Indian scientists develop software to detect cervical cancer
Now open: Certificate Course in Management of Covid-19 by Govt. Of Gujarat and PlexusMDKnow more...Now open: Certificate Course in Management of Covid-19 by Govt. Of Gujarat and PlexusMDKnow more...
Researchers from the Institute of Advanced Study in Science and Technology (IASST), Guwahati have designed a robust fully automated software-based solution using Pap smear images to detect abnormal growth or development of tissues in the cervix.

The current work has been recently published in the journal ‘Tissue and Cell’.

Pap Smear is the most popular technique for early diagnosis of cervical cancer. Diagnosis is done by critically analyzing the slides prepared from the smear collections, under a microscope.

The researchers put together a database of two types of indigenous Pap smear images databases, one each for conventional and Liquid Based Cytology (LBC) and used it to create algorithms with the images of different qualities. The algorithms helped develop robust software to detect cervical cancer.

The team then adopted two AI approaches, Machine Learning and Deep Learning. Applying the first approach on the conventional dataset, the group reported a novel algorithm for automatic segmentation of cervical cells, removing debris like inflammatory cells and red blood cells from a whole slide image, including improvement towards dealing with poor stained images which are some of the major challenges faced by researchers.

The group achieved 96.5% accuracy using an ensemble machine learning algorithm where shape, color, and texture features of the Pap smear images were analyzed for automated classification of cervical dysplasia.

Source: https://twitter.com/IndiaDST/status/1273935310263447552?s=20
Like
Comment
Share