Automated segmentation of lungs and lung tumors in mouse micro-CT scans
Gregory Z. Ferl,
Kai H. Barck,
Jasmine Patil,
Skander Jemaa,
Evelyn J. Malamut,
Anthony Lima,
Jason E. Long,
Jason H. Cheng,
Melissa R. Junttila,
Richard A.D. Carano
Affiliations
Gregory Z. Ferl
Preclinical & Translational PKPD, Genentech, South San Francisco, CA 94080, USA; Department of Translational Imaging, Genentech, South San Francisco, CA 94080, USA; Corresponding author
Kai H. Barck
Department of Translational Imaging, Genentech, South San Francisco, CA 94080, USA; Corresponding author
Jasmine Patil
Genetic Science Group, Thermo Fisher Scientific, South San Francisco, CA 94080, USA
Skander Jemaa
Data, Analytics and Imaging, Product Development, Genentech, South San Francisco, CA 94080, USA
Evelyn J. Malamut
Preclinical & Translational PKPD, Genentech, South San Francisco, CA 94080, USA
Anthony Lima
Department of Translational Oncology, Genentech, South San Francisco, CA 94080, USA
Jason E. Long
ORIC Pharmaceuticals, South San Francisco, CA 94080, USA
Jason H. Cheng
Department of Translational Oncology, Genentech, South San Francisco, CA 94080, USA
Melissa R. Junttila
ORIC Pharmaceuticals, South San Francisco, CA 94080, USA
Richard A.D. Carano
Data, Analytics and Imaging, Product Development, Genentech, South San Francisco, CA 94080, USA
Summary: Here, we have developed an automated image processing algorithm for segmenting lungs and individual lung tumors in in vivo micro-computed tomography (micro-CT) scans of mouse models of non-small cell lung cancer and lung fibrosis. Over 3000 scans acquired across multiple studies were used to train/validate a 3D U-net lung segmentation model and a Support Vector Machine (SVM) classifier to segment individual lung tumors. The U-net lung segmentation algorithm can be used to estimate changes in soft tissue volume within lungs (primarily tumors and blood vessels), whereas the trained SVM is able to discriminate between tumors and blood vessels and identify individual tumors. The trained segmentation algorithms (1) significantly reduce time required for lung and tumor segmentation, (2) reduce bias and error associated with manual image segmentation, and (3) facilitate identification of individual lung tumors and objective assessment of changes in lung and individual tumor volumes under different experimental conditions.