Related Experiment Video
Updated: Aug 5, 2026
06:48
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
Automated Fetal Brain Segmentation in Slice to Volume Reconstruction in Clinical Practice
Usha D Nagaraj1, Joshua S Greer2, Jean A Tkach2
1From the Philips Healthcare Cincinnati (J.S.G.), OH; Department of Radiology and Medical Imaging (U.D.N., J.A.T., B.K.-F.), Biostatistics and Epidemiology (B.Z.), Cincinnati Children's Hospital Medical Center, Cincinnati, OH. usha.nagaraj@cchmc.org.
AJNR. American Journal of Neuroradiology
|July 29, 2026
Summary
Automated slice to volume reconstruction (SVR) of fetal brain MRI provides accurate volumetric measurements. These data show strong correlations with manual 2D measurements, suggesting a potential replacement for clinical practice.
Area of Science:
- Medical Imaging
- Radiology
- Fetal Medicine
Background:
- Manual 2D measurements of the fetal brain are standard in clinical practice.
- Slice to volume reconstruction (SVR) offers potential for automated volumetric data acquisition.
- Evaluating the accuracy and utility of SVR-derived fetal brain measurements is crucial.
Purpose of the Study:
- To assess if automatically derived volumetric data from SVR MRI of the fetal brain can replace manual 2D measurements.
- To determine the feasibility and accuracy of automated fetal brain segmentation using SVR.
- To compare SVR-generated volumetric data with traditional 2D measurements.
Main Methods:
- Retrospective study of fetal MRI with SVR in pregnant women.
- Acquisition of 2D T2-SSFSE images as input for SVR.
- Implementation of an SVR pipeline using PRIDE for automated segmentation and volumetric analysis.
- Radiologist assessment of segmentation map accuracy and comparison of automated vs. manual measurements.
Main Results:
- SVR was performed in 70 fetuses (mean gestational age 28.2 ± 4.3 weeks).
- Reconstruction and volumetric analysis time was 8.6 ± 3.4 minutes, correlating with gestational age.
- Good segmentation quality was achieved in 70% of cases; poorer quality was associated with lower gestational age.
- Strong positive correlations were found between SVR-derived volumes (supratentorial, cerebellar, brainstem) and corresponding manual 2D measurements.
Conclusions:
- Automated fetal brain segmentation using SVR is successful in most cases.
- SVR-generated volumetric measurements strongly correlate with manual 2D measurements.
- Automated SVR shows potential as a practical alternative to manual 2D measurements in clinical practice.
