Build your COVID-19 Research Program with Flywheel and Google Cloud
Apply for COVID-19 research offers and credits from Flywheel and Google
Flywheel has partnered with Google Cloud to provide special pricing and credits for qualified COVID-19 research projects.
Flywheel streamlines and accelerates your COVID-19 research program with a comprehensive research data management solution for medical imaging and related non-imaging data. Today, Flywheel works with leading global research institutions to manage their medical imaging data (MR, CT, PET, Ultrasound, and others), clinical data, and associated processing for large-scale academic and clinical research. Flywheel enables data collection and de-identification, secure multi-site collaboration, curation and annotation, preprocessing and pipelines, and analysis deployed Google Cloud Platform (GCP).
Flywheel is offering special pricing and credits for qualified projects, including use of Flywheel on Google Cloud at no cost while longer term funding is being established. Flywheel is also offering services to help speed deployment of your project.
Google is also offering credits for qualified COVID-19 researchers. Google Cloud research credits can be used for most computing services on GCP such as storage, compute, and data analysis, including those related to the deployment of Flywheel’s research data management platform.
To apply, interested researchers should describe their research problem, project timeline, and the GCP tools they intend to use. Applicants must be verified researchers from a commercial, nonprofit, government or academic research institution. Applications will be reviewed on a rolling basis during the COVID-19 pandemic.
Apply for the Flywheel and Google Cloud Platform offering in this interest form.
Flywheel Launches Enterprise-Scale Medical Imaging Solution for Life Sciences
A digital transformation is underway in the Life Sciences and organizations are motivated to reduce R&D costs, improve operational efficiencies and shorten drug development timelines. To be successful, these organizations need to cost-effectively leverage high-quality data for the exploration of new solutions and reuse data collaboratively within data science & research teams. However, traditional R&D infrastructures are insufficient for large-scale data collection, complex data validation, automated quality controls, and scaling processing (and AI) pipelines.
Next-Gen Imaging Research Platform to Drive R&D Efficiency and AI Initiatives
Flywheel offers a scalable research-first platform that can capture imaging and related data from multiple siloed sources, including CROs and historical clinical trials. Flywheel integrates with existing data pipelines and software tools to ingest multi-modal data, and then consolidate, curate, validate, and disseminate that data to analytics and machine learning platforms. With the provenance to support regulatory approvals and scientific reproducibility, Flywheel supports secure collaboration to solve problems with multi-disciplinary teams internal and external to organizations.
Now with Flywheel for Life Sciences, the solution incorporates new enterprise scaling features that help support the needs of large, global life sciences organizations and their massive volumes of data from diverse sources all over the world. These features include high speed bulk loading, horizontal and vertical scaling, version control deployments and the provenance to support audit readiness.
Cloud-Scale Solution Backed by Professional Services
Flywheel for Life Sciences leverages the full power of modern cloud computing to deliver a fully managed service allowing digital, operations, data, and clinical teams to focus on research and results rather than IT. The Flywheel team partners with life sciences organizations to tailor and deliver a system that meets enterprise requirements and critical schedules.
“We are excited to launch our Life Sciences solution to streamline and leverage imaging data assets and machine learning workflows, helping organizations increase R&D efficiency and accelerate drug discovery,” said Travis Richardson, Chief Product Officer of Flywheel.
To learn more, please reach out to info@flywheel.io.
Improved Collaborative Workflows with Custom Roles and Permissions
Flywheel is committed to provide customization tools for a secure, collaborative workflow. Previously, Flywheel offered fixed, predefined roles and permissions for administrators to match to site users and project collaborators. Now, administrators can have complete control over user permissions and defining roles for projects using a simple interface.
Tailor Your Workflows With Custom Roles and Permissions
The Custom Roles and Permissions interface enables you to:
Align the Flywheel system with specific responsibilities of the users. Select user capabilities for project management, access to files and metadata, and computational permissions.
Ensure your workflow is consistent with your organization’s policies. Define roles that ensure your research process follows organizational policies for viewing, modifying, and deleting data.
Implement fine-grained control to prevent unauthorized use and reduce risk. Ensure data integrity by entrusting only specific users with the ability to modify data.
Easily coordinate on responsibilities in multi-site collaboration. Reflect the permissions collaborators need to have while observing multiple institutional procedures.
Flexible Controls Enable a Variety of Applications
For example, here’s how you might use custom roles and permissions:
- Data Managers in clinical trials can be restricted from viewing or modifying analyses.
- A statistician role can be created with permissions to run gears, perform analyses but are restricted from deleting or modifying underlying data.
- A compliance coordinator role can be created with limited permissions to view metadata and data only to ensure project contents are valid and complete.
Powerful Controls and Easy-to-Use
Custom roles are defined at the site level, enabling consistency in permission sets across the site. Controls over Flywheel permissions include a user’s level of access with data, which data permissions apply to, and other key operations like running analysis or downloading data.
