肺癌
人工智能
无线电技术
深度学习
肺癌筛查
癌症
医学
机器学习
医学影像学
精密医学
医学物理学
计算机科学
肿瘤科
病理
内科学
作者
Pranali Santhoshini Pachika,Srijan Valasapalli,Phuong Ngo,Goetz Kloecker
标识
DOI:10.1089/aipo.2023.0002
摘要
Lung cancer is the leading cause of cancer-related mortality for both men and women, and it is one of the malignancies for which treatment is highly individualized. In all stages of lung cancer, including screening, diagnosis, therapy selection, prognosis, and surveillance, artificial intelligence (AI) components such as machine learning and deep learning can be applied. Radiomics (quantitative mathematical analysis of images) uses data-characterization algorithms to extract many diagnostic details from medical images. Numerous AI-driven radiomic models have been developed to differentiate benign from malignant lung nodules. New deep learning AI modules such as Sybil are even able to detect the future cancer risk based on a low-dose computed tomography. Numerous prediction models based on AI have been used to estimate the response to targeted therapy and immunotherapy. Al can also be used for active surveillance, and models for predicting the high-risk characteristics of recurrence have been developed. Some of the deep learning modules can detect the presence of molecular alterations and tumor histology based on tumor images. AI can be used in the field of genomics to discover new biomarkers that can help in prognostication as well as the development of new targeted therapies. Although AI can be used at every stage of lung cancer treatment, from diagnosis to survivorship, its lack of standardization is its greatest drawback. The intent of this review article is to provide a comprehensive overview of AI as regards the rapidly evolving field of lung cancer, encompassing all stages from screening to treatment, to highlight the latest advancements and their potential to revolutionize the management of lung cancer. We discuss the various AI models that are being utilized for lung cancer, as well as the potential future advancements. In addition, we highlight a few of the challenges associated with using these models in daily practice.
科研通智能强力驱动
Strongly Powered by AbleSci AI