Fig. 5: Comparison of segmentation results of nuclei segmentation models trained with different types of training data.

A SEG and DET segmentation scores of nuclei segmentation models trained with different types of training data. Scores can range from zero (worst possible) to one (best possible). Three manually corrected image patches from different image regions serve as test data. Maximum indicates a model trained on the test data and is considered an upper boundary. Based on this score, a dotted gray horizontal line is drawn to indicate this maximum. Dark blue color indicates training data generated by manual annotation. The nuclei model provided by Cellpose is depicted in light blue color. Dark and light orange colors indicate the pure use of synthetic training data generated with physical simulation-based and GAN-based approaches, respectively. Gray color indicates the use of synthetic data generated by the GAN-based transformation of synthetic nuclei labels. The error bars represent the standard deviation across six training runs of the segmentation models. Since the Cellpose nuclei model is pretrained, no standard deviation is provided. B Qualitative comparison of nuclei segmentation results. Representative single optical sections of ground truth patches are shown for enhanced clarity. The first and second columns display the raw image signal and its corresponding ground truth, while subsequent columns show the segmentation masks obtained with the segmentation models. Additionally, the last row visualizes the DET-related errors of the third row, including false-negative, false-positive, and required splitting operations. A complete visualization of DET errors across all ground truth patches is given in the Supplementary material (Fig. A2). Scale bar: 25 μm.