The brain How the digital revolution is transforming EU-funded brain research - June 2019 - mediaTUM
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
ISSN 2599-8293
CORDIS Results Pack on
the brain
A thematic collection of innovative EU-funded research results
June 2019
How the digital revolution
is transforming EU-funded
brain research
Research and
Innovation
Editorial
Contents How the digital revolution is
transforming EU-funded brain research
As the world continues to be remodelled by the digital revolution,
3 19 we are seeing digital technologies embed themselves firmly
into all aspects of our lives, including into the mainstream
A digital disease Shaping the ethical of neurosciences research to help unravel the complexity of
classification system fit direction of the HBP for brain disorders. In this cross-cutting CORDIS Results Pack, we
for the modern medical the public interest highlight innovative projects from across the full spectrum
era of EU-funded research programmes that are helping science
21 make sense of arguably the most complex and mysterious
5 The Virtual Brain
machine in existence – the human brain.
Neuroimaging that helps simulates the brain to
us fix our brains reveal the origins of Around 165 million Europeans are afflicted with a brain disorder and it is
disorders estimated that one in three people will suffer from a neurological and/or
7 psychiatric disorder at some point in their lives. In 2011, the global cost to
Breaking through barriers
23 European healthcare budgets was estimated to be around EUR 800 billion
per year and is set to only increase further as Europe’s population ages and
for a revolution in brain A high-performance PET
becomes more susceptible to brain disorders.
scans scanner for integration
into current MRI systems
9 Brain disorders come in many forms, from neurodegenerative diseases, with
EU-AIMS: Towards
25 some of the most well-known being Alzheimer’s, dementia, schizophrenia
and Parkinson’s disease. However, other diseases and conditions that are
personalised approaches Wearable devices to also classed as brain disorders include epilepsy, depression, stroke, migraine,
for autistic people help prevent sudden sleep disorders, pain and addiction. Many brain disorders have no cure, such
unexpected death through as Alzheimer’s and Parkinson’s. Others can be managed, but still have wide-
11 epilepsy ranging effects on the quality of life of patients and their carers.
Combining neuroscience
and computing, the
27 An EU research priority
Human Brain Project ERC research lends an Now the research focus on developing new treatment options for patients
reveals the brain’s secrets ear to the voices heard by with brain disorders has taken an inevitable digital turn – harnessing the
schizophrenia patients power of digital technologies and the increasing clout of computing as a
13 means to push forward the boundaries of brain research, cognitive
Collaborative data
29 neuroscience and brain-inspired ICT advances. Alongside basic research,
technological advances are also lending themselves to new solutions that
underpins advancing New research highlights help patients better manage their conditions, thus leading to improvements
knowledge about the brain biological similarities
in quality of life – wearable technology, for instance.
between psychiatric
15 and neurodegenerative
disorders
This is why the European Commission has been supporting brain research
through successive research and innovation Framework Programmes:
Understanding how EUR 3.1 billion in the previous programme, FP7 (2007-2014), and another
the brain’s structure
and functions generate
31 EUR 3.2 billion so far through Horizon 2020.
consciousness Novel algorithms can EU support for brain research spans across the various Commission funding
predict psychosis before mechanisms, from Horizon 2020’s research focus on health (indeed
17 it strikes one of the six specific Societal Challenges addressed through
Horizon 2020) and the industry co-supported Innovative Medicines
When virtual brain models
meet real and simulated
33 Initiative (IMI), to the European Research Council (ERC), the dedicated
SME Instrument and the Future and Emerging Technologies (FET)
robot bodies Night-time seizure
programme. A selection of projects from all of these initiatives are featured
detector improves quality
of life for epileptic together within this Results Pack.
patients and carers
Power to the projects
In total, this CORDIS Results Pack features 16 articles, with six dedicated
to the Human Brain Project (HBP), a EUR 1 billion Flagship foreseen to be
half-funded by the FET programme.
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
A digital disease
classification system fit for
the modern medical era
Classifying diseases is a practice still stuck in the past. The Innovative Medicines
Initiative AETIONOMY project brings it into the 21st century.
© Atthapon Raksthaput, Shutterstock
The classification of diseases is still largely based on patient project co-coordinator. “We are still following principles that go
symptoms, an approach that dates back to an era when doctors back way before we even knew what a ‘gene’ was.”
relied on spoken information and visual cues.
The EU and industry-funded AETIONOMY project therefore designed
“You have to be aware that our current classification system for a ‘mechanism-based taxonomy of disease’ with a particular focus
diseases has its roots in the middle of the 19th century,” says on neurodegenerative diseases: namely Alzheimer’s and Parkinson’s.
Professor Dr Martin Hofmann-Apitius from the Fraunhofer Institute The concept was based on an intriguing idea: “If we can classify
for Algorithms and Scientific Computing SCAI, the AETIONOMY patients according to their disease mechanisms, we are in the posi-
(Organising Mechanistic Knowledge about Neurodegenerative tion to predict with high accuracy whether they will benefit from
Diseases for the Improvement of Drug Development and Therapy) a drug that targets each mechanism,” says Dr Hofmann-Apitius.
3
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
A 21st century model AETIONOMY has prepared the ground for a computational
approach that drives subsequent experimental work in labs. But
AETIONOMY systematically captures and represents knowledge there is a long way to go to develop better, more efficient and
on neurodegenerative diseases in a computable format, as a cheaper drugs based on the insights and principles worked out
graph model that represents causes and effects and that can be during the project.
analysed using algorithms. It was designed to take into account
molecular biomarkers (e.g. proteins), recognising the role of per- The project also developed the concept of ‘Virtual Patient Cohorts’:
sonalised genomes with their individual genetic variations and synthetic data sets that give a new way of sharing patient-level
including various other underlying fea- data without compromising the privacy of any real person on
tures that have only been feasible to Earth. The team also gained new insights into misfolded proteins
measure in the past few years: using as a potential new mechanism for the ‘spreading’ of disease in
Our current neuro-imaging, for example. the human brain.
classification system
“The real challenge, however, in this field The AETIONOMY concept was quite revolutionary and a bit unusual
for diseases has its
is: nobody knows what mechanism(s) to many experimental clinical researchers and biologists, explains
roots in the middle of are responsible for these neurodegen- Dr Hofmann-Apitius, and there were several competing ideas about
the 19th century. erative diseases,” says Dr Hofmann- how a mechanism-based taxonomy could and should be constructed.
