Introduction: Lung cancer is the most common and growing kind of cancer. Physicians must include genomes, proteomics, immunohistochemistry, and imaging data into patient therapy recommendations, along with histological, clinical, and demographic data. Deep learning/Machine Learning (ML) and convolutional neural networks have been employed to build medical AI in recent years, which cancan reduce subjectivity and improve the efficiency in treatment. Methods: This investigation makes use of a bibliometric strategy and knowledge mapping, using CiteSpace and R Biblioshiny to conduct a quantitative and visual analysis current picture of development of AI in lung cancer. In this particular research endeavour, evaluations concerning authorship, nations, institutions, reference articles, keywords, and reference journals were carried out. Results: Through the review, total of 1868 articles with contributions of a total of 9974 writers were analysed. There is a 25.3 % increase in the annual rate of scientific production. Frontiers in Oncology journal is determined to be the most globally cited journal with 83 articles and Jiying Wang and Yu wang have the highest h-Indexes, the top authors, are also the most influential ones. The most important keywords are “radiomics”, “convolutional neural network” and “feature selection”. The highest contributing country is held by China, which has an impressive 2103 publications; in second place is USA with 1858 publications. Conclusion: The Pulmonary nodule classification, Radiomics, circulating mirna biomarker, Random Forest vector model are the top clusters and themes with good Q and silhouette value. Thus, we recommend that anyone interested in the field can start with these topics.
Key words: Bibliometric, Scientometric, Lung cancer, Biblioshiny, CiteSpace, Artificial Intelligence, Adenocarcinoma
|
