Neurological

Machine studying predicts the onset of dementia

Researchers from the Center for Healthy Aging of the Brain (CHeBA) and the School of Computer Science and Engineering at UNSW Sydney have conducted the largest comparison of survival analysis methods to date to predict the onset of dementia using machine learning.

The comparison published in Nature Scientific Reports is the first to apply these methods to CHeBA’s Sydney Memory and Aging Study, examining a wide variety of data in a study on dementia.

There is currently no cure for dementia or treatment that can successfully change the course of this disease.

Machine learning models that can predict the time to develop dementia are important tools for our understanding of dementia risks. Using data from the Sydney Memory and Aging Study, we found that we can create models that predict the onset of Alzheimer’s disease and other dementias with relatively high accuracy. Analysis of this data requires a technique known as “survival analysis” that predicts the time to an event, e.g. B. the diagnosis of a disease. We used machine learning techniques that have been adapted to deal with censored data rather than the more traditional statistical techniques.

Main author and computer scientist Annette Spooner

Recent research has shown that various sources of clinical data can provide complementary information about dementia. The integration of multiple data sources leads to a better prediction of the cognitive decline.

“Machine learning can provide more accurate results than traditional statistical methods when it comes to modeling high-dimensional, heterogeneous clinical data,” said Ms. Spooner, whose research was overseen by Professor Arcot Sowmya and supported by honorary student Emily Chen.

CHeBA co-author and co-director Professor Perminder Sachdev said the models they developed predicted dementia survival using data from the Alzheimer’s Disease Neuroimaging Initiative and the Sydney Memory and Aging Study.

Using machine learning, we found that neuropsychological scores are the best predictors of the occurrence of dementia.

Professor Perminder Sachdev, co-author and CHeBA co-director

Future research through this collaboration will aim to improve the stability whose variables are selected by the models as the most predictive of dementia.

Read the whole article

Related Articles