Automated Detection of Malignant Tissue in Specimens of Thyroid Gland Using Spectral Imaging Technology

Malignancy is the second most common reason of death in most countries. Diagnosis of thyroid cancers is one of the most difficult medical problems. Therefore, it important to use diagnosing methods that bring about quick results, are less costly, and prevent errors based on personal factors. An example of such new methods is the use of automated spectral imaging, which gives more information about the structure of tissue, through the formation of fingerprint histological diagnosis used in the automated machine. This study included 25 cancer cases and 26 specimens of non-malignant thyroid or normal. The credibility of method is reflected in the high rates of correct diagnosis in cancer samples (TP) = 92% and the negative result in sections of non-malignant (FN) = 96%. This method represents the use of automation technologies and artificial intelligence to facilitate and speed up the diagnosis of cancers and can be used to diagnose other organs and other tissues.