Application of DLIR Algorithm in Improving Abdominal CT Image Quality: A Case Study on Hepatic Cyst Diagnosis

Research Background :Abdominal CT is an important means for the diagnosis and treatment of abdominal diseases. However, affected by complex organ structures and movements, image quality is often unsatisfactory. Although the Revolution CT spectral imaging system has made progress, noise and artifacts still exist. The DLIR algorithm is superior to traditional methods in noise reduction and detail preservation and has been applied, but research is limited and lacks systematic verification. This study explores its role and potential in optimizing abdominal CT image quality in this system to assist clinical diagnosis and treatment. Objective : This study aims to compare the image quality of abdominal computed tomography (CT) under different reconstruction algorithms under conventional dose and contrast agent conditions, and evaluate the application value of the Deep Learning Image Reconstruction (DLIR) algorithm in the diagnosis of hepatic cysts. Methods : A total of 100 patients with suspected hepatic cysts were randomly divided into an experimental group and a control group. The control group underwent magnetic resonance imaging (MRI), while the experimental group underwent CT examination. After verifying the diagnostic validity, the images of the experimental group were optimized using DLIR, Iterative Reconstruction (IR), and Adaptive Statistical Iterative Reconstruction (ASIR). DLIR was divided into three reconstruction levels: low (DLIR-L), medium (DLIR-M), and high (DLIR-H), while ASIR was set to 30% ASIR-V and 70% ASIR-V. The Contrast Noise Ratio (CNR) and Signal Noise Ratio (SNR) of different reconstruction algorithms were compared. Results : The sensitivity, positive predictive value, and accuracy of CT imaging alone for diagnosing hepatic cysts were 92.54%, 93.12%, and 91.02%, respectively, showing a slight disadvantage compared to MRI. There were statistically significant differences in SNR, CNR, and subjective scores among CT images reconstructed by different algorithms (P < 0.001). DLIR-M and DLIR-H achieved the highest subjective evaluation scores, with DLIR-M scoring 4.87, 4.85, 4.87, and 4.88 in noise, contrast, edge, and fine structure, respectively. Conclusion : The application of the DLIR reconstruction algorithm can optimize the quality of abdominal CT images based on CT spectral imaging, improve diagnostic accuracy, and enhance the work efficiency of physicians. This research method provides a new development direction for abdominal examination techniques.