Prediction and Detection of Cancer Through Machine Learning: A review

• Published in: 2024 1st International Conference on Advances in Computing, Communication and Networking (ICAC2N) pp. 417 - 423
• Main Authors: Saxena, Nikita, Yadav, Sudeept Singh, Dujawara, Asmita, Kumar, Avneesh
• Format: Conference Proceeding
• Language: English
• Published: IEEE 16.12.2024
• Subjects: Dataset; Deep learning; Feature Selection; Forecasting; Machine Learning; Machine learning algorithms; Mortality; PRISM; Program processors; Reviews; Sensitivity; Surveys; Taxonomy; Tumors
• DOI: 10.1109/ICAC2N63387.2024.10895035

Expertise in the medical field is growing thanks to artificial intelligence. Free health data has led to the development of ways by experts to help in tumor detection and prediction. For those difficult illnesses, deep learning and machine learning models offer a trustworthy, quick, and efficient solution. Dissertations via Web of Science, EBSCO, and EMBASE that became available between 2009 and 2021 were selected using PRISMA criteria. Leveraging a useful discovery technique, the scientific publications for this investigation that employed AI-based learning algorithms for tumor forecasting were located. A collection of 185 studies demonstrate the substantial influence of neural network-based processors and conventional neural network-based taxonomies on carcinoma forecasting. The survey also looked at earlier studies and pointed out flaws in those studies. Forecasting rate, preciseness, specificity, sensitivity, rolling rating, recognition velocity, territory illustrated accuracy, recall, and F1-score were among the metrics utilized to gauge the results. The five scheduled examinations' remedies have been looked into. Considering the high likelihood of success for multiple of the treatments suggested in studies, the death rate from carcinoma has not dropped. Therefore, more investigation is needed to address the problems with carcinoma forecast.