H&E, Tumor Detection H&E, Tumor Detection

Process APP

Revert

#10197

RUO

H&E, Tumor Detection

The H&E Tumor Detection APP uses deep learning to automatically identify and outline tumor regions in H&E-stained brightfield images of different tissue types. It delivers accurate tumor ROIs and a quantitative total tumor area measurement (mm²).

Using our Tissuealign module, users can register the H&E with different stained sections (other brightfield or any IF) to transfer the tumor regions for further detailed analysis onto the aligned stains. This enables downstream analysis being limited to tumor-specific regions, also on stains where tumor detection is not possible by itself.

Trained on a diverse set of tissues and laboratories, the model is robust to staining variability across multiple organs. Users can easily adjust the minimum area of detected tumor or fine-tune the model to their own data, making the app adaptable to a wide range of research applications.

Figure 1

Figure 1

Analysis on breast cancer tissue

Figure 2

Figure 2

Analysis on melanoma tissue with exclusion of an artefact

Figure 3

Figure 3

Analysis on lung cancer tissue

Figure 4

Figure 4

Analysis on colon cancer tissue

Details

Quantitative Output variables

Output is the total tumor area in mm².

Related APPs

Success
Your message has been successfully sent!