Prediction of Sleep Patterns in Young Breast Cancer Patients Using Machine Learning Algorithm

Abstract

This research utilized machine learning techniques to provide a solution to young breast cancer patients in a form of a web application through which the patients can assess their sleep quality and know their sleep patterns at earlier stages of treatment. Before proceeding with this research, a pilot study was conducted by contacting 182 oncologists pan India among which 42 responded to the questionnaire circulated to know post- treatment symptoms burden, especially sleep disorders, its impact and need of a predictive tool that help knowing the symptoms at earlier stage. 88.9% oncologists responded positively regarding the need of a computational tool that can help in improving the sleep quality of breast cancer patients. Literature review of 121 worldwide publications was conducted. In total 930 patients aged between 18-49 years were included in study. Results indicated, the factors became more layered with increase of treatment combinations. The proposed algorithm SwSo SP (Sleep with Serenity and Observance Sleep Prediction) was compared with other machine learning algorithms i.e. CNN, KNN, SVM. The performance metrics indicated that SwSo SP model outperforms with highest accuracy i.e. 96.24%, 98.13% precision, 96.77% F1 score, 95.45% recall, and 23.12 RMSE. In addition to this, Phase III discussed the development of web application and its features through graphical representation indicating prediction of sleep patterns. The study emphasizes the importance of sleep and need for its continuous assessment as it highly affects their well-being and quality of life. I envision that this work will serve as a stepping stone for further innovations in predictive healthcare, and palliative care. newline

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced