Deep Learning for Lung Cancer Diagnosis, Prognosis and Prediction Using Histological and Cytological Images: A...

Area of Science:

• Oncology
• Digital Pathology
• Computational Biology

Background:

• Lung cancer is a leading cause of cancer death globally, particularly in smokers.
• Accurate diagnosis relies on histology and molecular data for personalized treatment.
• Pathologists face challenges classifying lung cancer from limited biopsy/cytology samples.

Purpose of the Study:

• To systematically review Artificial Intelligence (AI) approaches for lung cancer interpretation using histological and cytological images.
• To assess the potential of Deep Learning (DL) in assisting pathologists in routine practice within Digital Pathology.

Main Methods:

• Systematic review of AI-based methods applied to lung cancer histology and cytology images.
• Analysis of studies focusing on classification, pattern determination, prognosis, and predictive marker estimation.

Main Results:

• Most AI studies focus on differentiating major lung cancer types: adenocarcinoma, squamous cell carcinoma, and small cell carcinoma.
• Algorithms have been developed for predicting lung adenocarcinoma patterns, prognosis, mutational status, and PD-L1 expression.

Conclusions:

• AI, particularly DL, shows significant potential to enhance lung cancer diagnosis and analysis from digital pathology images.
• AI tools can assist pathologists in routine practice, especially with limited diagnostic material.