Automated Deep Learning Pipeline for High-Throughput Preclinical PET-CT Analysis

A deep learning-based pipeline automates organ and tumor segmentation and quantitative analysis for longitudinal preclinical PET-CT studies, reducing manual variability. Demonstrated in a murine case study from acquisition to uptake curve extraction. Performance metrics vs. manu…

• Protocol paper describing an end-to-end pipeline for high-throughput small-animal imaging, illustrated with a single murine longitudinal PET-CT case; no sample size or reader study.
• Limitation: entirely preclinical; no human data and no external validation of segmentation accuracy against a reference standard.
• Covers acquisition, reconstruction, multi-subject sorting, atlas alignment, deep learning segmentation, 3D visualization, and quantitative reporting.

RadPigeon summaries are original and for information only. They are not clinical advice.