Welcome to Shanoir!
Shanoir-NG, next generation, bringing Netflix-style technology into medical imaging research
Shanoir-NG is a complete technological remake of the first version of the Shanoir application, but maintaining the key concepts of Shanoir
Why did we take this big effort to reimplement Shanoir?
- Over the years of the existence and usage of Shanoir the technological basis of Shanoir has become quite outdated and most of its original technical frameworks (JBoss 4, Java Server Faces (JSF), Richfaces, JBoss Seam) are not supported and maintained anymore.
- Furthermore the architectures and technologies of how to develop web applications have dynamically changed as well in the last 5 years. The arrival of Single-Page-Applications (SPAs), like Gmail and Twitter, the Docker containerization technology and microservices architectures have dramatically changed the way we develop web applications today.
- This leads to the consequence, that only migrating Shanoir to newer versions of the existing technology stack and to add a REST web service layer were far from being sufficient to extend the lifetime and long-time usage of Shanoir. That is why we work hard on Shanoir-NG to deliver soon a first version:
Delivery date of a first version of Shanoir-NG: foreseen for June 2018.
As promised above, a demo of Shanoir-NG has been given on the APPNING workshop in Paris end of June.
Soon you will find a video here.
What is new in Shanoir-NG?
As we store medical data we take security very seriously, that is why we worked a lot on this side for Shanoir-NG:
- Keycloak (http://www.keycloak.org)
Instead of again and again redeveloping and securing our home-made user management we use a standardized and tested Open Source Identity and Access Management tool and its protocols
- OpenID Connect as authentication protocol
- Full control for principal investigators who can access to what data with which rights
- A new DICOM anonymization module configurable and based on the 2010 DICOM Supplement 142
- Data encryption on production servers
- Security audits on production servers
- Conformity to new European rules, starting from 25.05.2018
- General Data Protection Regulation, GDPR (in French: règlement général sur la protection des données (RGPD))
- To come: certification CNIL and ANSM
To comply with new technological standards and a better user experience Shanoir-NG is a 100% Single-Page-Application:
- Angular 5 as front-end technology (https://angular.io)
- To support as many browsers as possible, even mobile browsers, we choose to implement Shanoir-NG on using Angular 5 as base framework for the front-end
- New designed user interface that simplifies the usage of Shanoir
- Optimized browser viewer for DICOM and NIfTI images even in the browser, based on Papaya
- starting with STOW-RS to do a pre-import of files in Shanoir-NG, Shanoir-NG
will propose a DICOMweb support
- later to come: WADO-RS, QIDO-RS depending what mappings are possible with the
information model of Shanoir (as Shanoir is designed for multi-center clinical studies,
the highest information entity is a clinical study, that has nothing to do with a DICOM
study, which is below a patient in DICOM; in Shanoir clinical studies group unlimited
numbers of patients)
New DICOM and NIfTI viewer + medInria
established tool DICOM and NIfTI files in a web browser (http://ric.uthscsa.edu/mango/papaya.html)
- medInria (https://med.inria.fr), a desktop application to view and process images,
remains the viewer for data within Shanoir and remains its connection to access to Shanoir-NG
Pipelines and processings – directly connected
- On using the interface CARMIN, Shanoir-NG, connects to processing platforms,
especially VIP (https://www.creatis.insa-lyon.fr/vip/) and Boutiques (https://boutiques.github.io)
to access to a large scope of pipelines
- Individual pipeline integrations are possible
- The pipelines execute in a dedicated cluster environment
- The processing results are integrated back into Shanoir-NG
BIDS-native, brain imaging data structure
- Shanoir today uses a proprietary file data structure to store its data.
- With the arrival of Shanoir-NG Shanoir will natively use BIDS to structure its internal data.
- This will help to easily import and exports BIDS-validated and standardised data
on using Shanoir-NG.
Old features, well liked, but refreshed
- Ontology based data organization (OntoNeurolog) for multi-center clinical and preclinical studies
- Automatic conversion of DICOM files to NIfTI files
- Harmonization of DICOM meta-data during import
- Clinical scores management
- Command line access to shanoir on using ShanoirTK
New search and mass data download
- A simplified search field, that searchs in all data has been added based on Solr search technology
- On using this search simplified mass data downloads on using a “shopping cart” will be possible
- All DICOM meta-data will be indexed and searchable
Microservices architecture: ready for population imaging
Before Shanoir was a monolithic application deployed on one single server instance on using old Java Enterprise architecture, now this has changed:
- The entire Shanoir-NG application consists of multiple independent microservices, where each of them has its autonomous functionality
- This makes it a lot easier for developers to know where to work on
- This has lead to a very modular architecture that has nearly no limits in scaling up to whatever data volume may be achieved
Each microservice of Shanoir-NG runs in its own Docker container, that makes Shanoir-NG deployable and better runable on new cloud infrastructures, that scale up to no limits if required.
REST services and swagger
- REST is the new technology to contact the API-Gateway of Shanoir-NG
- All microservices provide swagger interface definitions to even generate clients to call Shanoir-NG in whatever client implementation language desired
Human and small animal data and images in one server
- Easily a plugin for preclinical data can be integrated into Shanoir-NG
- This leads to one Shanoir server available in the browser, that manages in the same application clinical (human) and preclincial (animal) meta-data and images
What is Shanoir?
Shanoir (Sharing NeurOImaging Resources) is an open source neuroinformatics platform designed to share, archive, search and visualize neuroimaging data. It provides a user-friendly secure web access and offers an intuitive workflow to facilitate the collecting and retrieving of neuroimaging data from multiple sources and a wizzard to make the completion of metadata easy. Shanoir comes along many features such as anonymization of data, support for multi-centres clinical studies on subjects or group of subjects.
Shanoir offers an ontology-based data organization (OntoNeuroLOG). Among other things, this facilitates the reuse of data and metadata, the integration of processed data and provides traceability trough an evolutionary approach.Shanoir allows researchers, clinicians, PhD students and engineers to undertake quality research projects with an emphasis on remote collaboration. As a secured J2EE web application, it therefore allows you safely storing and archiving, with no more requirements than a computer with an internet connection!
Furthermore, Shanoir is not only a web application: it is also a complete neuroinformatics platform in which you can easily integrate your existing processing tools or develop your own ones: see ShanoirTk.
Shanoir is a project carried out by the VisAGeS Team, based at IRISA (INRIA Rennes – Bretagne Atlantique Research Centre). This software is released under QPL 1.0 license.