Outcome

The event was a success with close to 75 participants (project instigators included) and 8 projects prototypes. You can see below a slideshow of groups hard at work. The results of each project (when available) can be accessed simply by clicking on the associated GitHub icon in the projects section. Courtesy of all the participants supporting Open Science !
PROJECTS

Team

We are here to help. Don't hesitate to ask us any question.

Michael Dayan

Fondation Campus Biotech Geneva

Ellie Shamshiri

Geneva University Hospitals (HUG)

Sebastien Tourbier

University Hospital of Lausanne (CHUV)

Maria Rubega

University of Geneva

Philipp Koch

EPFL

Anne-Dominique Lodeho

Fondation Campus Biotech Geneva

Speakers

Participating to provide examples of cutting-edge neuroimaging research.



"Machine learning for clinical brain imaging: promises, challenges, and the slow but steady march towards ML-augmented radiology" by Jonas Richiardi

Abstract: Computer vision and machine learning have seen tremendous increases in performance on natural images in the past 10 years. In medical imaging at large, there has also been impressive achievements, but 3D and 4D imaging data commonly found in brain imaging is slightly more tricky to handle, in particular due to image sizes. Site hardware differences, small dataset sizes, low data quality, clinical heterogeneity general cost of data acquisition are further difficulties in using clinical brain imaging data for predictive modelling. Thus, despite Geoff Hinton's famous 2016 quip that 'people should stop training radiologists now', much work remains.

In this talk, I will highlight some of the exciting ongoing work at the Lausanne University Hospital and worldwide, focusing on efforts to automate diagnosis, differential diagnosis, and prognosis using radiology data. I will also highlight practical and theoretical challenges, and discuss how human radiologists will increasingly collaborate with ML tools in their daily routine.

" How to optimize our memory consolidation: effects of sleep, reward and sport" by Kinga Igloi

Abstract: We all hope to stay sharp across our lifetime and defy the rules of time and aging by keeping good memories throughout our life. In this talk I will combine research from animal and human background to show how sleep, reward and sport are three key factors to boost our memories and underlying neural circuits at all ages.

"The lesion method as a tool for the falsification of hypotheses derived from functional neuroimaging studies" by David Rudrauff

Abstract: We will introduce the human lesion method and discuss its interest and limitations for the falsification of hypotheses derived from functional neuroimaging studies. We will illustrate the issue around the mainstream belief in the critical role of the insular cortex in interoceptive awareness, the experience of pain and consciousness.

Projects

Virtual reality to improve brain lesions drawing

by Louis Albert, Giulia Bommarito and Michael Dayan

Can brain dynamics predict IQ?

by Raphael Liegois and Thomas Bolton

Novel neuronavigation to guide brain stimulation

by Takuya Morishita, Philipp Koch and Olivier Reynaud

EEG and fMRI connectivity for predicting epilepsy surgery

by Ellie Shamshiri, Margherita Carboni and Maria Rubega

Machine learning for dementia clinical forecasting

by Mazen Mahdi and Jonas Richiardi

PlasticBrain: real-time brain activity on a 3D printed brain

by Manik Bhattacharjee and Victor Ferat



You can even propose your own by filling the form here!

Project 1. Virtual reality to improve brain lesions drawing by Giulia Bommarito, Louis Albert and Michael Dayan

Multiple sclerosis (MS) is a disease characterized by lesions in the brain, particularly visible on MRI data. The gold standard to mark these lesions still consists in looking at MRI slices one by one and manually draw/color the lesions, leading to unnatural shapes when seen in 3D. This is concerning as the resulting "lesion masks" are often used to characterize the disease, including in drug clinical trials. This project aims at creating a virtual reality tool to draw MS lesions directly in 3D, for a chance to possibly improving the current gold standard in MS lesion segmentation!

Project 2. Can brain dynamics predict IQ? by Raphael Liegois and Thomas Bolton

The way cerebral function shapes human behaviour remains largely unknown. In this project, we propose to explore if brain dynamics, evaluated from functional MRI measurements, can be used to predict a simple measure of fluid intelligence.

