The Challenge
The research project undertaken by TYKS, focusing on the "H&N MRI of retropharyngeal edema," aimed to assess the diagnostic accuracy of radiologists and clinicians in detecting specific edema patterns on neck MRIs related to acute infection. These edema patterns possess prognostic and clinical significance, making the study critical. However, previous research was fundamentally limited because it typically involved only a few radiologists within the same hospital, using the same equipment. For TYKS’s crucial study, it was necessary to recruit a large number of expert readers. Attempting to send images to numerous external hospitals proved logistically difficult, highlighting the urgent need for a cohesive and compliant platform capable of managing a high volume of reads and collecting all resulting data in a single, accessible repository.
The Solution
TYKS, through the efforts of Principal Investigator Prof. Jussi Hirvonen, engaged Collective Minds to utilize their Research Platform. The solution provided Collective Minds' cloud-based research workspace to support the complex multireader study. The services provided by Collective Minds included setting up a dedicated research pipeline for extracting imaging biomarkers from MRI images, incorporating both multireader and automated image segmentations.
Key platform features unique to the Collective Minds solution that enabled TYKS's success include:
- Compliant Data Management: The platform enabled customer upload of anonymized neck MRI scans to a repository accessible only by TYKS. This process was facilitated by CM-Connect PACS connection and Automatic De-Identification & Pseudonymization capabilities, ensuring compliant data inclusion and sharing.
- Streamlined Workflow: Collective Minds set up the research pipeline and provided a Customized eCRF (electronic Case Report Form). The system established the multireader workflow where 30 expert readers could perform the "Radiology Review task + eCRF" efficiently.
- Comprehensive Data Export: Results, including classification data, were provided in tabular form (e.g., .csv) and DICOM RT format.
- Ease of Use: The system’s intuitive design drastically simplified the workflow for participants, including 20 external readers and even surgeons who rarely interact with MRI images in their daily work.
The Results
The adoption of the Collective Minds research platform delivered substantial and immediate results for Turku University Hospital, primarily through massive efficiency gains in reader workflows.
Efficiency Impact: This efficiency gain resulted from the massive acceleration of the reading process. The estimated time required for the reader workflow was reduced by 50%, from 14 minutes per task without CM to just 7 minutes with CM. This time savings amounted to a total of 245 hours saved across all readers and cases.
Accelerated Research Timeline: The complex multireader study, involving 70 patient cases and 30 readers, was completed in 30 days. This represented an enormous acceleration compared to the typical 4–6 months required for comparable trials. This efficiency translated directly to a faster time to publication, estimated to be 2–6 months sooner.
Workflow Quality Improvements: The use of the platform provided comprehensive achievements across the study management:
- Zero coordination delays were reported.
- The system achieved instant audit-ready compliance.
- All manual workflow bottlenecks were eliminated.
- The platform enabled crucial global collaboration without complexity.
.png?width=1024&height=1024&name=image%20(17).png)
