Thyroid problems are a common endocrine disorder affecting a significant portion of the population worldwide. Early detection and accurate diagnosis of thyroid abnormalities are crucial for effective treatment and management of patients. The detection process involves collecting blood samples to know the percentage of hormones (TSH, T3 and T4). In recent years, medical image processing techniques combined with Machine Learning (ML) algorithms have emerged as a promising approach to improve thyroid problem detection and diagnosis. The use of Machine Learning (ML) in thyroid detection processing has received huge interest in latest years due to its attainable to enhance the accuracy and effectiveness of thyroid ailment diagnosis. The thyroid gland performs an essential position in regulating metabolism, and abnormalities in its shape or feature can lead to a number of thyroid disorders. ML algorithms utilized to clinical imaging data, such as ultrasound or thyroid scintigraphy scans, provide the possibility to beautify the detection and classification of thyroid abnormalities.