• Mugdha Deokar
• Ojas Barpande

Thyroid Cancer is highly fatal and often diagnosis is time consuming due to the requirement of multiple tests and biopsies. In this project, this observation has been taken into consideration and a technique has been proposed that will present a more objective and accurate classification and would further offer a valuable second opinion to medical practitioners. A classifier is used that extracts features that quantify changes in the texture of the training images from both benign and malignant nodules. The resulting feature vectors are used to train the classifier. A VGG-16 model that is pre-trained on the ImageNet dataset is fine tuned using thyroid ultrasound images to classify malignant and benign lesions. This system is expected to be fast while being accurate at the same time in order to be useful in scenarios requiring speedy diagnosis and treatment.

Thyroid Malignancy, Deep Learning, Convolutional Neural Networks, Transfer Learning

"Thyroid Malignancy Detection", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.8, Issue 7, page no.g96-g100, July-2021, Available :http://www.jetir.org/papers/JETIR2107743.pdf

"Thyroid Malignancy Detection", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.8, Issue 7, page no. ppg96-g100, July-2021, Available at : http://www.jetir.org/papers/JETIR2107743.pdf