A new computational model helps in the early detection of cervical cancer

A new computational model could improve the diagnosis of cervical dysplasia, or abnormal growth of cells on the surface of the uterus, and could be used for early detection of cervical cancer.

Correct identification and classification are important for the diagnosis and treatment of cervical dysplasia.

Scientists at the Guwahati-based Institute of Advanced Science and Technology (IASST), an autonomous institution under the Department of Science and Technology (DST), have begun developing an almost unparalleled model. The shorter the calculation time, the higher the accuracy.

Doctor  Libby B. Mahant and her team experimented with different color models, transformation methods, feature representation systems, and classification methods to build a robust machine learning (ML) framework.   The purpose of this comprehensive analysis and testing is to determine the appropriate combination for the diagnosis of cervical dysplasia.

The performance of the model was tested on two datasets: one collected from medical centers in India and the other publicly available dataset.

Using non-sampling contour transform (NSCT) image processing method and YCbCr color model (using image color representation method), the average accuracy of the new model reached 98.02%.

This innovative model could improve the diagnosis of cervical dysplasia and provide healthcare professionals with more accurate tools to increase diagnostic accuracy and improve treatment outcomes.

Reference

Author: MedBlog18 Blog Team

Views: 13