Detection and Classification of Upper Aero Digestive Tract Tumour using Deep Convolutional Neural Network

Abstract

Upper Aero Digestive Tract (UADT) tumour is a type of oral and nasal cavity tumour that causes death in any gender. WHO statistics reports that, around 6, 75, 000 new cases were diagnosed as cancer positive and 3, 30, 000 deaths are attained every year. Nasopharynx is one of the deadliest cancers when compared to other life-threatening cancers. In the year 2019 cancer observatory database reported that head and neck have 5.2 % cancer positive cases and 5.4 % death rate throughout the world. American oncology journal called Healio reports that the oral and nasal cavity cancers are diagnosed positive for 1, 10, 000 people every year. Early detection and diagnosis of cancer will help in reducing the death rates. Diagnosing dysplastic tumours using manual microscopy procedures from biopsy tissue samples will be a tedious process due to morphological characteristic loss which may lead to misclassification. Screening the metastatic tumours with less morphological characteristics is more complex. Analysing the malignant tumour tissue samples with less morphological characteristics may lead to higher risk of patient s survival newline