Prediction of Cervical Cancer Using Machine Learning
Ashish Kumar, Revant Singh Rai, Mehdi Gheisari · 2021
Cancer remains one of the most prevalent diseases humans are confronting. Even after taking huge strides in medicine, humans have not been able to properly treat cancer. Cancer of the cervix is counted among the deadliest diseases faced by women of the world. It grows in the lower lining of the cervix. Cervical cancer is the fourth most prevalent form of cancer in women around the globe; it propagates in 6% of patients who are diagnosed with cancer around the world. Less than half of the patients screened get flagged at an early stage. The low success in finding cancer may be due to the lack of awareness and vigilance of the disease. Deficiency of appropriate medical apparatus and infrastructure further deteriorates the problem, and this can lead to huge complications. In a huge country like India, the five-year survival rate goes as low as 50%. Even if the cancer is eradicated efficaciously, a minority of cancer cells can remain hidden. This can lead to the recurrence of cancer. Patients showing symptoms of recurrence have a low survival rate, as low as 22%. As patients show diverse growth dynamics, researchers are trying to find the root cause of the recurrence of cancer. They are trying to find neoteric and innovative ways to identify the origin of recurrence before cancer once again wreaks havoc in people’s lives. Researchers are now looking to facilitate their research by using high-level machine learning/artificial intelligence models. Organizations like BioGPS and University of California, Irvine (UCI) machine learning repository are keeping huge datasets, which can be used for research and training models. Researchers are working with SVM, C5.0, J48 graft, artificial neural network, Naive Bayes, and extreme learning machine to assist physicians with flagging, finding, and pointing out problems and discrepancies of cancer at early stages. By devising this methodology, we can find out which approach is the most effective and accurate way of helping future female patients through early detection and decreasing the chance of relapse.