Data analysis tools to help in early diagnosis of dementia
As part of SalWe´s Mind and Body Programme, novel methods to detect and diagnose Alzheimer´s disease at an early phase have been developed. The results indicate promising prospects and the methods are already available for research purposes.
Dementia is a growing health problem with enormous costs to society: the costs currently equal one percent of the global gross-domestic product. Dementia cases are expected to have a three-fold increase to 115 million cases by 2050.
Unfortunately, for dementias there is no cure yet, but intensive research is carried out to find therapies delaying its onset or slowing down its progression. The therapies that allow more patients to remain in milder phases of the disease should be started as early as possible to be effective. This requires tools to detect and diagnose persons at an early phase, and interventions for delaying disease progression.
Tools combine patient data from different sources
VTT Technical Research Centre of Finland has developed tools for the early diagnosis and follow-up of Alzheimer’s disease (AD) that is the most common form of dementia. The tools combine patient data from different sources, such as cognitive tests, magnetic resonance imaging (MRI) images or cerebrospinal fluid samples. The method calculates an index that describes whether the patient’s data resemble data recorded from healthy people or AD patients.
The further development of the tools was carried out in SalWe´s Mind and Body Programme.
“Within Salwe, we developed methods to analyse and visualize time-related changes in the index values and thus help detect subtle increases that could be linked to early development of AD,” states Docent Mark van Gils, Research Team Leader at VTT.
Helping to detect patients who can benefit from interventions
The method was successfully applied to data collected from patients with mild cognitive impairment (MCI) in 50 sites in the United States and Canada (the Alzheimer´s disease neuroimaging initiative, ADNI data set).
“The results show that the index values behaved differently over time between MCI patients who developed AD and MCI patients who did not develop AD. The index values of Alzheimer´s disease converters increased on average five times faster than theirs with no AD, meaning that their disease progressed considerably faster. Thus, this method for estimations of changes in variables potentially allows us to detect persons who would benefit from interventions earlier than currently possible,” states Mark van Gils.
“The mathematical methods are application-independent and can thus also be used for other data analysis problems. We have implemented our methods into a MATLAB-toolbox that is freely available for research purposes.”
According to van Gils, the Salwe cooperation has given the unique opportunity to collaborate with state-of-the-art partners in this field, thus leading to results based on valid health and wellbeing expertise especially at Finnish Institute for Occupational Health and Jorvi Hospital and taking into account the business view, GE Healthcare being the industrial partner.
Mark van Gils, Ph.D.
Research Team Leader
+358 20 722 3342
Runtti H, Mattila J, van Gils M, Koikkalainen J, Soininen H, Lötjönen J. Quantitative evaluation of disease progression in a longitudinal mild cognitive impairment cohort. Journal of Alzheimer’s Disease. 2014: 39: 49–61.
Cluitmans L, Mattila J, Runtti H, van Gils M, Lötjönen J. A MATLAB Toolbox for Classification and Visualization of Heterogenous Multi-Scale Human Data Using the Disease State Fingerprint Method. Proc. of pHealth 2013, 2013.
Alzheimer’s Disease Neuroimaging Initiative, http://adni.loni.ucla.edu, accessed on 10 November 2014.
You can find more SalWe´s stories here http://www.salwe.fi/home-en.html#blog