Related Experiment Video
Updated: Jul 7, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Applying deep learning to segmentation of murine lung tumors in pre-clinical micro-computed tomography
Mary Katherine Montgomery1, Chong Duan2, Lisa Manzuk1
1Comparative Medicine, Pfizer Inc., 10646 Science Center Drive, La Jolla, CA 92121, United States.
Abstract:
Lung cancer remains a leading cause of cancer-related death, but scientists have made great strides in developing new treatments recently, partly owing to the use of genetically engineered mouse models (GEMMs). GEMM tumors represent a translational model that recapitulates human disease better than implanted models because tumors develop spontaneously in the lungs. However, detection of these tumors relies on in vivo imaging tools, specifically micro-Computed Tomography (micro-CT or µCT), and image analysis can be laborious with high inter-user variability. Here we present a deep learning model trained to perform fully automated segmentation of lung tumors without the interference of other soft tissues. Trained and tested on 100 3D µCT images (18,662 slices) that were manually segmented, the model demonstrated a high correlation to manual segmentations on the testing data (r2=0.99, DSC=0.78) and on an independent dataset (n = 12 3D scans or 2328 2D slices, r2=0.97, DSC=0.73). In a comparison against manual segmentation performed by multiple analysts, the model (r2=0.98, DSC=0.78) performed within inter-reader variability (r2=0.79, DSC=0.69) and close to intra-reader variability (r2=0.99, DSC=0.82), all while completing 5+ hours of manual segmentations in 1 minute. Finally, when applied to a real-world longitudinal study (n = 55 mice), the model successfully detected tumor progression over time and the differences in tumor burden between groups induced with different virus titers, aligning well with more traditional analysis methods. In conclusion, we have developed a deep learning model which can perform fast, accurate, and fully automated segmentation of µCT scans of murine lung tumors.
Insights
Scientists developed a deep learning model for automated lung tumor segmentation in micro-CT scans of mice. This AI tool significantly speeds up analysis, improving accuracy and consistency in cancer research.
Area of Science:
- Biomedical Imaging
- Artificial Intelligence in Medicine
- Cancer Research
Background:
- Lung cancer is a major cause of mortality, with genetically engineered mouse models (GEMMs) offering valuable insights into human disease.
- Micro-Computed Tomography (µCT) is crucial for in vivo tumor detection in GEMMs, but manual image analysis is time-consuming and prone to variability.
- Automated segmentation methods are needed to improve the efficiency and reliability of µCT image analysis in lung cancer research.
Purpose of the Study:
- To develop and validate a deep learning model for fully automated segmentation of murine lung tumors from 3D µCT images.
- To assess the accuracy, speed, and consistency of the automated model compared to manual segmentation methods.
- To evaluate the model's performance in a real-world longitudinal study of lung cancer progression in mice.
Main Methods:
- A deep learning model was trained using manually segmented 3D µCT images of GEMM lung tumors.
- The model's performance was evaluated on independent datasets, comparing its segmentation accuracy (correlation coefficient, Dice Similarity Coefficient) against manual segmentations.
- The model's efficiency was assessed by comparing its processing time to manual segmentation time and its results against inter- and intra-reader variability.
Main Results:
- The deep learning model achieved high correlation with manual segmentations (r²=0.99, DSC=0.78 on testing data; r²=0.97, DSC=0.73 on independent data).
- Automated segmentation performance was comparable to inter-reader variability (r²=0.98, DSC=0.78 vs. r²=0.79, DSC=0.69) and close to intra-reader variability (r²=0.99, DSC=0.82).
- The model processed over 5 hours of manual segmentation work in just 1 minute and successfully tracked tumor progression in a longitudinal study.
Conclusions:
- A deep learning model enables fast, accurate, and fully automated segmentation of µCT scans for murine lung tumors.
- This automated approach significantly enhances the efficiency and reliability of quantitative analysis in lung cancer GEMM studies.
- The developed model holds promise for accelerating preclinical lung cancer research and drug development by streamlining image analysis.
More Related Videos
11:31Using Micro-computed Tomography for the Assessment of Tumor Development and Follow-up of Response to Treatment in a Mouse Model of Lung Cancer
Published on: May 20, 2016
06:51Utilizing 18F-FDG PET/CT Imaging and Quantitative Histology to Measure Dynamic Changes in the Glucose Metabolism in Mouse Models of Lung Cancer
Published on: July 21, 2018