Sustainable PErsonalized E-health for brain Diseases (SPEED)
Università degli Studi di Bari "Aldo Moro"

A. Problem Analysis

 1. What was the problem before the implementation of the initiative?
Despite substantial recent progress, our understanding of the principles and mechanisms underlying human brain function remains incomplete. There is no effective strategy to experimentally map the brain across all its levels and functions, yet. Modern supercomputing technology offers a solution, models and simulations allow researchers to understand novel physiological patterns, predict behaviors and trends, eventually unveil the etiology of diseases, in cases where measurements and experimental manipulations would be ethically or technically impossible. UN sustainable development goals promote well-being for all at all ages which means personalized medicine and innovation as natural instruments to manage people health. Taking care of patients no longer means just curing a disease, but also ensuring them a high standard of life; with this regard aging related diseases, such as the neurodegenerative ones, offer a tremendous example of a public concerns addressed by our initiative. It is well known how a brain disease could have a devastating effect not only on patients, but also on their families and yields huge costs, often not sustainable, for healthcare systems especially of least developed countries. According to the World Alzheimer Report 2016, 47 million people are affected by dementia and it is expected that over 131 million people will develop dementia by 2050. The Alzheimer’s disease (AD) represents up to the 70% of dementia’s cases. These diseases often have their origin many years before the manifestation of the illness and are preceded by a phase of mild cognitive impairment of variable duration. To date there is no cure for neurodegenerative diseases, but treatment can still help in reducing symptoms and providing a better quality of life. Thus, early diagnosis and treatment are of paramount importance and crucial to detain progression of the disease in its initial stages. Traumatic brain injuries too may cause structural and morphological damages whose locations depend on the impact and secondary events. The deficits range from impairment of higher level cognitive functions to comatose states. Evaluation of the damages is crucial for taking decisions about post-traumatic treatment. Quantitative neuroscience can infer alterations of the functional and structural brain connectivity, giving a precise measure which can support the physician during the diagnosis, overcoming the subjectivity of personal judgments. However, this kind of fine diagnosis needs sophisticated technological and data analysis tools. Neuroimages and computational algorithms require a big amount of data storage which is delivered by specialized datacenters. Besides, highly specialist support is needed to perform the analysis. Usually, all of them are not within the capacity of healthcare systems. The situation becomes even more critical in the Least Developed Countries which do not have expertise or enough economical means to afford such technologies. Understanding the human brain is one of the greatest challenges facing 21st century science for the first time in history, modern Infrastructure and Communications Technology seems to have brought this goal within sight. Setting up cutting edge technologies and computational infrastructures which support, through dedicated tools, medical responses specialized to people’s needs is then extremely important for facing such severe diseases.

B. Strategic Approach

 2. What was the solution?
This initiative concerns a strategic synergy between public health, research and technology to develop on-demand software solutions for personalized medicine basing on an open source cloud infrastructure. We design and implement data mining and predictive models to evaluate the clinical status of subjects based on genetic profiles and brain MRI data. We offer computing solutions and resources to manage and store big data and run the designed analyses, offering these workflows as a service.

 3. How did the initiative solve the problem and improve people’s lives?
The proposed solution has its foundations on three pillars: (i) the management of big medical data (genetics, imaging, biochemical, …); (ii) the definition of algorithms and models which extract from data personalized medical responses, such as diagnosis, patient assessment and follow-up prediction, just to mention a few; (iii) the customization of these analyses in terms of software as a service solutions. Medicine is experiencing a data explosion, primarily driven by advances in genetics and imaging technology. However, effective strategies to integrate the data and to identify the unique "biological signatures" of neurological and psychiatric diseases still lack. In this sense, the contribution of physical methodologies, deriving from the expertise gained in analyzing and interpreting big data, for example in high energy or nuclear physics, can be strategic. In fact, new databasing and data mining technologies now make it possible to federate and analyze the huge volumes of data accumulating in hospital archives, leading eventually researchers to identify the biological changes associated with disease and opening new possibilities for early diagnosis and personalized medicine. In the longer term, a multidisciplinary integrated approach will allow to modify models of the healthy brain to simulate disease. Disease simulation will provide researchers with a powerful new tool to probe the causal mechanisms responsible for disease, and to screen putative treatments, accelerating medical research and reducing the huge suffering and costs associated with diseases of the brain. The more urgent neuroscience objectives we addressed are: the federation of clinical, genetics and imaging data to extract unique disease signatures or to unveil patterns; to access multi-level data analysis and combine different information and sources; to define classification models based on biological features and markers for brain diseases for diagnostic purposes. We faced these objectives by addressing the following targets: the development of a new class of data classification techniques, new computational strategies and algorithmic solutions which could be used to tackle medical problems. Above all, these tasks require a comprehensive approach, able to take into account both scientific and computational concerns. The need for dedicated informatics infrastructures to manage workflows for "big data" analyses can only be faced in a multidisciplinary framework. Specific analytic methodologies and infrastructures that will conjointly enable researchers and physicians to determine the biological signatures of brain diseases have to be carefully planned to make an efficient use of computing distributed structures modern ICT provides nowadays. Our hope and belief is that this is the only rational approach to develop in future effective treatments for diseases such as Alzheimer’s disease or Parkinson’s disease. Accordingly, we designed and developed fully automated diagnosis support systems to evaluate cognitive impairment basing on MRI data and genetic analyses coming from people affected by traumatic brain injuries, Multiple Sclerosis, Schizophrenia and neurodegenerative disease, such as Alzheimer and Parkinson diseases. All these people are the target audiences of our initiative. We exploited the computational resources of the ReCaS data center ( made available thanks to joint efforts of the Italian University, the Apulia regional administration and the European funding programs for the infrastructural development. Finally, we offered these solutions as a service for free use to physicians of private and, above all, public hospitals, to reach the highest number of subjects, especially the poorest and most vulnerable. The motivations that inspire the proposed initiative coincide with the Target 3.8 from the SDGs.

C. Execution and Implementation

 4. In which ways is the initiative creative and innovative?
The initiative is innovative as it is the first and unique experiment trying to translate research and clinical practice in a remote service and on-demand. Thanks to the specificity of the initiative it combines resources and efforts from both public and private institutions. Above all, it addresses challenges for which no consolidated strategy exists yet and proposes novel solutions which gained acknowledgement from scientific community with publication in international journals and were awarded in international competitions among the best performing methods. In particular, our methodologies were awarded by the MICCAI Machine Learning Challenge 2014 (, MICCAI CADDEmentia Challenge 2014 (, Alzheimr’s disease Big Data DREAM Challenge #1 2015 (!Challenges:DREAM) and MICCAI Mild Traumatic Brain Injury Outcome Prediction ( For these competitions, we developed novel solutions for predicting Alzheimer’s disease basing on structural MRI analysis; we implemented complex network analyses to unveil relationships between genes and phenotypes in Schizophrenia; we studied dedicated models to understand how brain injuries can yield a cognitive impairment. All the proposed solutions exploit general methodologies which can in principle be adopted for other diseases, in fact our current activities deal with Multiple Sclerosis and Parkinson's disease.

