Related Experiment Video
Updated: Apr 16, 2026
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
Automated tumor volumetry using computer-aided image segmentation
Bilwaj Gaonkar1, Luke Macyszyn1, Michel Bilello1
1Department of Radiology, University of Pennsylvania, 3600 Market St, Suite 380, Philadelphia, Pennsylvania, 19104 (B.G., M.B., M.S.S., H.A., X.D., C.D.); Center for Biomedical Image Computing and Analytics (B.G., L.M., M.B., H.A., X.D., C.D.) and Department of Neurosurgery (L.M., M.A.A., Z.S.A., D.O.R., S.M.G.), University of Pennsylvania, Philadelphia, Pennsylvania; and Siemens Medical Solutions, Malvern, Pennsylvania (Y.Z.).
Rationale And Objectives:
Accurate segmentation of brain tumors, and quantification of tumor volume, is important for diagnosis, monitoring, and planning therapeutic intervention. Manual segmentation is not widely used because of time constraints. Previous efforts have mainly produced methods that are tailored to a particular type of tumor or acquisition protocol and have mostly failed to produce a method that functions on different tumor types and is robust to changes in scanning parameters, resolution, and image quality, thereby limiting their clinical value. Herein, we present a semiautomatic method for tumor segmentation that is fast, accurate, and robust to a wide variation in image quality and resolution.
Materials And Methods:
A semiautomatic segmentation method based on the geodesic distance transform was developed and validated by using it to segment 54 brain tumors. Glioblastomas, meningiomas, and brain metastases were segmented. Qualitative validation was based on physician ratings provided by three clinical experts. Quantitative validation was based on comparing semiautomatic and manual segmentations.
Results:
Tumor segmentations obtained using manual and automatic methods were compared quantitatively using the Dice measure of overlap. Subjective evaluation was performed by having human experts rate the computerized segmentations on a 0-5 rating scale where 5 indicated perfect segmentation.
Conclusions:
The proposed method addresses a significant, unmet need in the field of neuro-oncology. Specifically, this method enables clinicians to obtain accurate and reproducible tumor volumes without the need for manual segmentation.
Insights
This study introduces a fast, accurate, and robust semiautomatic method for brain tumor segmentation. It enables precise tumor volume quantification, addressing a critical need in neuro-oncology without manual segmentation.
Area of Science:
- Neuro-oncology
- Medical imaging analysis
- Computational anatomy
Background:
- Accurate brain tumor segmentation and volume quantification are crucial for diagnosis, monitoring, and treatment planning.
- Manual segmentation is time-consuming and not widely adopted.
- Existing automated methods often lack robustness across different tumor types and imaging variations.
Purpose of the Study:
- To develop and validate a semiautomatic method for brain tumor segmentation.
- To provide a fast, accurate, and robust solution for tumor volume quantification.
- To overcome limitations of existing methods in handling diverse tumor types and image qualities.
Main Methods:
- A semiautomatic segmentation technique utilizing the geodesic distance transform was developed.
- The method was validated on 54 brain tumors, including glioblastomas, meningiomas, and metastases.
- Validation involved both qualitative assessment by clinical experts and quantitative comparison with manual segmentations.
Main Results:
- Quantitative comparison using the Dice measure demonstrated strong agreement between semiautomatic and manual segmentations.
- Expert ratings indicated high-quality computerized segmentations.
- The method proved robust to variations in image quality and resolution.
Conclusions:
- The proposed semiautomatic method fulfills a significant unmet need in neuro-oncology.
- It allows for accurate and reproducible brain tumor volume quantification.
- Clinicians can achieve reliable results without relying on manual segmentation.