Apitius. “This is why they are called ‘idi-
opathic’, to indicate that their aetiology [the cause or underlying “At some point, we had to change the strategy, but finally the
causes of a disease or condition] is unknown.” project – and that includes the entire consortium with all its part-
ners – delivered on the promise,” he says.
The challenge was substantial, explains Dr Hofmann-Apitius,
especially because the pharmaceutical industry partners who
initiated the call wanted the new taxonomy to be validated in
a clinical study, meaning demonstrating its stratifying and clas- P R OJ E C T
sification potential. AETIONOMY – Organising Mechanistic Knowledge
about Neurodegenerative Diseases for the
As Dr Hofmann-Apitius says, the project’s goal was to explore Improvement of Drug Development and Therapy
if it was possible to identify patient-subgroups in Alzheimer’s or
Parkinson’s disease based on disease mechanisms. “The good C O O R D I N AT E D B Y
news is: we could demonstrate exactly that,” he reveals. UCB Biopharma SPRL in Belgium
FUNDED UNDER
A wider response FP7-HEALTH-IMI1
Because of the project, scientists are stepping back and starting to C O R D I S FA C T S H E E T
re-think neurodegeneration again, says Dr Hofmann-Apitius. “I am cordis.europa.eu/project/id/115568
convinced: AETIONOMY has paved the way for future strategies
in drug discovery and drug development in neurodegeneration P R OJ E C T W E B S I T E
research.” aetionomy.eu
4
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Neuroimaging that helps
us fix our brains
In most people’s minds, the purpose of medical imaging technologies is to diagnose diseases
or detect them before the first symptoms occur. The BRAINTRAIN project takes another
course by using functional magnetic resonance imaging (fMRI) to help patients with
neurodegenerative diseases and mental illnesses in regulating their own brain.
The BRAINTRAIN (Integrative neuroscience school on brain function the only techniques allowing us access to the function of the human
and disease) concept is built around that of neurofeedback train- brain in vivo. It is therefore a unique window into the human mind.
ing: patients can follow their brain’s activity on a screen, test spe-
cific strategies to regulate it, and see the results for themselves. This window can be used to improve our understanding of the
neural mechanisms behind psychiatric symptoms, for example
Professor David Linden, project coordinator and Head of Cardiff hallucinations or drug craving, as well as to evaluate the neural
University’s neuropsychiatric imaging group, believes that this effects of (psychological or pharmacological) therapies. Finally, and
solution could help patients affected with the likes of alcohol most importantly in the context of BRAINTRAIN, they can be used
dependency, autism, post-traumatic stress disorder or even for new treatment methods, especially neurofeedback.
Parkinson’s disease.
What would you say makes such use of
What makes you think that neuroimaging neuroimaging innovative?
can be a key weapon in the fight against
mental disorders? BRAINTRAIN brings together leading experts in real-time fMRI – a
functional imaging technique that allows for tracking the activity
Neuroimaging and non-invasive neurophysiology – electro- of specific brain regions and networks at high (millimetre) spatial
encephalography, magnetoencephalography, etc. – are presently resolution and temporal resolution in the second range.
© GrAl, Shutterstock
5
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Prof. Nikolaus Weiskopf from the Max Planck Institute of Cogni- and childhood anxiety. In other projects, we have also piloted
tive Neuroscience in Leipzig is an expert in ultrafast MRI, while neurofeedback for neurorehabilitation (Parkinson’s disease).
Prof. Rainer Goebel from Braininnovation is an expert in real-time
MRI data analysis. By combining these methods, we can acquire How do you plan to bring this to patients?
and analyse fMRI data in the sub-second range and display the
signal back to the participant, which is the basis of self-regulation Because of the non-invasive nature of neurofeedback, there are
training. The idea is that patients can use these signals, following few obstacles to bringing this technique to patients once efficacy
specific protocols, to change their brain activity, which in turn can has been demonstrated. Of course, the availability and cost of MRI
have a positive impact on their mental health. This would be the facilities constitute a constraint, but most of the protocols that
first therapeutic application of neuroimaging. we test only use a small number of sessions (1-6), which keeps
costs at a reasonable level.
Can you tell us more about the process
involved in the treatment of a patient? What has been the feedback from
practitioners so far?
Patients are in the MRI scanner and receive feedback about their
brain activity through the projection of a computer screen (it could Practitioners were grateful for the opportunity to mention our tri-
also be auditory feedback). This computer screen could just visu- als to their patients. There is generally a dearth of clinical trials in
alise the height of activation with a thermometer or similar dis- psychiatry, and most patients are keen to explore new non-invasive
play, or the feedback could be provided through the change of a treatment avenues, especially for disorders that are otherwise
disease-relevant stimulus or scene, for example a food or alcohol difficult to treat (e.g. alcohol dependence or autism).
cue as in a BRAINTRAIN study we’ve done on food craving.
Because neurofeedback involves very active collaboration of the
Participants/patients are then requested to change the brain activ- patients who have to develop their own personalised strategies for
ity in a desired direction (e.g. up- or downregulate it), and they will successful self-regulation, it leads to an active learning process,
see whether they are successful from the changes in the stimulus, and practitioners told us that they would be interested in sharing
which could see for instance the food picture become smaller. this with their patients and receiving more information about the
Learning can occur in a trial-and-error procedure, but many studies strategies that they used. We are currently thinking about ways
also provide some suggestions for potential strategies. of capturing these strategies and incorporating them into the
therapeutic process.
What kind of impact do you expect on
patients?
