OPTIMIZATION OF TOMOGRAPHIC IMAGE QUALITY USING AN ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM (ANFIS)

Abstract

Computed tomography (CT) is an advanced imaging modality widely employed in clinical practice for diagnostic and therapeutic planning purposes. However, due to its reliance on ionizing radiation, CT imaging may increase the long-term risk of malignancy, particularly among vulnerable populations such as pediatric patients, pregnant women, and individuals with chronic illnesses. Consequently, minimizing radiation exposure constitutes a fundamental principle in CT imaging and other radiation-based diagnostic applications. Nevertheless, reducing radiation dose often results in degraded image quality, thereby complicating disease detection and adversely affecting diagnostic accuracy in clinical practice. Although conventional filtering techniques are commonly utilized to enhance image quality, the associated loss of fine structural details highlights the need for more robust and efficient approaches. In this context, an Adaptive Neuro-Fuzzy Inference System (ANFIS) unit offers a promising solution by leveraging learning-based algorithms to suppress visual artifacts while preserving critical image information. Moreover, advanced implementations of such methodologies have the potential to reconstruct higher-quality images from low-dose acquisitions, thereby enabling reduced radiation exposure without compromising diagnostic performance. This advancement holds substantial significance for contemporary diagnostic clinical applications.