Fig. 3: Label-free multimodal nonlinear images of human ovarian carcinomas. | Light: Science & Applications

Fig. 3: Label-free multimodal nonlinear images of human ovarian carcinomas.

From: Deep learning autofluorescence-harmonic microscopy

Fig. 3: Label-free multimodal nonlinear images of human ovarian carcinomas.

a–d Full view of the large field (5.4 × 5.4 mm2) at FIGO Stage IA, IC, IIC, and IIIC, with the white dashed lines separating the registered input (top left) and the network output (bottom right). White squares indicate ROIs magnified in (e)–(h) (left: input, right: output) for the highest metric enhancement. The white solid line in each inset image refers to the line of the shown cross-section. i–m Quality metrics for the input (pink) and output (pale green) large images are compared in the histograms. Scale bars, 500 μm.

Back to article page