<![CDATA[Background: Nucleus Pulposus Hernia (HNP) in the cervical region is a degenerative condition that can cause pressure on the spinal nerve. Magnetic Resonance Imaging (MRI) is the primary imaging modality for the diagnosis of HNP, but it has a long acquisition time limitation, which risks causing artifacts due to patient movement. The use of GRAPPA parallel imaging techniques can speed up acquisition, but lower the signal-to-noise ratio (SNR). This study aims to evaluate the effect of wavelet denoising filters on improving the quality of images and cervical MRI information of T2 TSE GRAPPA sequences in HNP patients.Methods: The study used a pre-experimental was conducted on 16 cervical HNP patients at Budi Rahayu Hospital Pekalongan. Acquisitions are carried out with acceleration factors 2, 3, and 4. Subjective analysis of image quality by three radiologists was performed on five anatomical structures (disc, corpus, spinal cord, CSF, and spinesus process). Data were assessed using the Cohen's Kappa test for fit between assessors and the Friedman test to compare SNR and CNR values between acceleration factors.Result: The application of wavelet denoising filters significantly increased SNR and CNR values (p = 0.0174 and p = 0.0175). R-factor 3 produced the best overall image quality with the highest Kappa value of 0.681 in the R1*R2 grader pair, indicating substantial conformity. The highest mean rank for visual quality was found in the spinal cord (R-factor 2 = 1.17), while the CSF, corpus, and disc showed the best quality in R-factor 3. This technique has been proven to retain critical diagnostic information with more efficient acquisition times.Conclusions: The combination of GRAPPA with wavelet denoising filters on T2 T2 sequence cervical MRI GRAPPA significantly improved image quality, especially at R-factor 3. These results show that this method is effective in reducing noise and preserving anatomical details, thereby improving the accuracy of cervical HNP diagnosis and patient comfort during the examination.]]>
Copyrights © 2025