 5. Who implemented the initiative and what is the size of the population affected by this initiative?
Our initiative is mainly administered through the action of the Italian ministry of Public Instruction and Research. In particular, this was born as an academic effort to lead cross-disciplinary research in a region, Apulia, which was recognized by the European Union to be one of the Italian regions which most lagged behind in terms of development. Accordingly, Apulia became one of the so-called “Convergence regions”. Our mission was to support this effort, mainly economic, driving into the academic environment of the Physics Department of Bari University a genuine spirit of both scientific and technologic innovation aiming at combining our skills and expertise with the personalized medicine field. In collaboration with the local section of the Italian Institute of Nuclear Physics we established a network of relationships and knowledge with public and private hospitals of our regions. We collaborate with: “Policlinico di Bari - Ospedale Giovanni XXIII” public hospital (, “Azienda Ospedaliera Card. G. Panico” private hospital ( and “Istituto Tumori Bari - Giovanni Paolo II”, an institution dedicated to both research and healthcare. Our services can now reach the whole Apulian population (about 4 million people) and potentially, through the Italian public healthcare system, the whole Italian population (about 60 million people). Moreover, our services are freely available and accessible to every private or public institution. There is also a portal in which whoever makes a request can register and access on the demand some of the services we offer (
 6. How was the strategy implemented and what resources were mobilized?
Over twenty years, Bari team has maturated a long lasting experience in developing algorithms and methodologies in medical physics. The feasible solutions Bari built up made the group generate new and important collaborations with partners from hospitals. Medical applications required by our clinical partners are usually extremely demanding both in terms of scientific background and computational resources, particularly those requiring the analysis of genetic data and imaging. Thus, a particular attention has been paid to ICT, in parallel with the development of sophisticated analysis tools. In particular, Cloud technology appears to fit the requirements of such applications. For example, those technologies are able to provide easily and seamlessly the needed computational power as well as the storage resources to record the data produced. The technological side of our initiative was funded since 2013 to dedicate computational resources to medical applications. The ReCaS-Bari computing farm was built by the ReCaS project (, PRISMA project (, funded by the Italian Research Ministry of Education, University and Research to the University of Bari and INFN (National Institute for Nuclear Physics) with about 20 million euros. In particular, the data center offers over 180 servers for a total amount of 12000 cores. Each new server hosts 256GB RAM, 4GB RAM per core. Additionally, it offers about 3.5PB of disk space and 2.5PB of tape space. The data center is one of the biggest Italian supercomputers built with public funds. It's very well integrated into national and international infrastructures like WLCG (World LHC Computing Grid), EGI (European Grid Infrastructure), EGI Federated Cloud. The ReCaS Farm supports several life science communities and projects (Medical Physics, Elixir project, LifeWatch), etc. In particular the life-science community uses up 6% of the farm resources. The centre offers: · the possibility to require specific services on virtual machines; · the possibility to execute parallel jobs based on MPI library; · facilities to manage the execution of a bunches of independent jobs (automatic grid scheduling) also by means of WebServices; · “Cloud Storage” services based on WebDav and ownCloud; The Bari ReCaS data center hosts a cloud infrastructure that is an open source software platform providing a federated IaaS/PaaS cloud computing solution. This cloud environment exploits the hardware infrastructure of Recas-Bari datacenter and respects all the paradigms at the base of the definition of cloud computing: on-demand self-service resources that are pooled, can be accessed via a network, and can be elastically adjusted by the user. We actually are experimenting novel technical solutions to connect all the institutional partners (hospitals) by optic fibers internet connections, by which magnetic resonance images could be rapidly sent to the data center, elaborated by the algorithms and give a response in real time. We are able now to offer dedicated solutions for personalized medicine within an open source cloud infrastructure.

 7. Who were the stakeholders involved in the design of the initiative and in its implementation?
The initiative was designed by the Bari Medical Physics Group, a cross-disciplinary research group led by prof. Roberto Bellotti, associate professor in Applied Physics at the Bari University. The research group has a consolidate collaboration with the local branch of the National Institute of Nuclear Physics, in particular it is involved with the neuroimaging experiments led by dr. Sabina Tangaro. The research group is involved on activities related to the analysis and understanding of images and patterns, in pattern recognition, machine learning, complex network analysis and related applications to diagnostic medical imaging. The aim of the research group is to develop and deploy related applications in the field of diagnostic medical imaging. Besides, a particular attention is given to the "Big Data" approach, distributed computing on grid environments as the European Grid Infrastructure (EGI) and cloud services for large scale. The research group expertise includes (but is not limited to): - Computer Aided Detection (CAD) systems for the analysis of biomedical images, in particular brain and mammographic images. - Pattern Analysis of morphological and genetic alterations due to brain disease. The scientific productivity of the group is available online at the group site: A not secondary role was played within this initiative by the collaboration instituted by the group with private and public hospitals or research institutions. For example, we mention: “Policinico di Bari - Ospedale Giovanni XXIII” public hospital (, “Azienda Ospedaliera Card. G. Panico” private hospital ( and “Istituto Tumori Bari - Giovanni Paolo II” an institution dedicated to both research and healthcare.

