Article Preview
TopIntroduction
Thyroid nodules are common in the general population with a prevalence of 20-60% (Dean & Gharib, 2008) and can be either malignant or benign. The etiologies of thyroid nodules are a simple overgrowth of normal thyroid tissue, inflammation, or tumor. Thyroid cancer is one of the most common types of cancer in the endocrine system. It is the fifth most common cancer of women worldwide and the fourth most common cancer of women in Thailand (Pellegritit et al., 2013). Recent research showed that the annual incidence of thyroid cancer has gradually increased and now accounts for 1,500 new patients per year in Thailand (Tangjaturonrasme et al., 2018). Papillary thyroid carcinoma, the predominant subtype of thyroid cancer showed an increase in incidence in Thailand in a decade. (Bychkov, 2017; Bychkov, et al., 2017).
Nowadays, there are many diagnostic imaging modalities to detect thyroid nodules, including ultrasonography, computed tomography, magnetic resonance imaging, positron emission tomography, and scintigraphy. In clinical practice, thyroid ultrasonography is mainly used for both the detection and characterization of thyroid nodules. It is preferable due to noninvasiveness, convenience, no radiation exposure, and relatively low price, and intervention procedures can be performed concurrently. However, ultrasonography has some limitations to differentiate benign thyroid nodules from malignant ones due to the complex structures of thyroid nodules. Consequently, thyroid ultrasonography remains highly subjective and depends on physicians' experience, which causes a greater risk of misdiagnosing cancer and increases the number of FNA biopsies.
In the healthcare field, novel technologies have been developed in many countries for supporting a diagnosis such as medical records, medical text analysis, interprofessional team collaboration, and AI. These electronic medical records (EMRs) are useful for data sharing among medical departments (Taewijit & Theeramunkong, 2021). Medical text analysis uses knowledge management for disease prediction (Menaouer et al., 2020). The systems approach Interprofessional Team Collaboration (IPC) enhances outcomes of healthcare services, as well as improves the safety and quality of healthcare setups (Matsushita, et al. 2021). AI is used for diagnostic and therapeutic purposes in medical imaging. AI has shown impressive accuracy and sensitivity in the identification of imaging abnormality and tissue characterization. Thus, AI is expected to play an essential role in assisting radiologists in characterizing thyroid nodules. This can reduce errors caused by subjective factors, assist the diagnostic performance in avoiding unnecessary FNA biopsies, benefit further treatment plans for patients, and reduce healthcare costs.