Artificial intelligence, machine learning, and radiomics in lung cancer classification

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Lung cancer is a highly heterogeneous disease that presents significant challenges in accurate diagnosis and classification due to its diverse histological and molecular characteristics. Traditional diagnostic methods, while valuable, are often limited by invasiveness, subjectivity, and an inability to fully capture tumor complexity. Recent advancements in artificial intelligence (AI), machine learning, and radiomics have transformed the field, offering enhanced precision, efficiency, and objectivity in lung cancer classification. These technologies enable detailed analyses of imaging data, histopathological findings, and molecular profiles, facilitating improved subtype identification, outcome prediction, and personalized treatment strategies. Cytopathology remains a cornerstone of lung cancer diagnostics, particularly for small biopsies and cytological samples, which are often the only materials available in advanced stages. The integration of AI-driven methods into cytopathology and radiomics workflows has shown substantial potential to overcome the limitations of traditional approaches, reduce interobserver variability, and accelerate the diagnostic process. This review underscores the transformative role of AI and radiomics in lung cancer management, highlighting their synergy in advancing precision oncology. As ongoing research continues to refine these methodologies, the future of lung cancer care is poised for significant advancements, offering improved diagnostic accuracy, personalized therapies, and better patient outcomes.

This publication has 34 references indexed in Scilit: