Computer Intelligence in Detection of Malignant or Premalignant Oral Lesions: The Story So Far

Abstract

Cancer is a highly morbid disease affecting 20% of the world population. The third most common type of cancer is oral cancer which develops in the tissues of the mouth and neck and has different survival rates in men and women. Oral squamous cell carcinomas (OSCC), are the most common type of oral cancer and its early detection is key to longevity, successful outcome of treatment, and better quality of life. Computational intelligence (CI) may be a good tool in the detection of lesions in the early stage. The studies based on CI depend on images obtained through various light or resonance-based detection techniques as well as clinical images, while only one study has shown the accuracy of CI in detecting the gene expression patterns. Besides, there are several studies that have shown the efficacy of machine learning and deep learning in the field of radiology and spectroscopy. In totality, CI may be described as an upcoming tool that holds the promising potential to be used for the early diagnosis of malignant or premalignant lesions. The present review work is an attempt to propose a patient workup plan based on the concepts and models designed so far for the early detection of cancerous lesions. Thus, it may then help identify patients for closer follow-up, suggestive behavioral, and lifestyle modification to reduce the incidence of OSCC.