Digital Biomarkers in Head and Neck Pathologies
Research type
Research Study
Full title
Digital Biomarkers in Head and Neck Pathologies
IRAS ID
252094
Contact name
Syed Ali Khurram
Contact email
Duration of Study in the UK
4 years, 11 months, 28 days
Research summary
Head and neck cancers (HNC) are among the top ten most common cancers in the world with increasing frequency and poor survival. Diagnosis involves subjective tissue analysis by pathologists under a microscope with minimal objective prognostic information provided. Human Papilloma Virus (HPV) associated as well as salivary gland cancers can present in patients who dont smoke or drink alcohol. Diagnosis of these requires special tests with an associated delay and cost. In addition, when HNC spread to neck lymph nodes, patient survival gets even worse. The analysis of these 'distant' tumour deposits requires a significant amount of time and effort. Even then, subjectivity remains in deciding on HPV infectivity, salivary tumour sub-types and presence of tumour in lymph nodes.
Artificial intelligence (AI) is being increasingly used these days in and apps such as Alexa, Siri, as well as face unlock functions in phones. AI has been shown to be quite helpful in detecting tissue patterns to aid pathologists. Studies have even shown AI to be more accurate and consistent at recognising these patterns than experienced pathologists after appropriate training. In addition, AI can quantify things within tissues and even recognise features and patterns that are not seen by the naked eye. However, their use in diagnosis of head and neck cancers remains largely unexplored.This study proposes to explore the use of AI and determine whether it can be used to identify HPV infection, salivary gland tumour sub-types and behaviour and distant tumour deposits in lymph nodes. If successful, this will be an extremely helpful aid to pathologists improving efficiency, saving a significant amount of money, time and resources whilst providing an objective, consistent and reliable measure of features and risk prediction about future behaviour.
REC name
West of Scotland REC 1
REC reference
20/WS/0017
Date of REC Opinion
27 Jan 2020
REC opinion
Favourable Opinion