Creating an "Analyst” role with limited project permissions but has the ability to work with analyses
Research groups may then select from the site’s defined roles for the roles that fit their workflow. Users are assigned a specific role or multiple roles at the project level.
Setting roles at the project level - Note that users can be assigned multiple roles
You may find additional information about setting User Roles & Permissions in our documentation.
New Web DICOM Uploader with Configurable De-Identification
Flywheel has extended the functionality of the Web-based DICOM Uploader in version 11.2 to support configurable rules for de-identification at the project level. With the new Web Uploader, users have an easy to use, zero-footprint solution for remote data collection with the ability to de-identify data in the browser, before upload to the Flywheel system. De-identification rules may be specified to meet the specific needs of any project, ensuring that PHI protocols are adhered to from the point of data ingestion.
Simplify Data Collection for Multi-Center Studies
The Web Uploader provides an easy-to-use solution for collecting data for multi-center studies including:
- Uploading DICOM data without needing to install additional software
- Configurable de-identification to meet project-specific requirements
- Automatically enforcing institutional de-identification policies
Rich Functionality, No Installation Required
No installation is required to upload data to Flywheel, simplifying the process for researchers with administrative or technical constraints at their location. Users upload uncompressed DICOM data via their web browser by picking a target project via the Select Project dropdown and dragging and dropping standard DICOM files into the Drop Zone.
To access the Web Uploader, select “Upload DICOM” in the Left Navigation Panel in the “DATA” group. After you drag studies or series to the Drop Zone, Flywheel will automatically parse the DICOM headers and present the target subject, session and acquisition labels for editing as needed. You may upload multiple DICOM studies or series at once.
De-Identification Enforced Before Upload
After editing/confirming your subject, session, and acquisition labels, data is de-identified locally before being uploaded to the Flywheel system.
De-identification rules are attached to a project. Administrators can configure de-identification rules for a given project by attaching a file named deid_profile.yaml in the Project → Information → Attachments pane the project menu. In order to ensure clear requirements for de-identification, the web uploader will not be available for a project unless there is a de-identification profile installed. If no de-identification is required, simply provide a profile with no rules.
Powerful Tool for Configuring De-identification Rules
De-identification profiles are simple YAML files that specify a DICOM tag and the de-id action you’d like to take on that tag.
Examples of possible de-identification actions include:
- Removing DICOM tags such as PatientID or AccessionNumber
- Replacing tag values with a hash or other value
- Date offsets to obscure dates
- Convert dates to age in years
- And more
Example:
Once de-identification preferences are set, the de-identified profile can be validated by uploading a DICOM and then viewing the tags in the native DICOM viewer. Testing of de-identification rules can also be done locally using the Flywheel CLI.
Example:
You may find additional information about the Web Uploader or uploading Non-DICOM data in our documentation.
Advanced Search for Finding and Repurposing Data
Flywheel has released a powerful new Advanced Search capability in version 11.2. This tool extends previous search functionality to allow users to construct complex queries and quickly pinpoint the data they need.
Key features include:
- The ability to search any metadata including ROIs
- New SQL-like query language for complex AND/OR queries
- An easy-to-use visual query builder
- The ability to save and manage queries
Advanced Search enables a variety of applications including:
Exploring project data to ensure consistency and quality. Search on any metadata to find cases that meet or don’t meet required criteria using metadata on any object in the Flywheel database.
Finding and repurposing data from multiple projects for secondary applications and research. Search for relevant data sets meeting requirements for new applications using standard metadata, custom metadata and attributes of sessions and experiments.
Creating machine learning data sets. Create training sets from search results using metadata and ROIs created using Flywheel’s integrated DICOM viewer, for example.
Accessing Advanced Search
To access Advanced Search, simply click ‘Advanced Search’ in the left navigation panel from the search results screen.
A Simple, Powerful Visual Query Builder
Flywheel’s new Visual Query Builder makes it easy to construct complex queries combining search terms for projects, subjects, Gears, and file metadata.
Metadata fields can be easily added to search queries from within the Visual Query Builder. As you click to define the field you are searching for, Flywheel offers dropdowns for easy selection.
Powerful SQL-Like Query Language
Users that are more familiar with SQL can manually construct queries using Flywheel’s simplified query language: FlyQL. FlyQL enables access to all metadata, including DICOM tags and custom metadata.
While clicking through the Visual Query Builder, you will see a FlyQL query being built on the left. You may choose to write your queries in this editor as well. The type-ahead feature, which suggests text to autocomplete the query, allows you to quickly find the data points you need.
Manage and Save Queries
After easily constructing a specific query, click Save Query to preview, share, and reuse the query. Find your saved queries below the FlyQL Query Editor.
