During the year, the Neuroimaging Network organises meetings for its members (also at the annual ECNP Congress).
Data sharing agreement and database
- Accessible data repository project
This project involves collaboration across the different members of the ECNP Neuroimaging Network with the goal to produce an accessible data repository for analysis efforts. The collected detailed information from the participating sites will be used to create a distributed data resource for MRI-based biomarker extraction and validation that builds on safe, temporary data sharing (ViPAR platform) and distributed data analyses techniques.
We encourage ECNP members to participate in this exciting sharing data project and be part of the consortium data analysis.
For further information on how to get involved please send an e-mail to Nikolaos Koutsouleris and Adriana Herrera.
- Clinical data analysis report
Clinical data has been successfully collected and uploaded in ViPAR from 6 sites. Data from approximately 4.200 patients of the participating sites have been collected using different batteries including the GAF, HDRS, MADRS, CTQ and PANSS scales, as it is described below:
- Analysis proposals
A Standard Operating Procedure (SOP) was created to submit Analysis proposals to the ECNP Neuroimaging Network Accessible Data Repository Project Steering Committee.
A query home page was also created where the research data is available. In this platform the sites could fill in an analysis proposal form, selecting the data that wants to be analysed and generating a proposal document that will be sent to the Steering Committee.
The analysis proposal tool and query can be found here.
- ECNP site control pilot study — September 2020
A central problem across multiple funded research projects is MRI site control. As part of an ECNP-affiliated pilot study, we systematically investigated multiple site control techniques in a large cohort of publicly available cases (N=XX) within a machine learning competition. Results demonstrated the potential of competitions for science generation and collaboration. This is similar to the gains found for ADHD-200 and ABIDE datasets that have generated >30 papers each. The winning team used a combination of brain atlases with a mean correction technique (Fig. 1). The Follow up results showed that the competition indicated a underlying function that links dimensionality with reliability (Fig. 2). These pilot results will be expanded within the ECNP neuroimaging network and funding will be secured to continue the work.
- Online Machine Learning School, 19-23 September 2022
This virtual school was for clinicians, PHDs and early career researchers and it was suitable for different level of machine learning experience.
25 ECNP members could register for free.
To the programme
- NeuroMiner Machine Learning School September 2020
The aim of the NeuroMiner Machine Learning School was to introduce the participant to psychiatric machine learning with a specific focus on diagnostic and prognostic neuroimaging. NeuroMiner is a machine learning software written by Prof. Nikolaos Koutsouleris in MATLAB that can be operated with little coding experience with a text-based menu system. This gives clinical researchers access to cutting-edge tools that can be used for predictive medicine and has been used for multiple publications. For the second time in September 2020, the NeuroMiner Machine Learning School was conducted by the ECNP Neuroimaging Network, the Section for Neurodiagnostic Applications, and the Mx Planck Fellow Group for Precision Psychiatry. A total of 124 participants were registered in the 4 day course and 50 participants attended specific tutorials conducted by 5 instructors. It was a successful event with excellent feedback from the participants.
1. Accessible data repository project query and analysis proposal
2. NeuroMiner model library
||Journal and link to article
|Dwyer et al.
||Clinical, brain, and multilevel clustering in early psychosis and affective stages
|Pigoni et al.
||Classification of first-episode psychosis using cortical thickness: a large multicenter MRI study
|Haidl et al.
||The non-specific nature of mental health and structural brain outcomes following childhood trauma
|Koutsouleris et al.
||Multimodal machine learning workflows for prediction of psychosis in patients with clinical high-risk syndromes and recent-onset depression.
|Dwyer et al.
||An investigation of psychosis subgroups with prognostic validation and exploration of genetic underpinnings: the psycourse study.
|Popovic et al.
||Traces of trauma: a multivariate pattern analysis of childhood trauma, brain structure, and clinical phenotypes.
|Sanfelici et al.
||Individualized diagnostic and prognostic models for patients with psychosis risk syndromes: a meta-analytic view on the state of the art.