Abstract
Pelvic bone cancer is a serious medical condition that often remains undetected in its early stages due to non-specific symptoms and the complex anatomical structure of the pelvic region. Common clinical symptoms include persistent pelvic pain, swelling, restricted movement, unexplained weight loss, and fatigue, which are frequently mistaken for musculoskeletal disorders. Factors such as genetic predisposition, previous radiation exposure, bone disorders, and long-term inflammation are considered significant contributors to the development of pelvic bone malignancies. Delayed diagnosis increases disease severity, highlighting the importance of early detection and public awareness. Magnetic Resonance Imaging (MRI) plays a vital role in visualising pelvic bone abnormalities due to its superior soft-tissue contrast. This study proposes an automated framework for pelvic bone cancer detection that integrates image filtering, region extraction, and clustering-based segmentation. During preprocessing, median filtering and Gaussian filtering are applied to MRI images to suppress noise, smooth intensity variations, and enhance structural visibility. This filtering stage improves image quality and supports accurate identification of abnormal tissue regions. A Region of Interest (ROI) extraction step then isolates tumour-suspected pelvic areas, reducing interference from surrounding tissues. The extracted ROIs are segmented using K-means clustering and Fuzzy C-Means (FCM) algorithms based on intensity and spatial characteristics. While K-means performs hard clustering, FCM enables soft classification through membership values, resulting in improved tumour boundary delineation in complex pelvic structures. Experimental results show that FCM outperforms K-means in handling overlapping tissue intensities. This automated, filtering-assisted approach can serve as a supportive diagnostic tool for radiologists. Moreover, the study emphasises the importance of early symptom recognition and timely medical consultation to reduce fear, increase awareness, and improve survival outcomes among individuals at risk of pelvic bone cancer.
Keywords