Managing Search Results
For the following query looking for male Subjects between 40 and 80 years, these example results are obtained.
| subject.sex = male AND session.age_in_years >= 40 AND session.age_in_years <= 80 |
Use the left navigation panel to further filter results. You may choose to see results in Sessions, Acquisitions, Files or Analyses at the top.
After selecting results, users may click the Actions dropdown to choose to download selected results, add them to a collection, or run a batch gear on search results.
Search Image Annotations and ROIs
AI Developers may search for annotated images, including by regions of interest (ROIs), and create machine learning training sets with ease.
| file.type = dicom AND file.info.roi.label = Lesion AND file.info.roi.area > 300 |
Additional Examples
| Project.label IN [“PSY Study”, “NIMH Project”, “JBM Project”] AND session.created >= 2020-02-01 AND session.created <= 2020-02-29 AND session.satisfies_template is false |
Search for all sessions in three projects created over the last month that do not adhere to their project’s template
| analysis.label CONTAINS afq AND file.name CONTAINS analysis_summary AND subject.cohort = Control |
Search for analysis summary reports for all Automated Fiber Quantification analyses performed on a subject cohort.
You may find additional information about Advanced Search in our documentation. Please reach out to our Support Team with any questions about this feature or about other updates in 11.2.
Feature: Radiology-Grade Viewer with Image Annotation
Flywheel has made significant upgrades to our native DICOM viewer to meet the needs of radiologists performing safety reads or participating in blind multi-reader studies. The integrated viewer can now load full studies with a simple click of the button at the project and session level. The new work list interface and radiology-grade tools make loading, reading, and annotating cases fast and easy.
Our ROI and annotation tools allow users to efficiently annotate imaging data sets for applications such as machine learning. ROI labels may be customized to meet project-specific needs and ensure that annotations are consistent and standardized. Qualitative summary forms may be created on any project to capture the results of radiology reads in a structured manner, thereby eliminating the need to use an external spreadsheet or similar approach. All ROIs and qualitative summaries are searchable and easily accessible via Data Views and the Flywheel SDKs for subsequent analysis.
Flywheel is also releasing a new WADO (Web Access to DICOM Object) service that allows you to access imaging data through your viewer of choice or integrate with existing DICOM workflows.
Four AI Workflow Trends from RSNA 2019
The Biggest Trend: Maturing Implementation of AI
Attendees who visited our booth last year were interested in learning about AI capabilities. This year they were bringing questions about implementing infrastructure needed for AI and how to scale AI research in their organizations. Scaling access to clinical data and interoperability appears to be a rising concern this year. Organizations are also gradually accepting cloud scaling as a secure option.
Radiologists are beginning to plan for AI in their standard workflows. There were many radiologists in our booth asking questions with respect to AI research in their current clinical workflows.
Data Curation for Research Still Falls Short
The focus in many workshops and presentations from radiologists was “data wrangling” and data set quality. We received many questions from attendees regarding metadata management and labelling tools. At the same time there is growing recognition that clinical systems don’t meet the needs of the research and AI development communities. Additionally, an entirely new class of solution that supports the research workflow is needed.
We recommend Dr. Paul Chang’s (University of Chicago) AuntMinnie interview during RSNA: “AI is like a great car … Most cars still need gas and roads. In the context of this analogy, gas is vetted data and the road is workflow orchestration that is AI-enabled... The only way to make a transformative technology real is to do the boring stuff, the infrastructure stuff.”
Everyone Noticed the Busy AI Showcase
The AI Showcase was very active this year. In 2018, there were roughly 70 vendors in the AI Showcase, but this year there were 129, including many international AI vendors. We noticed growth in AI development for cardiac and brain imaging.
It’s Imminent: Equipment Vendors are Integrating AI Workflows
AI is moving beyond the desktop as imaging equipment manufacturers have their eye on supporting research workflows. Leading equipment manufacturers like Philips and Canon displayed developments in their interfaces to support AI or analysis tools in a disease specific applications. Flywheel is expanding partnerships with AI vendors and equipment vendors in addition to supporting clients performing imaging and clinical research.
Four Takeaways from BioData World West 2019
BioData World West wrapped up its third year! A mix of experts from industry, academia, and government mingled and mused on the data management supporting the healthcare industry.
Below are the insights from our own Chief Technology Officer, Gunnar Schaefer and Director of Sales, Marco Comianos, who attended.
Gunnar Schaefer, Co-Founder and CTO of @Flywheel_io presents on scaling medical imaging and machine learning in clinical research
Share quality data within your organization
The main focus among conversations at BioData this year was making data accessible across departments and organizations. Letting data flow freely between labs in life sciences organizations creates a feedback loop from health network partners and previously unprofitable drug trials. In health networks, data scientists can highlight opportunities where patients are underserved to create better experiences and processes that can be streamlined to cut costs.
