The technological advances of recent times make us wonder whether a question such as –Hey Siri, do I have Parkinson’s– will ever be a real alternative available to the public to diagnose that or other complex neurological diseases. We can certainly respond –not in the short term. But evidently we are witnessing a large expansion of novel digital technologies to evaluate people affected by movement disorders. The reasons for this are manifold: the miniaturization of sensors, the power of artificial intelligence and what is known as machine learning, and importantly, the need for developing more objective evaluation methods. All these factors have led to the everyday growth of new tools/resources in the news and academic journals. To evaluate the real possibility to reach the point of effective application of these technologies to replace the specialized professional diagnoses of Parkinson’s or other movement disorders, we have asked two of our society’s renowned experts and leaders of the MDS Technology Taskforce to share their views.
Q1: Will my mobile phone be able to diagnose if I have a movement disorder and why/why not?
Dr. Espay: We are a long way to using technology to include or exclude a "diagnosis" of the conditions we most often care for. The phenotypic variability of each condition invariably creates an inherent source of error for mobile health technologies, which can capture only a narrow slice of behavior bringing, inevitably, false positives and false negatives diagnoses. Also, the concept of a diagnosis for neurodegenerative disorders is due to evolve. "Parkinson's disease", "Alzheimer's disease", and other disorders lump many molecular disease subtypes. We will want more sophistication if we are to fully embrace the era of precision medicine.
Dr. Maetzler: I am optimistic that mobile phones will be able to do that in due course. They will not only diagnose movement disorders, they will most probably do it even earlier as the doctor can do it to date. There is already some literature available indicating that subtle motor changes exist in the prodromal phase of, e.g. Parkinson’s disease (1-3). It is well imaginable that (a) these subtle motor changes are detectable with high-quality motion sensors and specific algorithms and (b) that these sensors and algorithms can be integrated at some point into mobile phones. The quality of sensor technique that is integrated in currently available mobile phones may already be sufficient to do the trick.
Q2: What are the main opportunities around the use of mobile health technology in movement disorders?
Dr. Espay: The main opportunities are in monitoring a range of behaviors of interest over time to obtain an ecologically accurate picture of the fluctuations in a day and over longer periods of time. At the Task Force on Technology we are working toward planning the creation of a platform capable of integrating and synchronizing data from devices with different proprietary systems, coming together into "channels", similar to those of an electroencephalograph, in which data from one or two channels may provide valuable information about an individual's function over time –but irrelevant to another, who may be best served by data from other channels.
Dr. Maetzler: In my view the most important opportunity that comes with mobile health technology is that patients can do self-assessments, can learn more about their disease and what is good and bad for the actual state and (non-) progression of the disease, and can integrate this information into their own (I mean that literally, i.e., patient-controlled) electronical health record that will soon be available in most countries. I see this development in the broader picture of the future of digital medicine: Data collection for diagnosis and management of diseases –including movement disorders- will be based on interoperable electronical databases, with the patient-controlled electronical health record in the center (4). As illustrated in Figure 1 (adapted from www.neurologie.uni-kiel.de/en/neurogeriatrics/research), patients will use passive and active data collection devices –including motion sensors- and feed all information into the electronical health record. Additional databases, such as clinical data warehouses (institutional-located databases) and population health analytics will also feed into the health record and support health-related decision making by the professional team.
Q3: And what are the open questions around their use?
Dr. Espay: The unresolved questions once technology integration is adequate will deal with data ownership and data sharing. Limits to data sharing for the purposes of security and privacy protection, for instance, may also limit the usefulness of what these technologies can bring.
Dr. Maetzler: Although the above-mentioned aspects draw an optimistic picture of the future of mobile health technology, many questions have to be answered before this resource will be implemented in clinical routine. First, we do actually not have algorithms that detect human movements with acceptable accuracy in the home environment. For example, detection of movements that are not performed with own muscles (e.g., public transport) are difficult to differentiate form self-performed movements based on actually available algorithms. Second, data protection issues are not sufficiently solved, although enormous progress has been made in recent years. Third, medical professional teams, policy makers and especially patients have to get adapted to this rapid and, at least partly, game-changing development. However, the future of digital medicine is, in my view, extremely bright especially for patients, with better treatment options due to, for example, continuous and highly granular assessment of mobility.
Final Remarks
From these answers by two experts, we can see that there are many opportunities around the use of mobile health technologies to aid in the diagnosis of movement disorders. These technologies have opened up the possibility to perform evaluations outside the clinic and in the natural environment of the patient. These digital health technologies also seem to perform with higher precision and objectivity than traditional means. On the other hand, there are many aspects that will need to be addressed before these technologies are fully embraced, namely how the information generated by most mobile sensors is evaluated, stored or transmitted.
So the corollary is,
Q: Hey Siri, do I have Parkinson’s?
A (Siri): “Thanks for asking, but I am not a doctor (not yet, I am still in school). You should get an appointment with a movement disorders specialist”.
This is where we are.
References
1. Mirelman, A., et al. Gait alterations in healthy carriers of the LRRK2 G2019S mutation. Ann Neurol 2011; 69:193-197.
2. Postuma, R. B.,et al.; How does parkinsonism start? Prodromal parkinsonism motor changes in idiopathic REM sleep behaviour disorder. Brain 2012; 135:1860–1870
3. Schrag, A., et al. Prediagnostic presentations of Parkinson's disease in primary care: a case-control study. The Lancet Neurology 2015;14: 57 – 64
4. Mandl KD, Kohane IS. Time for a Patient-Driven Health Information Economy? N Engl J Med 2016;374:205–8