Project 3. Package an application for super resolution MRI by Sebastien Tourbier

The essence of this project is to develop the next generation of the open-source Medical Image Analysis Laboratory Super-Resolution ToolKit (MIALSRTK), a set of C++ image processing tools necessary to perform motion-robust super-resolution fetal MRI reconstruction. This new version will integrate all new advances in the neuroimaging field over the past three years with the advent of:

  • the Brain Imaging Data Structure (BIDS), a standard to organize and describe neuroimaging data,
  • and the BIDS App framework, which promotes reproducibility and portability. all state-of-the-art solutions for enhanced portability, reusability, reproducibility and replicability in neuroimaging.

Project 4. A novel approach to optimize Brain Stimulation: structural connectivity guided individual neuronavigation by Takuya Morishita, Philipp Koch and Olivier Reynaud

Transcranial magnetic stimulation (TMS) is used to study electrophysiological fundaments of brain function. Besides this TMS has the outstanding opportunity of non-invasive neuromodulation of the cortex, a promising treatment concept for multiple diseases like e.g., stroke or depression. Hereby, specific brain areas are targeted using T1 weighted MR images of the subject guided by anatomical landmarks. Still, this does not consider individual underlying anatomical and functional features including e.g., white matter connections. To move towards personalized precision medicine in neuromodulation, incorporating individual patient characteristics in treatment planning is inevitable. Taken toegther, this project aims to precise the targeting of brain areas for TMS using individual structural connectivity estimates.

Project 5. EEG and fMRI connectivity for predicting epilepsy surgery using advanced signal processing for connectivity visualization by Ellie Shamshiri, Margherita Carboniand Maria Rubega

The goal of the project is two-fold: - to manipulate and display brain connectivity data (resolved in space, time and frequency) of epilepsy patients and normal controls acquired with electroencephalography (EEG) and functional MRI - to leverage these data to predict surgery outcome (success vs failure) of patients suffering from intractable epilepsy

Project 6. Machine learning for dementia clinical forecasting from imaging and multimodal data by Mazen Mahdi and Jonas Richiardi

The goal of this project is to rapidly develop and test algorithms to forecast disease evolution in Alzheimer's disease, using pre-extracted brain imaging markers from MRI (volume, thickness...), PET (metabolism, amyloid load...), Diffusion (diffusivity), liquid biomarkers, and clinical scores, provided by the TADPOLE challenge project.

Project 7. PlasticBrain: real-time brain activity displayed on a 3D printed brain by Manik Bhattacharjee and Victor Ferat

PlasticBrain is a 3D-printed brain built from an MRI. During the hackathon we would like to display realtime brain activity as recorded with an EEG headset onto it (e.g. power in alpha/beta/gamma frequency bands). The goal is to visualize functional activity so that it can be used in a biofeedback context for research and clinical applications, and for teaching.

Project 8. Misfolded protein spreading on brain connectome by Alessandro Crimi

Connectomics have been used so far to look for quantifying global and local differences in the functional or structural brain networks [1], or alternatively, to simulate/study brain hemodynamics [2] . Very few studies have used connectomes to investigate the spreading of misfolded proteins which is at the basis of Parkinson’s (PD) and Alzheimer’s disease (AD) [3,4]. It is believed that diseases as AD and PD are spread by misfolded proteins or agents which moves along brain connections (axons and dendrides of the neurons) starting from specific regions to others [5]. For instance, AD has a progression of tau pathology consistently beginning in the entorhinal cortex, the locus coeruleus, and other nearby noradrenergic brainstem nuclei, before spreading to the rest of the limbic system as well as the cacingulate and retrosplenial cortices. While Parkinsons starts from the brainstem and spread to the neocortex [5]. A previous study investigated this mechanism on connectome comparing simulated tau deposits on connectome to those detected by PET scans specific for Alzheimer’s [3].

In this project, we aim at carrying out a similar study bu t for PD. In particular, we want to simulate deposits/spreading of alpha-syn proceeding via the brain’s anatomic connectivity network. We will use human and mice data provided by the supervisors. The main challenge of the project is defining a proper model of spreading along the connectome. Data from the PPMI dataset (https://www.ppmi-info.org/access-data-specimens/) and given by the project supervisors will be given. Those include case (PD patients) and control subjects along with clinical data and genetics.

