Description

Thyroid detection using Thermal Image Processing

Abstract

       The infrared (IR) thermal imaging could be utilized as a non invasive tool for detecting body disorders based on an abnormal temperature rise. In the present study, the IR thermography is employed for the detection of malignant thyroid tumors. Previous studies on the thyroid thermography have confirmed the higher temperature of thyroid tumors in comparison to the thyroid gland, which appears as hot spots and disturbs the symmetry of the thermogram. However, the thyroid thermography has clinical significance if thyroid cancer could be estimated by studying the IR thermal image.

Proposed System

       IR thermal images are captured from 18 human subjects including 10 healthy and 8 cancerous cases. The contrast between image components is increased by a cooling process during a dynamic thermal imaging. The captured thermal images undergo specific image processing algorithms for the image segmentation and the noise reduction. The study of the processed thermal images indicates a local temperature rise of 1–1.5 °C in front of the thyroid cancerous tumor compared to the surrounding healthy tissue. Parameters of the cancerous tumor are determined by the thermal classification of the processed thermal images.