IDENTIFYING FOLLICULAR THYROID CANCER USING THE YOLOV5 ALGORITHM: A COMPARISON USING FUZZY C-MEANS AND SINGULAR VALUE DECOMPOSITION

This investigation offers a comprehensive approach for the detection of Follicular Thyroid Cancer (FTC) using the YOLOv5 algorithm. The dataset utilized in this study consists of one thousand photos from the UCI dataset. To enhance the quality of the input photos, a Wavelet-based filter approach is applied during the pre-processing stage. The primary objective is to evaluate the performance of the proposed YOLOv5-based system and compare it with other well-known techniques such as Fuzzy C-Means and Singular Value Decomposition (SVD). With an F1 score of 95%, accuracy of 98%, precision of 97%, and recall of 97%, the suggested technique produces impressive results. These metrics demonstrate how well the algorithm can detect cases of Follicular Thyroid Cancer in medical photos. The YOLOV5 algorithm outperforms other current approaches, such as Fuzzy C-Means and SVD, in terms of detection accuracy and resilience, as demonstrated by the comparison study. The findings of this study advance the area of thyroid cancer detection strategies by offering a dependable and effective method that can assist medical professionals in making an accurate and timely diagnosis. The excellent accuracy, recall, and F1 score values of the proposed YOLOv5-based system indicate that it has potential to be clinically relevant in medical image analysis and pathology.