Abstract
Breast cancer is the most prevalent cancer in women worldwide. Early presentation, detection and prompt treatment limit morbidity and mortality due to breast cancer. Conventionally, breast cancer screening techniques and diagnosis have relied upon interpretation of radiologists and pathologists. However, advancement in artificial intelligence can lead to further enhancement in accuracy and efficiency of these diagnostic techniques, thereby, reducing incidence of morbidity and mortality.
AIM This systemic literature review is conducted to ascertain whether artificial intelligence (AI) can be used to complement existing breast cancer screening techniques. Objective of this review is to determine whether AI can enhance sensitivity of screening techniques, enable more accurate classification of benign and malignant tumors and improve assessment of response to neoadjuvant therapy
METHODS This systemic review is conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines. A comprehensive computer literature search in database of PubMed was performed using a combination of search terms: ‘Artificial intelligence’ OR ‘AI’ OR ‘Machine learning’ AND ‘Screening’ AND ‘Breast cancer’. 843 articles were identified through search in PubMed database. Following removal of 4 duplicate papers, titles and abstracts of 839 articles were reviewed. 115 articles with relevant titles and abstracts were analyzed. Following thorough analysis, 15 papers were included in this literature review.
RESULTS AI algorithms exhibited capability in classifying breast lesions and identification of malignancy in otherwise suspicious lesions across different imaging techniques. The integration and assistance of AI algorithms in interpretation of MRI, mammography and thermography has led to significant improvement in diagnostic accuracy and classification of breast lesions. AI complements radiologists and aids in improving performance, thereby, generating better results. AI has the capability to predict response to neoadjuvant chemotherapy in breast cancer patients, leading to safer, more effective and more cost-effective treatment for breast cancer patients.
CONCLUSIONS Artificial intelligence has the potential to revolutionize medicine in 21st century. Artificial intelligence has widespread potential in breast cancer screening. It can aid in improving radiologists’ ability to detect cancer on radiograms, classifying breast lesions, and predicting response to neoadjuvant therapy.
