Espectroscopia de infravermelho com transformada de Fourier com Refletância Total Atenuada (ATR-FTIR) e quimiometria como adjuvantes no diagnóstico de câncer de tireoide

• Autor(a) principal: Cruz, Ingrid Gabriela Bezerra de Lima
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFPB
• Texto Completo: https://repositorio.ufpb.br/jspui/handle/123456789/25128
• topic: Câncer de tireoide; Espectroscopia de infravermelho; DD-SIMCA; Classificação; SIMCA; Thyroid cancer; Infrared spectroscopy; Classification; CNPQ::CIENCIAS EXATAS E DA TERRA::QUIMICA

Thyroid cancer is the most prevalent malignant neoplasm affected in the head and neck region in recent years. According to US projections, it will be the fourth most frequent type cancer in society by 2030. The lack of professionals specialized in distinguishing normal and cancerous thyroid tissues in more remote or poorer regions makes it difficult to offer diagnosis. Given this, offering a diagnostic or screening method that can be performed by a more readily available professional can reduce the deleterious impact that this shortage causes. In this sense, infrared spectroscopy in conjunction with chemometrics may be a viable alternative for the development of such a diagnostic method. Thus, this study aimed to develop a method employing attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) in conjunction with chemometric classification techniques, the soft independent modelling by class analogy (SIMCA) and data driven SIMCA (DD-SIMCA), to aid in the diagnosis of thyroid cancer. The study employed fine needle aspiration (FNA) samples from 49 patients (n = 38), with 27 samples diagnosed as benign, 11 samples diagnosed as thyroid cancer and 11 samples diagnosed as undefined. With the models developed, it was possible to evaluate the prediction of 11 undefined samples, to predict the similarity of the pathological classification with the proposed model, values of global SENS 61% and global ESPEC 89% were obtained for the test set and undefined. The proposed methodology represents a promising tool in the diagnosis of thyroid cancer, being possible to conclude that spectroscopy and chemometrics can distinguish between healthy and cancerous thyroid tissues.