 8. What were the most successful outputs and why was the initiative effective?
Our initiative supports the early diagnosis of cognitive impairment and in general the possible occurrence of brain diseases. The first result we are proud to mention is that the methodologies developed by our group now support the diagnosis of several neurodegenerative diseases, such as Alzheimer’s disease, for which, at the present, no cure exists. This is probably the first and most important output of this initiative. A genuine combination of both personalized medicine and innovation, within the spirit of UN SDGs. In 2010, the number of people over 60 years of age living with dementia was estimated at 35.6 million worldwide. This number is expected to almost double every twenty years. Early and accurate diagnosis has great potential to reduce the social and economic costs related to care and living arrangements as it gives patients access to supportive therapies that can help them maintain their independence for longer and delay institutionalization, achieving a better quality of life. Thus, a second output of this initiative is the development of personalized medicine strategies for early diagnosis, a benefit which could greatly increase the economic sustainability of healthcare. The methodologies developed by our group allow robust morphological analyses of the human brain on demand, thus our technical background and the computational resources we have are made available to other scientific or clinical communities which can take benefit from our services. We think that sharing our knowledge with other communities and offering for free services of analysis and support to diagnosis is one of the most important achievements we obtained. A third output of our initiative is the sharing of knowledge and the birth in Apulia of a truly cross-disciplinary research environment. A fourth output is the boost in innovation, let us think to ReCaS, that our initiative has carried in a territory, such as Apulian region, which stays behind in terms of development. A fifth output is the virtuous circle our initiative has generated. Let us think to the ongoing collaborations with healthcare institutions and, more importantly, the improvements that are ongoing, as the new project with the “Azienda Ospedaliera Card. G. Panico” private hospital, according to the hospital and ReCaS will be connected via ultra fast fiber optics so physicians could rapidly store and access to all their medical data, require automated analysis and get results in real time.

 9. What were the main obstacles encountered and how were they overcome?
The difficulties we encountered were twofold. Firstly, a particular mention deserves the search for funds, the management, the public relationships, the political and marketing efforts performed. The design of the initiative lasted for a decade, culminating in the awarding of about 40 million euros from the European Community for the construction of the ReCaS data center and the implementation of the PRISMA cloud platform. This work was really hard for a public research group, whose main activities used to be the study of high energy physics or the Higgs boson. In fact, we had to win friction forces even inside our community to explain this initiative and convince the community of its goodness. Secondly, this initiative would not be a success without the work of many young researchers and their team leaders driving the every-day research activity. The methodologies we have developed required thousands of hours of work from more than 20 people from the Bari University and the National Institute of Nuclear Physics. It was of fundamental importance, as facing scientific challenges still open, the development of innovative strategies, but at the same time, it was important to keep these solutions friendly for the use of communities, let us think about clinicians, whose scientific background is far from those of physicists or computer scientists. Finally, we had to develop customized technological solutions according to the specifications of the end-users. Even in this case a remarkable effort was made to overcome existing technical limitations.

