Correlation analysis between preoperative MRI image features based on image reconstruction algorithms and incision healing in patients with benign hepatobiliary diseases undergoing postoperative drainage tube placement

This study was designed to examine the association between preoperative abdominal MRI features derived using a deep learning-based reconstruction (DL-Recon) algorithm and postoperative surgical incision-healing outcomes, while evaluating the role of subcutaneous drainage tubes in improving healing. One hundred and eighty patients undergoing abdominal surgery for benign hepatobiliary diseases (including symptomatic hepatic cysts and chronic cholecystitis with stones) were randomly assigned to the experimental group (DL-Recon-enhanced preoperative abdominal MRI, n = 100) and control group (no drainage, n = 80). The experimental group was further divided into Subgroup A (using 3M medical breathable tape in a cross pattern, n = 52) and Subgroup B (using 3M elastic tape with elevated platform design, n = 48). Primary outcomes included incision healing time and healing rate. Secondary outcomes comprised preoperative imaging characteristics, laboratory inflammatory markers, pain scores, and comparisons of complications associated with the two drainage tube fixation methods. The signal-to-noise ratio (SNR) of MRI images was significantly improved by the DL-Recon algorithm. The mean incision healing time in the experimental group was found to be significantly shorter than that in the control group ( P < 0.05). Within the experimental group, incision healing rates for grades A, B, and C were 82.0%, 16.7%, and 1.3%, respectively, compared to 52.6%, 28.5%, and 18.9% in the control group, with statistically significant variations between groups ( P < 0.01). Postoperative inflammatory indices (white blood cell count, C-reactive protein) and pain scores were also determined to be significantly reduced in the experimental group relative to the control group. Correlation analysis revealed a positive correlation between preoperative subcutaneous fat thickness and healing time, while the fat-to-muscle signal intensity ratio showed a positive correlation with postoperative inflammatory and pain indicators (all P < 0.01). Preoperative MRI feature analysis based on DL-Recon algorithm correlates with postoperative outcomes, suggesting its potential for preoperative assessment. Subcutaneous drainage tube placement after surgery for benign hepatobiliary diseases (hepatic cysts, chronic cholecystitis with stones) promotes incision healing.