A recent study from Cardiff showed substantial improvement P R OJ E C T
in symptoms of depression among patients who trained in the BRAINTRAIN – Taking imaging into the
upregulation of areas responsive to positive affective stimuli – and therapeutic domain: Self-regulation of brain
also in a control group training in the upregulation of a visual area. systems for mental disorders
Another study by our partners at the University of Coimbra tested C O O R D I N AT E D B Y
the feasibility of a social attention training programme for patients Cardiff University in the United Kingdom
with autism, using real-time processing and feedback of EEG signals
to track patients’ attention. This protocol trained social interaction FUNDED UNDER
in several contexts that are relevant to the everyday functioning of FP7-HEALTH
patients with autism, such as ordering a drink in a bar or talking to a
teacher. Although the training programme did not produce attention C O R D I S FA C T S H E E T
improvements, it led to improvements in several clinical secondary cordis.europa.eu/project/id/602186
outcome measures, especially in the domain of depression.
P R OJ E C T W E B S I T E
Can you provide other examples of disorders braintrainproject.eu
that could be alleviated with this method?
In addition to the examples above (depression and autism), we
are looking at post-traumatic stress disorder, alcohol dependence
6
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Breaking through barriers
for a revolution in brain
scans
Current brain scanning techniques have some critical limitations. The EU-funded
BREAKBEN project broke through these barriers, promising better and more
detailed brain scans.
© Semnic, Shutterstock
Brain disorders cause a great deal of suffering and a major eco- sensitive magnetometers – devices used to detect and measure
nomic burden to society. The annual cost in Europe alone has been extremely subtle magnetic fields. The innovative array has been
estimated at EUR 800 billion. Now, a European team aims to put designed to pick up and combine high quality signals for two
Europe at the forefront of a revolution in the scanning and meas- principal scanning techniques for the human brain: MEG (mag-
urement of the human nervous system, to improve the accuracy of netoencephalography) and ULF MRI (ultra-low-field magnetic
current techniques that characterise electrical activity in the brain. resonance imaging).
The BREAKBEN (Breaking the Nonuniqueness Barrier in Electro- MEG measures ultra-ultra-weak magnetic fields generated by
magnetic Neuroimaging) project has designed and built highly the brain, a non-invasive method used to inform clinicians about
7
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
neuronal function. MRI maps the tissue-dependent behaviour a qualitative technology jump with ULF MRI, its applications and
of spinning nuclei in water molecules, giving structural images combinations,” says Ilmoniemi. “This will revolutionise the way
of the brain. we do magnetism-based measurements of the nervous system.”
ULF MRI was first demonstrated in 2004 by the team of Professor The technique could help with
John Clarke at Berkeley University, using just one sensor – known the mapping of epileptic brain
as SQUID. BREAKBEN took the idea to the next step. “We have activity, and ULF MRI may help We believe that we are at
developed the first large SQUID sensor arrays that measure both perhaps even in cancer diag- the edge of a qualitative
MEG and ULF MRI signals with the same sensors,” explains project nostics. The MEG-MRI device
technology jump with ULF
coordinator Professor Risto Ilmoniemi, Head of Neuroscience and will also open up brain imag-
Biomedical Engineering at Aalto University School of Science. ing for new patient groups, such
MRI, its applications and
as those with metal implants. combinations.
Improved workflow and more
Breaking old barriers accurate diagnostics will likely shorten hospital stays and optimise
treatment – thus saving on costs, too.
The German physician Hermann von Helmholtz showed in the 1850s
that it is impossible to uniquely determine the electric current dis- The BREAKBEN team at Aalto University has received Innovation
tribution (in this case, neuronal currents) inside a conductor (in this Launchpad funding from the EU to work towards commercialisa-
case, the head) – even if both the electric and magnetic fields out- tion of the technique. “We are applying for further funding from
side the conductor are perfectly known. This translated into a major Business Finland to prepare the prototype and to demonstrate its
future problem for scientists using MEG to measure neuronal cur- benefits in patient studies,” concludes Ilmoniemi.
rents in the brain, a limitation known as the non-uniqueness barrier.
“We can never be sure on the basis of MEG alone which parts of
the brain are active in a subject or patient,” explains Ilmoniemi. P R OJ E C T
BREAKBEN’s innovation is to combine these two scanning tech- BREAKBEN – Breaking the Nonuniqueness
niques to help avoid this problem, thus breaking through the barrier. Barrier in Electromagnetic Neuroimaging
Simultaneous ULF MRI provides accurate knowledge of the loca- C O O R D I N AT E D B Y
tion of the head and the details of the cortical shape. This allows Aalto University in Finland
a priori information – such as that the source current is only in
the brain cortex – to be added reliably into the data analysis. This FUNDED UNDER
makes the scanning far more reliable. H2020-FET
C O R D I S FA C T S H E E T
Future applications cordis.europa.eu/project/id/686865
BREAKBEN’s breakthrough opens completely different work- P R OJ E C T W E B S I T E
flows for brain imaging. “We believe that we are at the edge of aalto.fi/en
8
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
EU-AIMS: Towards
personalised approaches
for autistic people
The Innovative Medicines Initiative-funded EU-AIMS project has helped to signal a
new era for effective therapies acknowledging the individual needs of autistic people. Its
work has considerably advanced our knowledge of the variability within the biology of
autistic people, identified areas for medical intervention and successfully demonstrated
that compounds can modulate biological differences associated with autism.
Autistic people say that there is no one-size-fits-all approach interaction, but not at the expense of positive aspects of autism.
to helping them. Some say their autism does not need treating Some prioritise treatment of conditions such as epilepsy, which
because they value it as a key part of their identity. Others want they experience alongside autism.
treatment for specific aspects such as sensory difficulties or social
© Dubova, Shutterstock
9
CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
EU-AIMS changes the playing pluripotent stem cells, were used to trace causal links from genes
and environmental factors to molecular changes and biological
field pathways. Researchers also tested whether certain compounds
were effective in rodent and cellular models, thereby identifying
The EU-AIMS (European Autism Interventions – A Multicentre Study potential new treatments based on an improved understanding of
for Developing New Medications) team has changed the playing underlying biological mechanisms. A key task was then to ‘trans-
field by showing there is no single characteristic shared by all late’ findings from animal models to humans by finding measures
autistic people at either that can be safely used in humans – even babies – such as high-
the behavioural, cogni- tech magnetic resonance spectroscopy, electroencephalogram
tive, neurobiological or (EEG), and respectively structural and functional MRI (sMRI/fMRI).
genetic levels. Instead,
We started by improving
researchers began to Another EU-AIMS focus was the identification of biomarkers. These
knowledge of the biological identify individual neu- are any objective measure, such as genes, patterns of brain activity
diversity of autism, translated rocognitive profiles and or a test score, that predicts how a person may develop or respond
results of rodent studies to make link them to patterns to interventions. Biomarkers could be used in autism diagnosis
them relevant to humans, of autism progression. to predict whether symptoms will change during development,
demonstrated that certain EU-AIMS focused on and to select the best interventions, support or treatment for
developing effective any given person. There were three studies with a total of over
compounds could affect the
treatments matched 1 200 participants, and they were assessed at a scale and in-
biological differences seen in to individual needs and depth level unprecedented in Europe. EU-AIMS was also the first
autism – including in autistic adults biological profiles and academic/industry group to obtain ‘qualification advice’ from the
– and searched for biomarkers that overcoming a number European Medicines Agency (EMA). This crucial step improved
could enhance clinical trials. of research obstacles, understanding between regulatory authorities, academics and
including: Traditional industry, and played a key role in developing new EMA guidelines
approaches attempted to treat symptoms without understand- on drug testing.
ing underlying causes or mechanisms; most prior studies only
included 15-30 autistic people, which is insufficient to identify Next, the consortium successfully launched AIMS-2-TRIALS to
‘subgroups’; research into the biology of autism has lacked reli- continue its work, adding a programme of education and engage-
able methodologies; and clinical trials have usually included small ment, particularly for early career researchers, and a group of
numbers of diverse participants, limiting likelihood of success Autism Representatives to ensure engagement of autistic people
because treatments may only work effectively in some subgroups. in research.
The EU-AIMS project took an unprecedented multi-faceted
approach. Declan Murphy, Project Lead and Professor at Kings
College London (KCL) says: “We started by improving knowledge P R OJ E C T
of the biological diversity of autism, translated results of rodent EU-AIMS – European Autism Interventions
studies to make them relevant to humans, demonstrated that – A Multicentre Study for Developing New
certain compounds could affect the biological differences seen in Medications
autism – including in autistic adults – and searched for biomarkers
that could enhance clinical trials.” C O O R D I N AT E D B Y
F. Hoffmann-La Roche AG in Switzerland
New treatment targets through FUNDED UNDER
FP7-HEALTH-IMI1
genes and biomarkers
C O R D I S FA C T S H E E T
One way to identify new treatment targets is to start with genes cordis.europa.eu/project/id/115300
associated with some forms of autism and ask what are their
effects on the development of nerve cells and communication P R OJ E C T W E B S I T E
between them? This impacts a child’s brain development and, in eu-aims.eu
turn, social, cognitive and emotional development. In EU-AIMS,
animal models and new non-invasive techniques, such as induced
10CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Combining neuroscience
and computing, the Human
Brain Project reveals the
brain’s secrets
As a European Flagship Programme, the Human Brain Project (HBP) embraces ICT
advances and pools distributed neuroscience and neuromedicine resources and
competences. This more unified view of the brain will benefit both people and technology.
Despite many years of research, the human brain remains one The HBP was set up to build a unique European technology plat-
of nature’s greatest secrets, partly as labs have typically studied form for neuroscience, medicine and advanced information and
different aspects of it independent of each other. communication technologies (ICT). By linking research areas on an
© Axer & Amunts, INM-1, Forschungszentrum Jülich
11CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
unprecedented scale, it seeks to unlock synergies and accelerate and the responsible use of new technologies, the HBP adheres to
progress. Responsible Research and Innovation (RRI) principles through its
ethics and society research area.
One of the EU’s most
Tackling brain disorders, while
ambitious projects
developing competitive ICT
The project’s twofold vision is mutually reinforcing: To utilise
advanced ICT in neuroscience to better understand the brain and The HBP has already developed a number of new approaches for
its diseases, while drawing inspiration from biology to improve ICT. technology-driven neuromedicine.
With more than 500 scientists from 120 European partner insti- One example is a new personalised brain avatar that has been
tutions across 19 Member States, the HBP is one of the largest developed to improve epilepsy surgery, soon to be tested in a
projects ever funded by the EU. The project has an initial run of 400-patient clinical trial. Another is the creation of the most com-
10 years (2013-2023) during which it will create a permanent prehensive 3D human brain atlas to date.
European neuroscience research infrastructure.
Two of the world’s leading neuromorphic (brain-inspired) comput-
“To meet emerging challenges, we’ve seen a shift from single ing systems, SpiNNaker and BrainScaleS, have been developed
specialised labs to larger interdisciplinary initiatives,” says Prof. under the HBP. The project has also established the FENIX infra-
Katrin Amunts, the Scientific Research Director of the project. structure, a Europe-wide network of supercomputing centres for
“This is partly due to computer-intensive simulations and the data the exchange of large-scale data.
quantities involved, terabytes to
petabytes, beyond the capacity of “We are laying the foundations now with a broad external user
a single research institution.” community for the integrated European research infrastructure.
Brain complexity is This means building not only hardware and software, but also a
breath-taking as it To enable a comprehensive under- culture of usage,” says Prof. Amunts.
standing of brain complexity and
encompasses vastly
apply it to neuromedicine, comput- The six platforms, first made available to the public in 2016,
different spatial and ing, robotics and AI innovations, already allow scientists access to a range of resources, including
temporal scales, from the data sharing according to FAIR different brain models and atlases. External users can access the
molecular level of principles plays a pivotal role in research infrastructure from anywhere in the world via a com-
synapses, cells and nerve the HBP. mon web portal. For all users, a High Level Support Team is on
fibres to connections standby. Scientists can also upload their own data with help from
The HBP approach integrates a curation team.
between whole brain
the experimental investigation of
areas, and from mouse and human brain organisa-
milliseconds to whole life tion, along with cognitive and theo-
spans. retical neuroscience. These benefit P R OJ E C T
six technology-driven platforms in Human Brain Project Flagship
the fields of: neuroinformatics, simulation, high-performance ana-
lytics and computing, neuromorphic computing, neurorobotics and C O O R D I N AT E D B Y
medical informatics. EPFL in Switzerland
“Brain complexity is breath-taking as it encompasses vastly dif- FUNDED UNDER
ferent spatial and temporal scales, from the molecular level of H2020-FET
synapses, cells and nerve fibres to connections between whole
brain areas, and from milliseconds to whole life spans,” says Prof. C O R D I S FA C T S H E E T
Amunts. cordis.europa.eu/project/id/785907
As advances in neuroscience and computing can be such emo- P R OJ E C T W E B S I T E
tive topics, prompting philosophical questions about the ‘self’ humanbrainproject.eu
12CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Collaborative data
underpins advancing
knowledge about the brain
The HBP’s data ‘ecosystem’ comprises software solutions for information collection,
organisation, analysis and sharing, enabling models and simulations to be built. These
depend on the brain’s structure and functions being mapped, from molecular to the
whole brain level.
© Axer & Amunts, INM-1, Forschungszentrum Jülich
Talking of ‘mapping’ the human brain doesn’t really do the endeav- Given the scale and scope of the Human Brain Project's (HBP)
our justice. Unlike say the liver – which contains fewer cell types, ambition, a unique characteristic is the broad range of expertise
organised similarly across the organ – the brain is very inhomo- available. This is critical, since outputs rely on productive interac-
geneous and tremendously more complicated. This makes not tions among these very different communities of researchers and
only description and location vital, but also capturing interactions. developers, impossible to achieve working in isolation.
13CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
“These interactions do present challenges on a scale possibly work together (basic science), to translate this to better diagno-
not seen by any other European project,” says Prof. Jan Bjaalie. ses and treatments of brain disease (applied science for health),
“There are barriers created by different practices, with everyone and also to use this understanding to develop new brain-inspired
bringing traditions from their own fields, and there are termi- technologies (technology).
nology barriers. These barriers actually explain the rationale for
the HBP, while highlighting the importance of getting the data Within these broad goals researchers are guided by some over-
infrastructure right.” arching research targets. Dementia, in particular Alzheimer’s dis-
ease, has been chosen as an area to demonstrate the Medical
Informatics Platform’s functionality, installed at hospitals across
Developing a generic data Europe. Combining this anonymised clinical data facilitates more
efficient analysis.
system
Another example is the use of personalised brain models and
Neuroscience has typically not managed and shared data at the atlases, navigable in 3D, for different applications like better plan-
large scale of the HBP, partly as the focus has previously been ning for epilepsy surgery or deep brain stimulation in Parkinson’s.
on the publication of papers which interpret collected data. For Identification and resection of a small portion of the brain where
the HBP to capitalise on research synergies across all areas of epileptic seizures originate requires a deep understanding of the
neuroscience, the project’s data has to be collaborative. seizure’s organisation. To improve these procedures, the HBP com-
bines a new multi-scale modelling and simulation environment
The data is stored alongside with the project’s brain atlases and other data analysis tools.
metadata tags and is searchable
Infrastructure through a web browser. Specific Critical to achieving the goals of the HBP is that it makes its
features within particular groups systems openly available to the wider research community. Forth-
developments go hand in
of data can be extracted for incor- coming this year will be a new online portal providing access to a
hand with the science, poration into computational mod- series of user-centric services. “While some of the information is
enabling researchers to els and these can be used to per- already available it is not so visible or easy to navigate. The new
find and access relevant form simulations, with the results portal will make it easy for the outside community to find and use
data, use it in clearly compared to data from real brain the tools and services delivered by the HBP,” says Prof. Bjaalie.
defined conditions, share experiments.
the results, and cite the
“Infrastructure developments go
work of the original data hand in hand with the science, P R OJ E C T
providers. enabling researchers to find and Human Brain Project Flagship
access relevant data, use it in
clearly defined conditions, share the results, and cite the work of C O O R D I N AT E D B Y
the original data providers,” says Prof. Bjaalie. EPFL in Switzerland
Given the quantities of data involved, these efforts are only pos- FUNDED UNDER
sible thanks to the use of high performance computing. H2020-FET
C O R D I S FA C T S H E E T
New technologies and clinical cordis.europa.eu/project/id/785907
applications P R OJ E C T W E B S I T E
humanbrainproject.eu
The three primary goals of the HBP are: to better understand how
the different elements of the human brain’s organisation fit and
14CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Understanding how
the brain’s structure
and functions generate
consciousness
Definitions and understandings of consciousness have been hotly contested down the
centuries, engaging philosophers and more recently neuroscientists. To make further
sense of it all, the HBP models both the neurological ‘wood and the trees’
simultaneously.
© cherezoff, Shutterstock
15CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
How is it that the same If we take a broad definition of whole brain. Incorporating both high-level, large-scale models
anatomical structure, our consciousness as experience of that replicate global brain dynamics (top-down) and detailed,
ourselves and the outside world, it biophysical models of realistic neuronal functioning (bottom-up)
brain’s intricate network
can be said to come and go. During will be a key step.
called the ‘connectome’, dreamless sleep it seems absent,
can at times host the seemingly reappears during vivid
complexity of dreaming, before more definitively Medical implications and
consciousness and at reappearing on awakening. But
other times exist how the brain transitions between beyond
these states is poorly understood.
seemingly as dull matter?
The project’s results, combined with others from the HBP, are
This is one of the greatest Acknowledging that answers may forming a more coherent picture. For example, empirical research
mysteries of biology, or lie within neuroscience, one of the with rodents has identified a key neuronal mechanism for the
even physics. objectives of the Human Brain conscious perception of sensory stimuli. This process, known as
Project (HBP) is to develop a com- ‘apical dendritic amplification’, was also detected in the visual
plex model characterised by realistic connectivity, detailed neural system of humans and can be replicated in computer simulations
dynamics and learning rules; one undertakes multiple cognitive as well as in neuromorphic chips to improve image recognition.
tasks by integrating different cortical areas.
In parallel, these activities have elucidated the mechanisms by which
recurrent, complex cortical interactions are disrupted upon falling
Replicating basic activity and asleep, whereby neurons cannot keep track of the inputs they receive.
cognitive tasks “Thanks to our common infrastructure, two lines of research may
soon incorporate a unifying mechanism, accounting for both sen-
The ‘Network models for consciousness’ programme within the sory perception of specific content and global brain state transi-
HBP set out to better understand the relationship between struc- tions,” says Prof. Massimini.
ture and function in the brain, to explain the emergence of the
complex network dynamics which enable perception, prediction, This work has significant medical implications for assessing patient
goal directed behaviour and other higher-end cognitive functions. consciousness and treating disorders, contributing to bedside obser-
vation during loss and recovery of consciousness in sleep, anaesthe-
“How is it that the same anatomical structure, our brain’s intri- sia, coma and related states. Currently, there is a lack of clear brain-
cate network called the ‘connectome’, can at times host the com- based behavioural guidelines; crucial in intensive care medicine.
plexity of consciousness and at other times exist seemingly as
dull matter? This is one of the greatest mysteries of biology, or The results are also of relevance for brain-machine interfaces such
even physics,” project researcher, Prof. Marcello Massimini, says. as those for sensory function restoration, as well as for future AI
architectures.
One of the challenges of building a realistic, data-driven, multi-
tasking consciousness model is that of getting the right structural
and functional modelling parameters to replicate the emergence
of balanced, complex patterns of activity. “Existing models repli- P R OJ E C T
cate either specific cognitive functions or global brain states, but Human Brain Project Flagship
can’t cope with both. This balance between differentiation and
unity is what makes the brain special with respect to conscious- C O O R D I N AT E D B Y
ness,” says Prof. Massimini. EPFL in Switzerland
The programme benefits from the unique range of HBP exper- FUNDED UNDER
tise to build a model using a common infrastructure assembled H2020-FET
from several assets including: atlases, neuroinformatics, brain
simulation, high-performance analytics and computing, medical C O R D I S FA C T S H E E T
informatics and neuromorphic computing. cordis.europa.eu/project/id/785907
This ‘backbone’ collects, curates and integrates structural and P R OJ E C T W E B S I T E
functional data on a scale ranging from single neurons to the humanbrainproject.eu
16CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
When virtual brain models
meet real and simulated
robot bodies
A crucial part of what gives the brain its power is its ability to learn from the
interactions of the body under its control with the world. Neurorobotics research offers
a unique opportunity to understand how this action-perception loop works, and to use
this knowledge to build the next generation of robotic systems.
© maxuser, Shutterstock
The Human Brain Project's (HBP) Neurorobotics Platform (NRP), is This provides researchers with new tools to study topics such as
a set of tools to connect and simulate brain and body together neural control of movement or learning mechanisms in the context
(‘embodiment’). It supports virtual experiments in physically real- of interactions with the environment. Eventually, the knowledge
istic simulated environments through which neuroscientists can derived from these studies will fundamentally change the way
evaluate functional performance of brain models in the context robots are designed.
of behavioural tasks, and so compare or refine them iteratively.
17CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Better understanding brain the means to build better robots, increasing their effectiveness in
areas where the current state-of-the-art is found lacking, such as
structure, function and with situational awareness.
dynamics Additionally, the NRP provides all the simulation tools required to
train robotic systems in simulation: not only does this minimise
The first step for embodied simulation is to define a task to be per- the risk of damage to expensive hardware, it also opens up the
formed and establish the relevant virtual and robotic setup. Next, possibility of using massively parallel training algorithms.
a control architecture (brain) for the
simulation is selected, alongside “Using High Performance Computing will greatly reduce the time
the relevant learning paradigms needed to train robotic systems. Additionally, virtual prototyping
and experimental scenarios. The of robotic platforms should reduce development costs and deliver
Virtualisation of NRP supports users at every step better designed, more robust products,” says Prof. Dr. Alois Knoll,
neuroscience is the of this process. project leader for the NRP.
natural next step, where
The closed-loop experiments then Some HBP partners are now investigating whether brain-inspired
the limits are those
afford neuroscientists a unique cognitive architectures may be useful as control systems for indus-
imposed by computation opportunity to explore the rela- trial robotic arms operating in collaboration scenarios (cobots).
power. In other words: tionship between brain structure, They also extend these research efforts to mechanically-compliant
the sky is the limit! function and neural dynamics in robots, which could provide ‘safety by design’ where robots must
specific behavioural contexts. For operate around humans – in factories for example.
example, several HBP partners are jointly investigating how ocu-
lomotor activity, visual processing and dexterous motor control The NRP is public, online and available for researchers who want
work together in tasks that involve grasping moving objects. Others to test their brain models or to build the brain-inspired robots of
are creating a virtual mouse for the NRP, complete with a highly- the future. “Virtualisation of neuroscience is the natural next step,
detailed musculoskeletal system; the objective is to create a virtual where the limits are those imposed by computation power. In other
model that is as close as possible to its biological counterpart, words: the sky is the limit!” concludes Prof. Dr. Knoll.
so that some animal experiments can be reproduced and carried
out virtually.
These advances may eventually provide significant benefits for P R OJ E C T
medical science. For example, members of the HBP team are Human Brain Project Flagship
modelling the neural circuitry of the spinal cord and simulating it
in order to understand how spinal stimulation can be optimised C O O R D I N AT E D B Y
and/or personalised. One of the principal aims is to assist patients EPFL in Switzerland
with spinal cord injuries in recovering some capacity for locomo-
tion, and neurorobotic developments could prove invaluable here. FUNDED UNDER
H2020-FET
Building better robots C O R D I S FA C T S H E E T
cordis.europa.eu/project/id/785907
Other experiments aim to unlock the secret of the biological
processes that enable the brain to adapt behaviour to changing P R OJ E C T W E B S I T E
environmental conditions. This would provide researchers with humanbrainproject.eu
18CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Shaping the ethical
direction of the HBP for
the public interest
Better understanding the human brain offers insights into what it means to be human,
the causes of brain-related diseases – alongside improved diagnosis and treatment –
while holding out the prospect of advances in artificial intelligence technologies. But at
what ethical cost?
© whiteMocca, Shutterstock
Sensitive to the ethical questions emanating from the research By engaging with external stakeholders, including citizens, the RRI
methodologies, results and potential applications at the interface programme studies specific research consequences, while continu-
of neuroscience and technology, the Human Brain Project (HBP) ing to address traditional ethical concerns surrounding issues such
incorporated a programme of Responsible Research and Innova- as animal or human experimentation. The work contributes to the
tion (RRI). governance of the HBP as a whole, for example by developing
standard operating procedures. It also ensures ethical issues can
19CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
be identified, raised and responded to, through formal and informal Danish Board of Technology. The programme also conducts empiri-
channels, including an independent Ethics Advisory Board (EAB). cal investigations through interviews, surveys and public engage-
ment exercises.
Ethics without borders
The practical and the profound
Ethics touch almost all aspects of most research projects, but the
scale and ambition of the HBP make ethical implications more The HBP research will benefit science, ICT and medicine in innu-
pronounced. Taking data governance, and questions about which merable ways, with important implications for citizens. Novel
data can be used for what purpose and by whom, generates a neuroscience-inspired technologies, such as neuromorphic com-
range of legal as well as ethical concerns. With the HBP, exchang- puting and neurorobotics could fundamentally alter the physical
ing data across national borders adds a further layer of complexity. and cognitive capabilities of humans.
In response, the HBP Ethics and Society research programme led Perhaps the biggest, but most challenging, impact will come with
by Prof. Kathinka Evers has established a working group which the development of better ways of understanding, diagnosing
collates all data-related policies and regulations. “This is a large and treating brain-related diseases. On a more philosophical
and multi-faceted problem that all large data-intensive projects level, better brain understanding may affect the way we
share. Our work in this field is seminal, with potential to shape think about ourselves, exerting a profound influence on social
how future international collaborations work,” says HBP’s Ethics structures and relations.
Director Prof. Bernd Stahl.
Forewarning that those implications are likely to be both negative
Another crucial area of work relates to policy and regulatory and positive, means that the right questions can be asked, in time,
compliance. In some cases, pre-existing EU regulations make by the right people.
the rules clear. “With biomedical
research for example, EU rules However, as Prof. Stahl says, “in most cases there is not one single,
about animal protection, along simple answer to ethical questions and there is no final ethical
with means of enforcement, are arbiter. We encourage open, inclusive and transparent commu-
In most cases there is not well established,” says Prof. Stahl. nication to reach consensus about how to best move forward.”
one single, simple answer “In other less defined areas of
to ethical questions and research, the EU has a clear role
there is no final ethical to lead a broader social debate.”
arbiter. We encourage P R OJ E C T
One such area is that of artificial Human Brain Project Flagship
open, inclusive and
intelligence (AI). While the HBP’s
transparent neuromorphic research may hold C O O R D I N AT E D B Y
communication to reach the key to developing stronger AI, EPFL in Switzerland
consensus about how to the roll out of ever-increasing AI
best move forward. capabilities could prove socially FUNDED UNDER
problematic. Within the field of H2020-FET
employment alone it has the potential to create unemployment
in some sectors, while also increasing workplace discrimination C O R D I S FA C T S H E E T
and bias. cordis.europa.eu/project/id/785907
The ethics and society programme contributes to these social P R OJ E C T W E B S I T E
debates with its own public engagement activities, led by The humanbrainproject.eu
20CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
The Virtual Brain
simulates the brain to
reveal the origins of
disorders
The human ability to use a ‘tool’ (our brain), to build another tool to explain the
workings of the first tool, is an evolutionary gift that distinguishes us from other
animals. Powerful simulation engines like ‘The Virtual Brain’ exemplify both this gift, as
well as its unwrapping.
© Jessica Palmer
21CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
A cognitive irony is that the human brain’s very complexity makes The open source Virtual Brain is freely available for download and
it difficult to theorise about its workings, using thought alone. even modification.
However, computer models can simulate the consequences of
theories, identifying problems and formulating new theories for
neuroscience testing. The hope of the ‘virtual
Within the Human Brain Project (HBP), large-scale brain simulation, human’
and specifically the included Virtual Brain, affords a better under-
standing of the emergence of so-called resting-state networks. Neurodegenerative disorders are one of the most pressing problems
Electroencephalography (EEG) and magnetic resonance imaging facing modern societies. In addition to the individual burden, with
(MRI) show that brains are active even when not engaged in spe- 14 million people across Europe predicted to have dementia in 2030
cific activities. Brain simulations explain these rhythmic networks alone, the cost is predicted to surpass EUR 250 billion by that year.
as spontaneously emerging from the interaction of large groups
of nerve cells via the brain’s white matter. Additionally, mental health conditions such as bipolar disorder,
schizophrenia, depression, anxiety, PTSD, ADHD, alcohol and drug
use disorders currently affect one in six people across the EU, and
Looking for emergent patterns this percentage is rising. The cost of healthcare, social security
and decreased employment/pro-
The brain is on average made up of approximately 86 billion nerve ductivity is EUR 620 billion annu-
cells and 1 quadrillion connections between them. Nerve cells and ally. Existing treatments for these
connections are in turn made up of even smaller elements, such conditions usually rely on medica-
as ion channels and spines, with a range of functional properties. tion, which suppresses symptoms
We are interested in
Simulating the brain down to this level would require the measure- rather than cures illness. high-level cognitive
ment of all the subtle properties of these components. functions like intelligence,
While the underlying mecha- decision-making, memory
But computer processing power is still too limited to perform nisms of these disorders remain and learning, to work out
these calculations in a practical amount of time. On HBP’s Brain unclear, evidence increasingly
the cause of impairments
Simulation Platform (BSP), teams work with simulation engines on points to complex system-
different levels of abstraction – from modelling smaller volumes atic physiological connections,
and to map out
to a high level of detail to more coarse-grained, but still full-brain which are hard to study with improvement strategies.
simulations of brain dynamics. One of the latter is The Virtual Brain, experimental methods alone.
“So, we are not trying to accurately simulate the brain, but rather “With full-brain simulation, and in the future full-body simu-
reveal the large-scale patterns that emerge from the interaction lation, we will better understand the whole human system.
of these elements, like those that emerge in a flock of birds,” says ‘Virtual humans’ would allow us to develop customised inter-
Prof. Petra Ritter, leading the work on the Virtual Brain Project and ventions targeting the combination of genetic, metabolic
connecting it to more detailed simulators on the BSP. and neuronal factors responsible for brain disorders,” concludes
Prof. Ritter.
Dividing the entire brain into areas, the researchers formulate
theories testable by computer modelling. As many details are little
understood or only vaguely specified, they use EEG and fMRI brain
imaging to constrain the models. These simulations enable the P R OJ E C T
team to estimate the connectivity between brain areas yielding Human Brain Project Flagship
so-called connectomes (strengths of interaction between different
brain areas) and so accurately predict brain activity. C O O R D I N AT E D B Y
EPFL in Switzerland
“We are interested in high-level cognitive functions like intelli-
gence, decision-making, memory and learning, to work out the FUNDED UNDER
cause of impairments and to map out improvement strategies,” H2020-FET
says Prof. Ritter.
C O R D I S FA C T S H E E T
So far, amongst other findings, the project has increased under- cordis.europa.eu/project/id/785907
standing about: recovery after stroke, the prediction and charac-
terisation of epileptic seizures, the progress of Alzheimer’s and P R OJ E C T W E B S I T E
the functional implications of brain tumours. humanbrainproject.eu
22CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
A high-performance PET
scanner for integration
into current MRI systems
Although it’s mostly known for its key role in detecting cancer and monitoring its
evolution during treatment, PET imaging actually has a multitude of other clinical
applications. Neurology is one of these. The MINDView project recently tapped into its
potential by developing a compact brain PET imager that can be combined with
existing MRI systems to better diagnose schizophrenia.
© kckate16, Shutterstock
23CORDIS Results Pack on the brain
How the digital revolution is transforming EU-funded brain research
Besides its high resolution and efficiency, the main advantages of Such design also brings cost down considerably, as tasks such as
MINDView’s (Multimodal Imaging of Neurological Disorders) new cutting, polishing, painting and gluing thousands of pixels back into
PET imager are its significantly reduced cost, its size and the fact a single block are avoided. This is a major incentive for hospitals
that it allows for conducting both Positron Emission Tomography which generally must pay between EUR 4 and 7 million for com-
(PET) and Magnetic Resonance Imager Radio Frequency (MRI RF) mercial PET/MRI systems. Smaller clinics that cannot afford a new
imaging at the same time. whole-body PET/MRI system might even be able to finally acquire
this powerful technology.
“MRI and PET provide complementary information,” explains Prof.
Jose Maria Benlloch Baviera, coordinator of the project and Director Last but not least, the new design allows for a portable system
of the CSIC’s Institute for Instrumentation in Molecular Imaging that can easily be removed when it’s not needed for a particular
(I3M). “MRI provides high quality and resolution images on the mor- MRI examination. As Prof. Benlloch Baviera points out, this repre-
phology of the soft tissue, which is very useful in localising lesions sents a major technical breakthrough since the mutual interfer-
in the body. On the other hand, PET ence of MRI and PET imaging modalities, in particular when being
images provide information on the physically so close to each other, would normally be substantial.
physiologic processes occurring in
different organs and substructures. “For instance, the high field (3 Tesla) produced by the main mag-
Smaller clinics that cannot The two technologies potentiate net of the MRI may completely degrade the performance of the
afford a new whole-body each other: For instance, MRI infor- photo-sensors and electronics of the PET scanner. Special photo-
PET/MRI system might mation can be used to better posi- sensors based on silicon technology have been developed and
even be able to finally tion physiologic information from non-paramagnetic electronics and connections have been used
acquire this powerful PET within the brain.” in order to avoid that effect,” he explains. “Furthermore, innovative
shielding of the PET modules and the RF coil have been designed
technology.
The potential for the precise and to avoid eddy currents that will distort the high homogeneity of
early diagnosis of mental disorders the magnetic field required for a top-quality MRI image.”
such as schizophrenia and severe depression is tremendous. Neu-
rotransmitter pathways-specific PET radiopharmaceuticals – such The device is currently being tested on patients with Alzheimer’s
as glutamatergic, serotonergic or dopaminergic – can be imaged, disease (AD).
while areas of the brain activated by performing a task are tracked
with functional MRI. According to Prof. Benlloch Baviera, this might
eventually lead to precise and quantitative diagnosis of mental
disorders that would be impossible to obtain with currently avail-
able techniques.
P R OJ E C T
MINDView’s PET system also stands out through its design. Com- MINDVIEW – Multimodal Imaging of
pared to current devices made of thousands of small crystal pixels Neurological Disorders
coupled to an array of photo-sensors, this one includes only 60
large monolithic crystal blocks coupled in one side to a matrix C O O R D I N AT E D B Y
plane of silicon photo-sensors. “This design has several perfor- Spanish National Research Council in Spain
mance advantages. Since the same amount of light is emitted in
all directions, we can determine the depth of gamma ray inter- FUNDED UNDER
actions by measuring the width of the light distribution. In other FP7-HEALTH
words, we can find the 3D position of the gamma ray impact. This
is a critical feature to avoid image blurring when using scanners C O R D I S FA C T S H E E T
that are close to the organ. Besides, light is directly detected by the cordis.europa.eu/project/id/603002
sensors instead of bouncing back and forth in the pixel. This allows
in principle for better energy and timing resolution, which in turn P R OJ E C T W E B S I T E
reduces signal noise at the image,” Prof. Benlloch Baviera explains. mindview.i3m.upv.es
24You can also read
