Methods & Models for fMRI Analysis


Would you like to obtain good knowledge of the theoretical foundations of SPM and DCM and their application to empirical fMRI data?

This course teaches state-of-the-art methods and models for fMRI data analysis. It covers all aspects of statistical parametric mapping (SPM), incl. preprocessing, the general linear model, frequentist and Bayesian inference, multiple comparison corrections, and event-related designs, and Dynamic Causal Modelling (DCM), a novel approach for identification of nonlinear neuronal systems from neurophysiological data. A particular emphasis of the course will be on methodological questions arising in the context of neuroeconomic and clinical studies.


Lectures (incl. practical sessions) take place on Tuesdays, 8:15 - 12:00, at the ETH - Building ETZ, Room E6

In this year's course, we will scan our own fMRI data in two sessions during the semester (more information will be communicated at the beginning of the semester).


The timeline might be subject to change depending on the exact dates of the scanning session.


First part - Basic analysis of fMRI data (UZH/ETH and MED):

25.09. 08:15-10:00 Foundations of functional MRI: neurophysiology and physics (Jakob Heinzle)

Intro-slides, Lecture

10:15-12:00 Setting up Matlab and SPM on own computers (Jakob Heinzle)

Tutorial slides

02.10. 08:15-10:00 Why is fMRI important for medicine? (Klaas Enno Stephan)


10:15-12:00  Basic SPM functions (Jakob Heinzle) 

Tutorial slides

09.10.  08:15-10:00 UZH/ETH: Scanning of own Experiment (Sandra Iglesias); MED: Running an SPM batch analysis (Stephan Frässle)

Tutorial slides

10:15-12:00  Introduction to Spatial preprocessing of fMRI images (Sam Harrison)


16.10.  08:15-10:00 MED: Scanning of own Experiment (Sandra Iglesias); UZH/ETH: Running an SPM batch analysis (Jakob Heinzle)

Tutorial slides

10:15-12:00 The General Linear Model for fMRI analyses (Frederike Petzschner)


23.10.   08:15-10:00 Classical (frequentist) inference and multiple comparison correction (Klaas Enno Stephan)

Classical inference lecture, Multiple comparison correction lecture

10:15-12:00: Analysis of own data (Jakob Heinzle, Sandra Iglesias)

Tutorial slides, MatlabScripts

30.10.  08:15-10:00 Experimental design and Resting State Analysis (Sandra Iglesias)

Experimental design lecture; Resting State lecture

10:15-12:00 Analysis of own data and preparation of short presentation (Jakob Heinzle, Sandra Iglesias)

06.11.  08:15-10:00 Event-related fMRI and design efficiency (Jakob Heinzle)


10:15-12:00 Short presentation of results of analysis of own data (MED)

Second part - Advanced topics (UZH/ETH):

13.11.  08:15-10:00 Group level analysis (Sandra Iglesias)


10:15-12:00 Tutorial

Tutorial slides; Code

20.11.  08:15-10:00 Noise models in fMRI and noise correction (Lars Kasper)


10:15-12:00 PhysIO


27.11.  08:15-10:00 Bayesian inference and Bayesian model selection (Klaas Enno Stephan)


10:15-12:00 Tutorial: BMA and BMS (Stefan Frässle)

Data; Tutorial slides; Code

04.12.  08:15-10:00 Computational Neuroimaging (model-based fMRI) (Andreea Diaconescu)


10:15-12:00 Tutorial

Tutorial Code

11.12.  08:15-10:00 Introduction to Dynamic Causal Modelling (Stefan Frässle)


10:15-12:00 DCM analysis


18.12. 10:00-11:30 Exam (UZH/ETH)

Exam info


Exam dates and testat:

First part (Medizin Mantelstudium): No formal exam, but in order to get the testat, you need to present your data analysis in a short presentation during the tutorial on Tuesday, 06.11.2018, 10:15-12:00, Room: ETZ E6.

Main exam (all other students): Tuesday, 18.12.2018, 10:00-11:30, Room: ETZ E6

Material for the exam: Bring with you something to write (pen), your Legi and an ID. No other material is allowed!


Sandra Iglesias:

Jakob Heinzle: