Promoting a better quality of life
Several brain disorders present with similar symptoms; therefore, obtaining an accurate diagnosis can take a while, and sometimes a definitive diagnosis is even not possible. By the time a patient is concerned about a symptom and goes or is taken to a doctor’s office, dementia may have been in development, possibly presenting as a minor symptom, for up to 7-8 years. Medication for dementia may slow disease progression and longer maintain the patient’s quality of life. A waiting time of 20 months for the diagnosis is essentially a waste of time because many actions are not undertaken until dementia is diagnosed.
Results of European R&D co-operation
VTT Technical Research Centre of Finland (VTT) participated in three EU-funded projects during 2009-2018: PredictAD, VPH-DARE@IT and PredictND, of which in two of them it was coordinator. These projects focused on developing decision support tools for dementias using artificial intelligence and machine learning. In the final project the accent was on validating the developed methods in four European hospitals involved in the project.
The PredictND project ended in early 2018, and an application tool to facilitate decision making for the early diagnosis of dementia is now available as part of the product portfolio of VTT’s spin-off company, Combinostics Oy.
Cost savings with artificial intelligence and machine learning
In order to apply artificial intelligence and machine learning tools in a meaningful way, we need large databases that combine patient data. Currently, hospital databases often only allow the use of the hospital’s own medical records, instead of sharing patient data across hospital networks. The optimum use of artificial intelligence in medical research and diagnostics also requires further development of data protection regulations.
It is important to note that artificial intelligence does not make an actual diagnosis; rather, it produces risk assessments and thus supports decision-making by doctors, that is, it is a tool for humans. Significant cost savings will be possible in the near future when routine tasks are transferred to services based on artificial intelligence. For example, when a patient visits the first-aid department, the reception bureau asks certain routine questions and determines whether a nurse or physician is needed. An artificial intelligence-based system may help in here: it bases its suggestion for a decision on the patient’s description of a symptom, which it compares to information found in the database.
Measurements on the home front
As a continuation of the PredictND project, research concentrates on developing methods to identify increased risks of developing dementia using measurements done at home during daily living that can be combined with the more traditional measurements done in the hospital setting.
The application is e.g. a browser-based memory game that can detect possible cognitive deterioration based on game results from a regular player. The free application would bring cost savings: expensive cognitive tests and magnetic resonance imaging could be targeted for high-risk individuals based on game results – thus providing the right tests for the right people.
Finnish expertise on a single platform
Finland has strong know-how and expertise in high-level computing and algorithmic capabilities. The establishment of e.g. the Neurocenter Finland, and biobanks that enable the linking and sharing of researcher knowledge are incredibly valuable; together, we can contribute substantially to research and development. It is foreseen that these initiatives will bring our research expertise and resources as neuroscientists together. And then, great things can happen on the international stage for neuroscience.