Open Access

ARTICLE

Hyperparameter Tuning Bidirectional Gated Recurrent Unit Model for Oral Cancer Classification

by K. Shankar¹, E. Laxmi Lydia², Sachin Kumar^1,*, Ali S. Abosinne³, Ahmed alkhayyat⁴, A. H. Abbas⁵, Sarmad Nozad Mahmood⁶

1 Department of Computer Science, South Ural State University, Chelyabinsk, 454080, Russia
2 Department of Computer Science and Engineering, Vignan’s Institute of Information Technology, Visakhapatnam, 530049, India
3 Computer Technical Engineering Department, College of Technical Engineering, the Islamic University, Najaf, 54001, Iraq
4 College of Technical Engineering, the Islamic University, Najaf, Iraq
5 College of Information Technology, Imam Ja’afar Al-Sadiq University, Al-Muthanna, 66002, Iraq
6 Computer Technology Engineering, College of Engineering Technology, Al-Kitab University, Iraq

* Corresponding Author: Sachin Kumar. Email:

Computers, Materials & Continua 2022, 73(3), 4541-4557. https://doi.org/10.32604/cmc.2022.031247

Received 13 April 2022; Accepted 25 May 2022; Issue published 28 July 2022

Abstract

Oral Squamous Cell Carcinoma (OSCC) is a type of Head and Neck Squamous Cell Carcinoma (HNSCC) and it should be diagnosed at early stages to accomplish efficient treatment, increase the survival rate, and reduce death rate. Histopathological imaging is a wide-spread standard used for OSCC detection. However, it is a cumbersome process and demands expert’s knowledge. So, there is a need exists for automated detection of OSCC using Artificial Intelligence (AI) and Computer Vision (CV) technologies. In this background, the current research article introduces Improved Slime Mould Algorithm with Artificial Intelligence Driven Oral Cancer Classification (ISMA-AIOCC) model on Histopathological images (HIs). The presented ISMA-AIOCC model is aimed at identification and categorization of oral cancer using HIs. At the initial stage, linear smoothing filter is applied to eradicate the noise from images. Besides, MobileNet model is employed to generate a useful set of feature vectors. Then, Bidirectional Gated Recurrent Unit (BGRU) model is exploited for classification process. At the end, ISMA algorithm is utilized to fine tune the parameters involved in BGRU model. Moreover, ISMA algorithm is created by integrating traditional SMA and Chaotic Oppositional Based Learning (COBL). The proposed ISMA-AIOCC model was validated for performance using benchmark dataset and the results pointed out the supremacy of ISMA-AIOCC model over other recent approaches.

---

Keywords

---