EASE: Clinical Implementation of Automated Tumor Segmentation and Volume Quantification for Adult Low-Grade Glioma

Area of Science:

• Neuro-oncology
• Medical imaging analysis
• Artificial intelligence in healthcare

Background:

• Quantifying non-enhancing low-grade glioma growth is crucial for prognosis but challenging due to irregular tumor shapes.
• Automated volumetric assessment offers a reliable solution for tumor growth quantification, contingent on fully automated segmentation.
• Recent advancements in automated tumor segmentation pave the way for clinical implementation of volumetric analysis.

Purpose of the Study:

• To describe the clinical implementation of automated volume quantification for low-grade gliomas using the EASE (Erasmus Automated SEgmentation) application.
• To evaluate the success rate of automated segmentation and its impact on clinical diagnosis.
• To establish protocols for safe and effective clinical use of volumetric measurements.

Main Methods:

• Implementation of the EASE (Erasmus Automated SEgmentation) software for automated tumor segmentation.
• Inclusion of radiologist visual quality control for segmentation accuracy.
• Development of clinical protocols for using volume measurements in diagnosis and algorithm updates.

Main Results:

• EASE was applied to 55 patients in the first 3 months.
• Radiologists successfully utilized EASE for volume-based diagnosis in 36 patients, based on three consecutive measurements.
• Volume-based diagnoses consistently aligned with conventional visual diagnoses.

Conclusions:

• Automated segmentation and volume quantification via EASE is feasible for clinical use in low-grade gliomas.
• The integration of EASE supports accurate diagnosis and prognostic assessment.
• This represents a significant step in translating automated segmentation techniques into routine clinical practice.