Skripsi
ANALISIS PENGGUNAAN ARTIFICIAL INTELEGENCE (AI) DALAM PEMERIKSAAN MAGNETIC RESONANCE CHOLANGIOPANCREATOGRAPHY (MRCP) DI RUMAH SAKIT UMUM PUSAT PERSAHABATAN
In routine clinical practice, Magnetic Resonance Cholangiopancreatography (MRCP) is a non-invasive imaging modality based on Magnetic Resonance Imaging (MRI) that is used to evaluate the biliary system and pancreas. However, MRCP examinations are limited by relatively long acquisition times and high sensitivity to motion artifacts, which may adversely affect the image quality. For this reason, advances in Artificial Intelligence (AI), particularly deep learning-based approaches, have the potential to improve image quality while simultaneously accelerating MRI acquisition.rnThis research aims to analyse the impact of Artificial Intelligence (AI)–assisted MRCP examinations on the image quality and examination time efficiency undertaken at Persahabatan Central General Hospital. This research applied quantitative method with comparative approach which was employed by comparing the MRCP results acquired using conventional T2 Turbo Spin Echo (TSE) sequences without AI and AI-based T2 TSE sequences. Meanwhile, image quality assessment was based on anatomical delineation and noise level, evaluated through among the other things questionnaires completed by qualified respondents, and subsequently analysed using statistical methods.rnAs the results, this research demonstrated that the use of AI in MRCP examinations improved anatomical clarity and reduced noise levels compared to the examinations performed without AI. In addition, AI implementation was shown to significantly reduce scanning time without compromising the diagnostic quality of the images.rnIn conclusion, this research suggests that the application of Artificial Intelligence (AI) in MRCP examinations provides many tangible benefits in improving the image quality and examination efficiency, thereby supporting the optimization of radiology services and enhancing the accuracy of clinical diagnosis.
Tidak ada salinan data
Tidak tersedia versi lain