What type of study is this?

This is a Literature Review study.

October 3, 2025

Advancing cancer care with artificial intelligence: early detection initiatives

Key Points

DL algorithms achieve high accuracy in cancer detection, often matching expert evaluations, enhancing early diagnosis.
AI models like Sybil and CAD colonoscopy improve cancer risk prediction and adenoma detection rates effectively.
Comprehensive literature review utilized databases such as PubMed and Google Scholar to evaluate AI's role in oncology.
Future research should address ethical considerations and model interpretability for better clinical integration.

Abstract

Background Artificial intelligence (AI) is revolutionizing early cancer detection, with machine learning (ML) and deep learning (DL) at the forefront. While ML requires human guidance, DL autonomously learns from data using neural networks, excelling in image and speech recognition, pattern detection, and natural language processing, making it highly valuable in oncology. The aim of this research article is to review the application of Artificial Intelligence, particularly DL, in the early detection of cancer, evaluating various models and their effectiveness in improving diagnostic accuracy and prediction. Methods A diversified literature review was carried out using PubMed, Google Scholar, and ScienceDirect. Keywords included "Artificial Intelligence," "Machine Learning," "Deep Learning," and "Cancer Detection." Inclusion criteria encompassed articles in English published in the last twenty years, focused on original research. Excluded were non-research articles and those with inaccessible full texts. Results AI models under investigation include: Sybil: Accurately predicts lung cancer risk from a single low-dose CT scan. CNN-based models: Effective in classifying cancers from endoscopy, radiology, and histopathology images. CAD colonoscopy: Improves adenoma detection rates in real-time colonoscopy. Paige Prostate Alpha: Detects prostate cancer in whole slide images. CUP-AI-Dx: Recognizes the primary tissue of origin in cancers of unknown primary using RNA-based data. M3Net: Integrates CT images and biomarker data for cancer risk prediction. GANs: Enhance image-based diagnostics through denoising and completion. Conclusion DL algorithms demonstrate high accuracy in cancer detection, often matching or surpassing expert evaluations. Despite challenges in model robustness, multimodal data integration, and ethical considerations, AI holds significant promise in oncology. Future research should focus on enhancing model interpretability and clinical integration.

Mark Helpful

Bookmark

Relay