessible sites in the body, like the brain or the lungs.

By implanting a porous plastic disc made of FDA-approved biomaterials, I can harness the body's natural responses to allow cells to migrate into the disc, survive at the site, and form a tissue.

Eventually, we're left with an engineered tissue with integrated immune cells, just the cells we need for diagnosis.

Although these tissues are complex and chronically inflamed, they're also innocuous, and after a few weeks, nearly imperceptible.

Our engineered tissues contain information not present in the blood, and they can help bridge the gap between what we can see on a traditional test and cellular changes we know occur in disease.

Step two is to read this signal.

Currently, I could take a biopsy of my engineered site and analyze it because I made them accessible just under the skin.

But it would certainly be better if we could incorporate and read a sensor non-invasively.

Within the next decade, rapidly converging technologies could enable diagnosis at such an implant by harnessing simple detectors like a blood pressure cuff or smartwatch does now.

The mechanisms for diagnosing and monitoring disease could be as simple as opening an app like Candy Crush on your phone.

Step three is to harness the huge array of knowledge in fields like engineering and materials science to improve these implants and our ability to read their data.

Eventually, tens if not hundreds of individual engineered tissues with integrated sensors may be implantable with a single application.

Now, this approach to diagnosis is unconventional to be sure.

But it is robust.