1. Griffa, Alessandra, et al. “Structural connectomics in brain diseases.” Neuroimage 80 (2013): 515-526. https://bit.ly/2Soox0S
2. Friston, Karl, Rosalyn Moran, and Anil K. Seth. “Analysing connectivity with Granger causality and dynamic causal modelling.” Current opinion in neurobiology 23.2 (2013): 172-178. https://bit.ly/2EsLPiz
3. Iturria-Medina, Yasser, et al. “Epidemic spreading model to characterize misfolded proteins propagation in aging and associated neurodegenerative disorders.” PLoS computational biology 10.11 (2014): e1003956. https://bit.ly/2E7bFav
4. Raj, Ashish, Amy Kuceyeski, and Michael Weiner. “A network diffusion model of disease progression in dementia.” Neuron73.6 (2012): 1204-1215. https://bit.ly/2GZWBhV
5. H. Braak et al., “Staging of brain pathology related to sporadic Parkinsons disease,” Neurobiology of aging, vol. 24, no. 2, pp. 197–211, 2003. https://bit.ly/2SUPFK2

Project 9. Visualizing brain connectomics using D3.js by Renaud Marquis

Current progress in neuroimaging allows to collect data from multiple imaging modalities and at multiple scales. While numerous software packages allow the processing of such data, visualization can become problematic with increasing number of dimensions. The goal of this project is to use D3.js, a JavaScript library relying on common web standards, to visualize complex brain connectivity from multiple imaging modalities (EEG, functional and diffusion MRI) at multiple scales in an interactive fashion to facilitate insights, data and knowledge sharing.

Project instigators

Here is the list of project instigators who will help giving the "big picture" of each project.

Giulia Bommarito

Giulia's expertise: neurology, neuroimaging, multiple sclerosis

Louis Albert

Louis' expertise: virtual reality, C/C#/C++, Unity

Michael Dayan

Michael's expertise: MRI image processing, brain connectivity, machine learning

Raphael Liegois

Raphael's expertise: brain connectivity, dynamical systems theory, functional MRI

Sebastien Tourbier

Seb' expertise: cmake/c++/itk, python/nipype, BIDS, BIDS App, super-resolution, brain connectivity

Philipp Koch

Philipp's expertise: neuroimaging, network analysis, medicine

Takuya Morishita

Takuya's expertise: transcranial magnetic stimulation, motor control

Olivier Reynaud

Olivier's expertise: brain imaging, MRI, brain stimulation, physics, engineering

Ellie Shamshiri

Ellie's expertise: Neuroimaging, Simultaneous EEG-fMRI, Cognitive Neuroscience, Epilepsy, Brain Connectivity

Margherita Carboni

Margherita's expertise: EEG, epilepsy

Maria Rubega

Maria's expertise: EEG, machine learning, Matlab

Mazen Mahdi

Mazen's expertise: data science, modelling, visualization, docker

Jonas Richiardi

Jonas' expertise: statistical modelling, machine learning, neuroimaging, predictive radiology, imaging genetics

Manik Bhattacharjee

Manik's expertise: neuroimaging, software engineering, signal processing, 3D printing

Victor Ferat

Victor's expertise: EEG, Signal processing, Electronics

Alessandro Crimi

Alessandro's expertise: machine learning, neuropathology, dynamical model, biomarkers

Renaud Marquis

Renaud's expertise: EEG, MRI, machine learning, connectivity

Thomas Bolton

Thomas's expertise: fMRI analysis, dynamic functional connectivity, hidden Markov models, graph signal processing, Roger Federer

Program (event concluded)

The Brainhack took place over two days, on March Friday 22nd and Saturday 23rd.

8:30
Breakfast
9:00
Welcome/Ignite talks
Breakfast
09:30
Project pitches
Ignite talk
10:00
Mingling & Coffee break (~15 mins)
Open Hacking
10:30
Open Hacking
Open Hacking
12:30
Lunch
Lunch
13:30
Open Hacking
Open Hacking
15:15
Break
Break
15:30
Open Hacking
Open Hacking
17:00
Open Hacking
Project prez
18:00
Open Hacking
Wrap up
19:00
Drinks/Dinner

Contact & Venue

Address

Chemin des Mines 9, 1202 Geneva, Switzerland

Phone Number

Email