D. Impact and Sustainability

 10. What were the key benefits resulting from this initiative?
Brain diseases can be very heterogeneous, Alzheimer’s disease and Parkinson disease are good examples. Alzheimer’s disease heterogeneity is evident in the disease's clinical, anatomic, and physiologic characteristics. The presence of considerable inter-subject and intra-subject heterogeneity suggests that subtypes of the disease exist. It is possible to define several Alzheimer subtypes in regard to the behavioral features, inheritance (familial or sporadic), time course of progression, age of onset (presenile or senile), many other characteristics. What is clear is that it would be of paramount importance the definition of biomarkers allowing an early diagnosis of the disease. Parkinson’s disease (PD) is a heterogeneous neurological disorder, firstly described almost two centuries ago, basically related with early death of dopaminergic neurons in the substantia nigra and characterized by classical motor features associated with Lewy bodies. It is recognized that age is the greatest risk factor for PD, its incidence reaches a maximum at about 80 years of age, thus the rising life expectancy is expected to increase the number of patients by more than 30% by 2030. One of the most important features of PD is its slow progression. It is known that the average latency between onset of early symptoms and occurrence of motor features is about 12-14 years. Thus, even in this case there would be in principle room for the development of treatments to reduce or stop the disease process before the death of dopaminergic neurons has triggered irreversible damages. Our initiative gives hope that a deeper comprehension of this kind of diseases can be reached. The possibility for scientists and clinicians (even from the poorest countries) to access sophisticated algorithms and computational resources is a great opportunity for them, but also for their patients; besides, our algorithms provide for each patient a personalized score assessing its cognitive impairment or the probability of disease: this information can be crucial if combined with other meta data to understand the clinical history and, hopefully, the disease outcomes. Another aspect we want to outline is that quantitative evaluation of cognitive impairment has direct applications for traumatic sports, an example is football. Traumas experienced during football matches could have serious consequences; a quantitative evaluation of this aspect still lacks, in fact we participated and won last September an international challenge on this theme. Applications are almost infinite, as the possibilities to experience a brain trauma. Our initiative forecasts the advent of smart cities, e-health and personal healthcare-as-a-service according to the consolidated cloud paradigm.

 11. Did the initiative improve integrity and/or accountability in public service? (If applicable)
All the proposed initiative is managed by public institution and funded by public resources, so the principal goal is not the profit but pursuing SDGs, particularly the target 3.8 which accounts on achieving universal health coverage for every people, particularly for the poor or vulnerable component of the society. Also, transparency, open sharing of knowledge, knowledge transfer are basilar aspect of our project. This spirit is surely opposite to corruption and bad public administration. It is also well known, for example when reading the UN Survey for e-government, that countries with a pronounced vocation for e-platforms (for government, health, ...) are usually more resilient to bad administrative practices, including corruption and concussion. Our initiative supports the use of innovation technologies and digitalization of health-care services, as an opportunity to trigger within our country, and especially in Apulia, a greater attention of citizens, private and public institutions and political stakeholders towards on demand services and smart cities.

 12. Were special measures put in place to ensure that the initiative benefits women and girls and improves the situation of the poorest and most vulnerable? (If applicable)
Our applications are publicly available on an open cloud infrastructure, according to the SaaS paradigm and it is accessible through a common web browser. In this way, anyone can use them by means of an internet connection without expensive computational resources. Our initiative provides powerful tools to help advance brain disease prevention strategies and treatment development. They represent a method to improve the quality of life of patients by using complex technology, freely and safely. Sophisticated computational tools used for the analysis and medical data need a great amount of data storage and computational resources which, in most cases, are not available within healthcare institutions because of their high cost and the complexity of their use which requires highly specialized personnel. This fact is even more true for Least Developed Countries. Our initiative encloses all these characteristics making them available from remote to the whole international community. It is worth mentioning that a free training for our services is available online through the moodle e-learning platform. In this sense, our initiative falls in target 3.8 of SNG: “Achieve universal health coverage, including financial risk protection, access to quality essential health-care services and access to safe, effective, quality and affordable essential medicines and vaccines for all”.

Contact Information

Institution Name:   Università degli Studi di Bari "Aldo Moro"
Institution Type:   Academia  
Contact Person:   Roberto Bellotti
Title:   Prof.  
Telephone/ Fax:   +390805443226
Institution's / Project's Website:  
Address:   Via Amendola 173
Postal Code:   70125
City:   Bari
State/Province:   BARI

          Go Back

Print friendly Page