News & Events
An intelligent digital service that will improve the care of Parkinson’s disease7 - April
In the coming months, Helsinki University Hospital (HUS) will pilot an intelligent digital service for patients and clinicians that will improve the care of Parkinson’s disease, optimising also the use of healthcare professionals’ resources. The aim is to transfer the follow-up of chronic disease to a normal living environment and provide decision-supporting tools to clinicians for early recognition of deterioration or progression.
On the occasion of International Parkinson’s day this month, we have interviewed Laura Mäkitie, one of the neurologists involved in the pilot implementation in HUS. Same as her, several neurologists are also having interviews with the development team, and their opinions and experiences on the test use are helping to gather relevant input for the support of the development work.
What is the current pathway for Parkinson’s disease patients at HUS?
Patients visit the outpatient clinics in 6-12 months intervals, where a multi-disciplinary team helps with the management of the disease. In between the visits, patients can be in contact with their nurse via phone or electronic channels. Information about the disease, treatments and support is available 24/7 in a digital care pathway.
Adjustments to the care and treatment plan are based on clinical evaluation by the neurologist during the visit and the history given by the patient. A wide variety of symptoms and fluctuation of them make the evaluation difficult.
Which will be the improvement through AICCELERATE?
Both doctors and nurses will have information on patients’ well-being and symptoms gathered also between the visits to the hospital. Data will be analyzed and served in an easy-to-use form as a support for clinical decision making. The treatment and care can be better designed and adjusted to meet the needs. Remote monitoring will make it possible to follow the patient also between the visits. This will help to reallocate the health care resources to time points where they bring the most value.
You plan to incorporate a remote monitoring app, right?
Yes, and apart from solutions for home monitoring, we will include a video-capturing system developed by Neuropath. It will help with detecting motor symptoms regularly in patients’ normal living environments. We hope to clarify the fluctuations of the symptoms and recognition of those symptoms that are difficult for the patients to recognize or describe. The slow progress of the disease might be better recognized with repeated measurements.
We also intend to use other manufacturer’s apps like the Cognitive app by the University of Padua to detect cognitive decline.
From the patient’s perspective, what will be the benefit of using an AI-based development? And from clinicians’ perspective?
The burden of the system for the patient is to measure and report symptoms on regular bases. The reward is better care and treatment and hopefully better quality of life. The benefit for a clinician is the analysis of a huge amount of data in a form that is usable for clinical judgement.
How do you plan to optimise the use of healthcare professionals’ knowledge with the project implementation?
We are scrutinizing and implementing the knowledge of the movement disorder specialists in AICCELERATE’s machine learning models. This way, the knowledge of these rare Parkinson-specialized neurologists will be scaled widely.
Are you putting into value the data of your Data Lake for this project?
Of course! In HUS, we can use the data of 5,000 Parkinson’s patients to develop an algorithm predicting the transit of the disease to the advanced phase and cognitive impairment. In addition to that huge amount of patient data, the beauty of the data lake is, that it includes various kinds of data. For example, we can investigate if a change in the number of contacts to the hospital or emergency visits indicates a deterioration of the disease.
Furthermore, the information in the data lake is normal clinical data which could be very different from data collected for research purposes. Prediction algorithms that would be used to support normal clinical work, should be based on data collected in a similar setting.
Does it appear to be reluctant to use Artificial Intelligence to predict Parkinson’s disease in its early stages?
Since we cannot cure or even hinder the progression of Parkinson’s disease, it is not ethical to provide predictions of deterioration years beforehand. However, patients are unsatisfied with the speed (or lack of it) of getting help, when their functionality and quality of life are declining. This is why we try to recognize and even predict this situation accurately or a couple of months in advance.
April 11th will be International Parkinson’s Disease Day. What message would you send to patients and carers reading this interview?
Just as doctors want to meet their patients, also in the future, regardless of AI or remote monitoring, we do need the involvement of Parkinson’s patients in the development of digital tools. Please, if you are asked, participate in development projects even though digitalisation or digital devices seem a little uncomfortable for you.