A Narrative Review on Large AI Models in Lung Cancer Screening, Diagnosis, and Treatment Planning

2025-06-08

AI Summary

Early lung cancer screening and precision oncology face challenges of high false-positive rates and suboptimal clinical decision-making efficiency. This study systematically reviews state-of-the-art applications of large foundation models in pulmonary nodule detection, driver gene prediction, multi-omics prognostic modeling, and therapeutic planning. We propose, for the first time, a taxonomy of medical large models based on architectural paradigms—modality-specific encoders, encoder-decoder frameworks, and joint encoders—and establish a standardized “clinical task–model–dataset” mapping framework to advance AI clinical translation. Leveraging multimodal models (e.g., CLIP, BLIP, BioViL-T) and benchmark datasets (e.g., LIDC-IDRI, NLST), we validate performance via multimodal alignment, prompt-based fine-tuning, and interpretability analysis. Several models achieve radiologist-level diagnostic accuracy; some have entered clinical pilot deployment, significantly improving diagnostic precision, personalized treatment decision efficiency, and cross-modal integration capability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Large Vision Models
• Natural Language Processing: (Large) Language Models

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Lung cancer remains one of the most prevalent and fatal diseases worldwide, demanding accurate and timely diagnosis and treatment. Recent advancements in large AI models have significantly enhanced medical image understanding and clinical decision-making. This review systematically surveys the state-of-the-art in applying large AI models to lung cancer screening, diagnosis, prognosis, and treatment. We categorize existing models into modality-specific encoders, encoder-decoder frameworks, and joint encoder architectures, highlighting key examples such as CLIP, BLIP, Flamingo, BioViL-T, and GLoRIA. We further examine their performance in multimodal learning tasks using benchmark datasets like LIDC-IDRI, NLST, and MIMIC-CXR. Applications span pulmonary nodule detection, gene mutation prediction, multi-omics integration, and personalized treatment planning, with emerging evidence of clinical deployment and validation. Finally, we discuss current limitations in generalizability, interpretability, and regulatory compliance, proposing future directions for building scalable, explainable, and clinically integrated AI systems. Our review underscores the transformative potential of large AI models to personalize and optimize lung cancer care.

Problem

• Applying large AI models to improve lung cancer screening and diagnosis
• Evaluating AI performance in multimodal learning for lung cancer tasks
• Addressing limitations in AI generalizability and clinical integration

Innovation

• Large AI models enhance medical image understanding
• Modality-specific encoders improve diagnosis accuracy
• Multimodal learning optimizes treatment planning