Gorilla Troops Optimizer with Deep Learning-Based Thyroid Cancer Classification on Histopathological Images

The thyroid gland serves a vital role in regulating various body functions, namely energy expenditure, metabolism,
and organ function, such as the heart and brain. Thyroid cancer refers to a cancer of the thyroid gland and is a commonest endocrine cancer. A pathologist can detect thyroid carcinoma on the basis of the visual inspection of tissue samples prepared on microscopic slides. Machine learning (ML) is increasingly employed in the medical imaging fields and for pathological diagnosis of various diseases. A deep convolutional neural network (DCNN) is a kind of ML, such as a specific artificial neural network resembling the multi-layered human cognitive system. Various studies have examined the application of DCNN to assess pathological images. This paper introduces a novel Gorilla Troops Optimizer with Deep Learning Based Thyroid
Cancer Classification on Histopathological Images (GTODL-TCHI) model. The presented GTODL-TCHI model majorly
analyses the HIs for the identification and classification of thyroid disease. Initially, the GTODL-TCHI model applies the image
denoising procedure using the non-local mean filtering (NLMF) technique. In addition, the pre-processed images are then segmented using a fully convolutional network (FCN). Besides, the GTO algorithm with a densely connected network
(DenseNet121) method can be implied to produce feature vectors. Finally, the classification of features takes place using a stacked sparse autoencoder (SSAE) model. The performance validation of the GTODL-TCHI method can be tested using the HI dataset. The results stated the significant performance over the recent state of art DL models.