The Write Diagnosis

Blood tests, tissue biopsies, X-rays, and ultrasounds — these are the sorts of things that come to mind when one thinks about medical diagnostics. But engineers at UCLA have a different idea about what these tools could look like in the future. They have developed a new diagnostic device that takes the form of a pen. No, not with a hidden needle inside like an EpiPen, but the sort of pen that one writes with. And early indications are that it can detect Parkinson’s disease sooner than other existing tests.

It has been estimated that tens of thousands of people develop Parkinson’s disease each year that goes undiagnosed. In the later stages this disease is highly debilitating, but the underlying neurological damage causing those serious symptoms takes time to accumulate. If earlier medical interventions could be initiated, the worst of the symptoms could potentially be delayed, giving impacted individuals better health for longer. But for this to happen, patients must be diagnosed sooner.

That is where the researchers’ new tool comes in. Existing diagnostic tests for Parkinson’s disease are costly and complex, requiring visits to specialists. As such, few people that do not suspect they have a problem would seek them out. But the pen created at UCLA is inexpensive, simple to use, and could be distributed widely. There are few barriers that would prevent it from being a regular part of a physical checkup.

Contained within the 3D-printed pen is a silicon magnetoelastic tip and a ferrofluid ink with tiny embedded magnetic particles. In conjunction with a thread of conductive yarn that is wrapped around the length of the pen, very precise movements can be detected. This is achieved by measuring how magnetic fields around the pen are changed as the user writes with it or traces it through the air.

Given the complexity of these signals, the team leveraged a neural network to interpret them. The network was trained to recognize the characteristic motor patterns that are associated with the disease. That gave it the ability to discriminate between healthy individuals, and those with Parkinson's disease with an average accuracy of 96.22%. It should be noted, however, that these early trials only involved sixteen participants, three of which had the disease, so further validation of the approach will be needed before this pen can be used in the real world.

Should the system hold up to further validation, the team hopes to extend its use in the future. They anticipate that in addition to Parkinson’s disease, the pen may also be capable of detecting other neurological conditions. With any luck, these pens will be available in physicians’ offices around the world one day.