Enhanced validation for brain research analysis methodsMagnetic resonance imaging (MRI) is a method that provides a lot of information on brain activities. MSc (Tech) Juha Pajula further developed the ISCtoolbox software that enables researchers to obtain information on brain functionalities using research designs that mimic real-life stimuli.
Functional magnetic resonance imaging (fMRI) is a modern, non-invasive method for studying human brain functions. FMRI has become a common imaging method in scientific and clinical research as it enables studying the functional areas of the brain, for example. A typical fMRI study could try to identify, for example, the brain areas that activate when the subject moves his fingers. With a different selection of analysis stimuli, it is possible to analyze basically any active brain functionality triggered by an external stimulus.
In everyday life, the human brain faces an extremely complex set of stimuli through all the five human senses. A simple, highly controlled test has little chance of effectively examining the complex brain functionalities related to real-life situations.
“The latest fMRI studies with naturalistic stimuli use such real-life impulses as movies, video games or music. Specifying a comprehensive mathematical model for an entire movie or musical piece is highly challenging, which limits the use of the typical model-based analyses in these research designs,” Juha Pajula explains.
In his dissertation, Pajula studied the ISCtoolbox calculation software, previously developed in the same research group. The software enables studying computational data from complex research designs. ISCtoolbox implements the inter-subject correlation (ISC) analysis method.
In his dissertation, Pajula examined the properties of ISC-based analysis and validated it as a reliable method for fMRI analyses. The ISC analysis is used for searching brain areas which are activated similarly between multiple subjects under the same stimulus.
“According to my research results, the ISC analysis is well suited for both typical, highly controlled analyses and modern fMRI studies using naturalistic stimuli. The ISC analysis and its user-friendly ISCtoolbox implementation create new opportunities for researchers around the globe to study the functional structure of the human brain,” Pajula says.
Juha Pajula further developed the ISCtoolbox software by enabling a cluster computing support for the software. Due to the vast computational load, effective computing is required: it may take days for an individual desktop computer to compute a single large ISC analysis.
“The largest computations in this dissertation would have taken nearly five years with a single computer, but with TUT’s Merope computing cluster, they were completed in roughly two months.”
Public defence of a doctoral dissertation on Friday, 22 April
The doctoral dissertation of MSc (Tech) Juha Pajula in the field of biomedical signal processing entitled ‘Inter-Subject Correlation Analysis for Functional Magnetic Resonance Imaging - Properties and Validation’ will be publicly examined at the Faculty of Computing and Electrical Engineering of Tampere University of Technology (TUT) in room TB109 in the Tietotalo building (address: Korkeakoulunkatu 1, Tampere, Finland) at 12:00 noon on Friday, 22 April 2016. The opponent will be Professor Christian Windischberger (Medical University of Vienna, Austria). Vice President Ulla Ruotsalainen from TUT’s Department of Signal Processing will act as Chairman.
Juha Pajula (32) comes from Tampere, Finland, and works as Research Scientist at the Systems Medicine group of VTT.