We are proud to share how DICOM Vision, our medical image visualization and analysis platform, has been successfully used in a research project at the renowned Giannina Gaslini Hospital in Genoa. This project, described in the scientific paper “An automatic tool performing functional analysis in MR urography in children“, published in the European Journal of Radiology Open, represents a significant step forward in the diagnosis and treatment of pediatric urological anomalies.
The context of the project
Gaslini Hospital, a leader in pediatric research, collaborated with our team to integrate DICOM Vision into an advanced analysis pipeline for magnetic resonance urography (MRU). The study focused on one of the most common urinary tract anomalies in children: congenital obstructive uropathy, which can severely compromise kidney function if not diagnosed and treated promptly.
In the paper, the authors describe an innovative automated pipeline based on a convolutional neural network (Attention U-Net) for the analysis of pediatric MRU images. This pipeline was implemented and tested using DICOM Vision as a management, visualization and reporting platform.
The Role of DICOM Vision
DICOM Vision was the key platform for:
- Automatic segmentation: The neural network accurately segmented the kidneys and renal pelvis in the MRU images. The average accuracy achieved, measured by the Dice Score, was greater than 90% for renal structures, even in complex clinical cases.
- Automated functional analysis: The pipeline allowed to calculate essential parameters such as:
- Split Renal Function (SRF).
- Renal excretion curves to assess the degree of obstruction.
- Automatic Report Generation: Results were integrated directly into the hospital’s PACS as detailed PDF reports, ready for clinicians to review and share.
The benefits obtained
Thanks to the adoption of DICOM Vision:
- Over 100 pediatric MRU studies were analyzed in significantly reduced time.
- The clinical workflow has been simplified, making functional analysis accessible even to teams without AI expertise.
- Doctors were able to view and analyze data directly from the browser, without having to install additional software.
The results demonstrated high reliability and significant clinical impact, improving the accuracy of diagnoses and allowing more targeted interventions.
Conclusions
This use case highlights the potential of DICOM Vision to support clinical research and optimize medical image management. We are excited to see our product contribute to projects that improve the quality of pediatric diagnosis and healthcare.
If you want to discover how DICOM Vision can transform medical image management in your organization, contact us or try the platform today!
Learn more and try a free demo and watch video tutorials
#DVisionLab #DICOMVision #RicercaMedica #Pediatria #Urologia #MedicalImaging #AIInHealthcare
