2023 World Conference on Lung Cancer (Posters)-EP06.01. Deep Learning Based Quality Assessment of DAPI Stained Nuclei Images Obtained from Routine FISH Test for Lung Cancer

EP06.01. Deep Learning Based Quality Assessment of DAPI Stained Nuclei Images Obtained from Routine FISH Test for Lung Cancer - PDF(Abstract)

Back to course

Pdf Summary

Fluorescent in situ hybridization (FISH) is a valuable tool for identifying genetic abnormalities in lung cancer patients. DAPI staining is often used to visualize cell nuclei in FISH experiments. However, the variability of tumor cell quality and the presence of non-tumor cells can make it challenging to accurately interpret the images. In this study, researchers developed a deep learning algorithm to help pathologists identify high-quality lung cancer DAPI images for FISH interpretation.<br /><br />The researchers created a dataset of 35,769 DAPI images from 1,788 cases of non-small cell lung cancer. Two expert pathologists manually classified the images into different quality classes based on tumor and storm composition, as well as image quality related to protein digestion and degradation. These classes were then grouped into a binary classification model.<br /><br />The researchers combined a CNN-based feature extraction layer with an attention-based multiple instance learning to develop the binary classification model. The model achieved a recall of 0.871, which was higher than the recall of the attention method used for comparison.<br /><br />The results indicate that automated deep learning methods can achieve high accuracy on diverse FISH platforms. This has the potential to enhance clinical tools by reducing false positive and false negative screening FISH results.<br /><br />In summary, this study developed a deep learning algorithm for quality assessment of DAPI stained nuclei images obtained from routine FISH tests for lung cancer. The algorithm demonstrated high accuracy and holds promise for improving the interpretation of FISH images by pathologists.

Asset Subtitle

Tae-Jung Kim

Meta Tag

Speaker Tae-Jung Kim

Topic Pathology & Biomarkers: Artificial Intelligence in Pathology

Keywords

Fluorescent in situ hybridization

FISH

genetic abnormalities

lung cancer patients

DAPI staining

cell nuclei

deep learning algorithm

pathologists

lung cancer DAPI images

binary classification model