Open Access Publication of the Month – Dr. Kaustubh Patil (INM-7), Julian Caspers (UKD) et al.
13. April 2023
To encourage us all to broaden our horizons, each month the Central Library selects one open access publication from the JuSER publications portal to be featured in the newsletter of Forschungszentrum Jülich.
In this month’s open access publication, Shammi More, Georgios Antonopoulos, Felix Hoffstaedter, Simon B. Eickhoff, Kaustubh R. Patil (all Institute of Neuroscience and Medicine – Brain and Behaviour, INM-7) and Julian Caspers (University Hospital Düsseldorf, UKD) report on a performance comparison of 128 combinations of machine learning workflows that are used to predict a person’s age based on their magnetic resonance imaging scan.
Magnetic resonance imaging (MRI) is a powerful tool for studying brain changes due to aging. However, interpreting the data can be difficult. Machine learning (ML) can help by analysing large amounts of MRI data to identify patterns in how the brain changes with age. By training an ML model to predict a person’s age based on their MRI scan, scientists can detect abnormal changes that might be associated with certain diseases.
To create an accurate ML model, scientists must consider factors such as the ability to make accurate predictions on data from different MRI machines as well as the issue of reliability over time, i.e. it should be able to predict age consistently – even as the brain changes with age. The authors evaluated 128 combinations of ML workflows and found a well-functioning workflow, demonstrating its usefulness using an Alzheimer’s dataset. They plan to use this workflow in other scenarios, such as studying schizophrenia, sleep deprivation, and brain changes associated with space travel.
The open access publication entitled “Brain-age prediction: A systematic comparison of machine learning workflows” was published in the journal NeuroImage.
JuSER publications portal – Brain-age prediction: A systematic comparison of machine learning workflows
Internet – Institute of Neurosciences and Medicine – Brain and Behaviour (INM-7)