When these different sources of data are merged, unconventional combinations of biomedical data can point to obscure patterns of disease. Scientists from organizations like GenomeAsia, Sidra Medicine, and AstraZeneca presented their findings from blending microbiome and genetic research, genotypic and phenotypic data, and imaging and text data.
In order for machine learning to power artificial intelligence applications, data must be routed, organized, cleaned, and standardized from the moment of creation. More important than proper data storage is the ability to query a system over and over for renewed insight. Genentech introduced the need to store data so it is FAIR: findable, accessible, interoperable, and reusable. That way, data are ripe for query and can integrate together for analysis.
However, it’s important to remember that no matter how well sources are linked together, data must be high-quality and machine learning investigations must be ethically supervised. As Faisal Khan of AstraZeneca put it: “Tortured data will confess to anything.”
Looking forward, expect life sciences companies to adopt better data principles in their data strategies, refine what’s working already, and search for software that bridges the gaps.
Being precise about requirements for precision medicine
Much of the groundwork for precision medicine is now being laid, though mostly in oncology. At BioData, speakers gave direction for its high-value applications.
Today’s genomics research can treat previously-untreated rare diseases. A panel addressed how data sharing must accompany public genomic projects to optimize therapeutic development for rare diseases. Presenters also reported on diversifying the pools for large genome projects. On the treatment side, analysts explained methods to match an individual’s genomic profile with one out of many pre-existing drugs, saving time for patients facing debilitating diseases.
These advancements require access to large amounts of data with well-defined interoperability. Looking forward, expect the general hype around precision medicine to fade, making way for discussions about infrastructure which enable answers to disease-specific precision questions.
Machine learning shortens both ends of drug trials
Beyond the potential for drug discovery using genetic markers, algorithms were showcased which had correctly predicted the pharmacokinetics and effectiveness of drug compounds. Not only does this technology assist researchers and cut costs for developing compounds or finding targets, once therapies are in clinical trials, AI can predict the likelihood of certain subpopulations having an adverse reaction to a drug. Clinical trial pools normally miss these portions of the population, which can result in a public perception crisis.
Looking forward, expect to see AI use with historical clinical data and patient data becoming a competitive factor in shortening the time horizon for successful drug launches. We’ll also see which AI vendors become the most productive partners for life sciences organizations.
AI specialists come ready to partner
If data scientists hold some healthy skepticism of practically applying machine learning, AI specialists showed up with the energy to compensate. AI specialists are drawing talent from universities to specialize in anatomical regions. Companies in this vertical are also starting to partner with each other to complement their deep expertise in one region.
Many AI companies at BioData specialize in genomics and digital slide pathology, so look forward to development and consolidation in this field. Fewer imaging analysis companies were present at BioData - stay tuned for the imaging market insights yet to come out of RSNA!
At RSNA’s Annual Meeting, Flywheel will be exhibiting from December 1st to December 5th in the AI Showcase. Schedule a demo and find us at booth #11618.
High Performance Computing (HPC) with Flywheel
Flywheel now provides additional support for customers’ existing infrastructure and workflows with two significant advances.
HPC with Singularity and Slurm
Flywheel has introduced support for High Performance Computing environments running singularity and Slurm. This integration enables users to save on computing costs and allows them to take advantage of significant investments made in existing computing resources. Flywheel Gears may be run on HPC clusters regardless of whether the Flywheel system is deployed on premises or in the cloud.
Rsync-like Data Access
CLI Sync provides users with an efficient method of syncing their Flywheel data with local file systems. This improved data portability allows customers to leverage pipelines and hardware not yet integrated with Flywheel. The rsync-like functionality ensures that only changes to data are migrated to local systems, thus saving time and cost.
Flywheel Introduces New Advanced Search
Flywheel augments its fast, powerful search tool with the introduction of Advanced Search, providing users the ability to construct complex queries and quickly pinpoint the data they need. When this feature is introduced, find Advanced Search by hitting ‘Enter’ in the Search Bar and clicking ‘Advanced Search’.
Visual Query Builder
Advanced Search features a visual query builder, providing even the least technical users with the ability to quickly and easily construct highly specific queries. Search queries may be saved for sharing and reuse.
Powerful SQL-Like Query Language
Users that are more familiar with SQL can manually construct queries using Flywheel’s simplified query language: FlyQL. FlyQL enables access to all metadata, including DICOM tags and custom metadata.
Search Image Annotations and ROIs
With more and more customers looking to Flywheel to manage large, diverse data sets for Clinical Trials and Machine Learning projects, the need for improved data management tools has come to the forefront. With advanced search, users will have the ability to quickly and easily locate data across multiple projects matching any number of criteria. AI Developers can search for annotated images, including by regions of interest (ROIs), and create machine learning training sets with ease. Researchers can quickly locate subjects based on custom metadata to create collections or run batch analysis directly from search results.