Figure 1 | Scientific Reports

Figure 1

From: Deep learning to automatically evaluate HER2 gene amplification status from fluorescence in situ hybridization images

Figure 1The alternative text for this image may have been generated using AI.

Pipeline of the Aitrox AI model for the automatic detection of HER2 amplification statuses in FISH images. First, a whole FISH image goes through the tumour cell nucleus detector with a feature extractor and predictor to localize the tumour cell nuclei. The detected tumour cell nuclei are cropped from the original image as patches. Second, the signal detector takes patches as input and outputs their HER2 and CEP17 gene signal quantities. HER2 amplification statuses are classified into 5 groups according to the ASCO/CAP guidelines.

Back to article page