Parkinson’s disease was first diagnosed as far back as the early 1800s, yet there remains so much more that needs to be learned about such a tragic illness. Just some of the various points of research into Parkinson’s revolve around communication, visual symptoms, patients’ samples – and of course, the existing medical literature.

Communication

A symptom of Parkinson’s is the reduced ability to talk in a way that is loud and clear (people with Parkinson’s are often unaware of their being quiet when they talk). The startup Speechvive offers a straightforward solution: an earpiece that plays noise in its wearer’s ear.

According to PU (Purdue University – where the co-founders of Speechvive, Steve Mogensen and Jessica Huber met), the technology challenges traditional speech therapy, because such conventional approaches rely on training and , which can sadly be forgotten by those with Parkinson’s.

Having an ever-present earpiece removes this pitfall by rendering one’s control of their own speech volume an instinctive – rather than conscious and therefore forgettable – process. In fact, PU compares the process to how most people automatically speak louder when they’re in a busy restaurant.

Visual symptoms

As Parkinson’s disease affects motor functions and other health aspects such as overall fitness, this means that some symptoms can thankfully be observed visually. This does not make diagnosis an easy process, however, especially as the disease can be mistaken for other signs of declining health, such as those found in the onset of other degenerative illnesses.

The startup Lookinglass offers homeowners a ‘smart mirror’ (namely a mirror that feeds back digital information to anyone in its reflection) that observes its users’ movements and is able to apply its machine learning and computer vision processes to detect early signs of Parkinson’s disease.

To quote the company’s website: “Lookinglass observes and then connects healthcare experts to grandad in his home and ensures he receives the best possible care when he needs it.”

Patient sample analysis

In the analysis of Parkinson’s, one diagnostic tool that has proved promising is the use of biomarkers, which is the collective name for human tissues, bodily fluids, or biological molecules that provide medical professionals with signs of patients’ ill health.

One example of a biomarker in the context of Parkinson’s is a distinctive protein that appears in those who have the illness. Sadly, in the past, this was observed in those who had already passed away – namely through the use of autopsies. And although there is the plus side that biopsies (where biomarker analysis is taken from a living person) were found to be effective diagnostic tools in a Michael J. Fox Foundation study, this is a laborious process without technological intervention.

Thankfully, the spinout known as PreciseDx, which offers artificial intelligence for clinical modelling in the biopsies of cancer patients, has found success in biopsy-based approaches to Parkinson’s diagnosis.

As John Crary, MD, a neuropathologist who worked on the study explains on neurologylive.com: “Throughout the entire time [of doing biopsies] I was thinking: “There's got to be a better way to do this. It'd be amazing if we could train an AI to do it.” That's where this project came from.”

Medical literature analysis

OccamzRazor is a medical startup by Katharina Sophia Volz, PhD, who wanted to find a cure for Parkinson’s when – soon after she’d completed her PhD in Stem Cell Biology – she learned that someone she knew had been diagnosed with Parkinson’s disease.

Knowing just how complicated the diagnosis of Parkinson’s is, Volz went on to become the CEO and Founder of the tech startup, which uses machine learning to mine the enormous amount of data found in the medical literature of the degenerative illness. According to OccamzRazor’s website, the new company “strives to supercharge human scientific reasoning”.

Two major areas of machine learning that the startup uses are natural language processing (or NLP) and graph representation learning. The intention is to achieve what the company calls a ‘multidimensional knowledge map’ – therefore offering the ideal combination of human understanding and artificially intelligent reasoning.

The benefits on the horizon

The advantages of AI and its many subsets, such as machine learning and image analysis, are of course crucial to society. But what is perhaps most exciting about artificial intelligence is its ability to improve the services of the healthcare industry. And given that the ingenious startups above are largely new entries to the medical field, the benefits that are on the horizon can only be unimaginable.