Integrated Biomedical Informatics for the Management of Cerebral (@neurIST)

Published on: 08/04/2010

@neurIST is focused on cerebral aneurysms and intends to provide an integrated decision support system to assess the risk of aneurysm rupture in patients and to optimize their treatments. @neurIST believes that the current process of cerebral aneurysm diagnosis, treatment planning and treatment development is highly compromised by the fragmentation of relevant data.

The project presents a new paradigm to understand and manage cerebral aneurysms. A complete IT infrastructure will be developed for the management and processing of the vast amount of heterogeneous data acquired during diagnosis. @neurIST will benefit patients with better diagnostics, prevention and treatment because it will combine efforts of clinicians and industry.

Through research clinicians will gain a greater insight in aneurysm understanding, while industry will be dragged by these achievements to develop more suitable medical devices to treat the disease. @neurIST will provide an IT infrastructure for the management, integration and processing of data associated with the diagnosis and treatment of cerebral aneurysm and subarachnoid hemorrhage. This infrastructure will:

  • Facilitate clinicians the diagnosis and study of the disease, as a result of providing a seamless access to patient data using data fusion and processing of complex information spanning from the molecular to the personal level;
  • Provide a better planning and personalisation of minimally invasive interventional procedures for patients, after linking modern diagnostic imaging to computational tools;
  • Collaborate in the development, extension and exploitation of standards and protocols at all project stages;
  • Share biomedical knowledge providing access to a set of software tools and platforms such as @neuLink, @neuFuse, @neuRisk, @neuEndo, @neuCompute and @neuInfo;
  • Create awareness through scientific dissemination and collaboration;
  • Explore the business opportunities directly arising from @neurIST.

@neurIST will contribute to a better knowledge of cerebral aneurysms and other diseases by:

  • Finding evidences of links between genomics and cerebral aneurysms;
  • Helping clinicians to take decisions and select more appropriate treatments.

@neurIST will improve patient care by:

  • Identifying those patients with high risk of rupture by assessing a personal risk factor, thereby reducing the patient's operation risks and anxiety;
  • Improving personalised design of endovascular devices.

@neurIST will result in major benefits to the healthcare system by:

  • Reducing healthcare costs through the suppression of an estimated 50% of unnecessary treatments. This will save of the order of thousands millions of Euros per year in Europe.

@neurIST will promote and validate:

  • A new diagnosis and treatment paradigm extendable to other disease processes using complex data fusion, information extraction, processing and inferential deduction;
  • The use and development of bioinformatics, medical informatics and medical devices standards and protocols.

Technology solution

@neurist will make use of platforms such as

  • @neuLINK
  • @neuFUSE
  • @neuRISK
  • @neuENDO

Main results, benefits and impacts

  • Project achievements:

In FP6 project @neurIST we set out to define a fully-structured approach to healthcare by integrating clinical management with continuously active research. In doing so we have constructed an information framework capable of uniting all aspects of disease management - from electronic patient records through imaging, complex analysis, genetic insights, knowledge synthesis and risk assessment – across multiple clinical institutions, and also introduced anonymised data transfer to the associated research community. Validated research results can then be used to inform further changes in clinical practice.

The @neurIST project has shown how a clinical workflow can be completely transformed by the use of Information Technology, to yield improved healthcare outcomes at lower cost. The approach combines sophisticated data, workflows and knowledge mechanisms with the latest 3D diagnostic simulations, automated research integration and secure international networking. The project’s results have implications for healthcare decisions at national level.

We chose to build our system in the clinical area of intracranial aneurysm (ICA), but the techniques are readily transferrable. ICA is an ideal condition for this work because clinicians must make a difficult choice between risk of rupture, which typically leads to death, and risk of neuro-impairment through costly intervention; improved decision-making has profound implications for health benefits and cost-saving. Improvements in clinical practice come from two sources: improved techniques and advancing knowledge. @neurIST has delivered multiple outputs in both categories.

  • Knowledge Discovery:

Medical research produces new results at an astonishing rate. Clinical practice must keep pace, but staying up-to-date by hand is an impossible task. @neurIST has developed automatic search-and-detect systems that continuously identify new research findings and allow the conversion of raw information into knowledge that can improve clinical practice.

Repositories of patient data come to life when suitably interrogated, offering fresh insights into disease and treatment. Tools have been built to allow findings from continuing research to be explored for meaning.

Knowledge Discovery is possible from both structured and unstructured data sources, across multiple scales (genes, gross anatomy, patient-level and whole populations) and multiple formats (text, images, raw, structured and derived).

  • Diagnostic Techniques:

Patient-specific data has always been the essence of diagnosis, but only now are the sophisticated methods of advanced aeronautical engineering becoming available to medical practice. @neurIST has developed an automated process that reconstructs each patient’s anatomy and investigates in remarkable dynamic detail the internal physical behaviour driving the disease process.

New predictive measures can be extracted that define the unique characteristics of the patient’s condition and play an essential role in building for the clinician a detailed patient-specific risk assessment supporting the treatment decision.

The project has shown how every aspect of clinical information processing can be streamlined – sometimes radically innovated – into focused workflows and how, in combination, such changes can transform practice and performance into a new, efficient, integrated methodology for the future. This work has implications at a national scale of healthcare provision, and is now being taken forward by a prestigious clinical institution.

  • Topic  Achievements:
    • Patient records: Research-aware electronic integrated clinical data collection tools and semantically annotated data repositories;
    • Data-sharing:  Federated security allowing simple but highly controlled institution-level access to shared data across Europe;
    • Pseudonymization: automated pseudonymised data transfer to the research community;
    • Genetics: Fresh insights into genetic predisposition to aneurysm formation and - separately - aneurysm rupture;
    • Complex analysis: an automated computer-modelling toolchain that processes images to extract the aneurysm and its vessels, and derive sophisticated measures of rupture risk;
    • Knowledge creation: automated knowledge discovery from structured and unstructured data, spanning multiple scales (genes, gross anatomy, patient-level and whole populations) and multiple formats (text, images, raw, structured and derived);
    • Treatment planning: computerized rehearsal of medical device placement and deployment, with post-operative performance prediction and customized device design;
    • Decision support – Risk: pre-operative risk assessment of intervention risk, allowing treatment optimisation and, importantly, the possibility of a no treatment strategy;
    • The Research-Knowledge-Clinic cycle: Full support for the process of clinical data driving research innovation, in turn providing improved knowledge back to the clinical workflow.

Academically, the work has currently resulted in more than 120 peer-reviewed publications and more than 150 participations in conferences.

Lessons learnt

As a multi-disciplinary research effort, @neurIST successfully faced the challenge of bringing together people form diverse scientific domains. An incremental approach to overall system development, as pursued by @neurIST, was well adapted to the time and effort required to reach a common understanding amongst stakeholders and helped in coping with the complexity of the system. Equally, adopting the SOA framework allowed developers of end-user applications the mandatory shielding from the details of the underlying distributed IT infrastructure and resources to focus on their applications.

Nonetheless, there are points where early misunderstandings have had knock-oneffects. The potential of the ontology to play a central role in the system for example, was one particularly difficult area for non-ontology experts to grasp and fully bring to fruition, thus slowing progress. Another challenge in developing the ontology was finding an appropriate granularity and level of detail, in particular for supporting annotation and integration of heterogeneous data sources.

Scope: Pan-European