Enhancing diagnostic precision and accuracy in invasive lobular carcinoma through machine learning approaches

Abstract

Invasive Lobular Carcinoma (ILC) is a type of breast cancer that forms in the lobules of the breast and is characterized by small, non-cohesive cells that invade surrounding tissues in a unique pattern. Invasive Lobular carcinoma mostly affects women compared to men. Various techniques are available to detect the presence of ILC, like mammography, Ultrasound, and MRI. Invasive lobular carcinoma is not present in a mass, making it difficult to detect ILC in some imaging techniques. Machine learning (ML) techniques are being used to improve the prediction and diagnosis of ILC. It involves data collection from electronic health records, imaging studies, and genomic data from Kaggle, and using different models, such as supervised learning and unsupervised learning, to predict ILC. In this current study, various algorithms have been used to predict and improve the accuracy and precision level of ILC diagnosis. Results found that Elastic Net and Logistic Regression have shown higher accuracy. ML is very useful for radiologists, oncologists, and patients in early-stage prediction of ILC, which is helpful in personalizing treatment plans.