Chinese Hepatolgy ›› 2025, Vol. 30 ›› Issue (9): 1244-1248.

• Liver Tumor • Previous Articles     Next Articles

The combination of MRI multi-modal parameters and MRI texture parameters has value in the diagnosis of background liver nodules in fatty liver

ZHANG Wei, WEN Lei-tao, LIU Wen-qian   

  1. Department of Radiology, Xi'an High tech Hospital, Xi'an 710075,China
  • Received:2024-06-11 Online:2025-09-30 Published:2025-11-05
  • Contact: WEN Lei-tao,Email:wenleitao999@163.com

Abstract: Objective To analyze the value of magnetic resonance imaging (MRI) multimodal parameters combined with MRI image texture parameters in diagnosing small liver nodules in the background of fatty liver. Methods A retrospective analysis was performed on medical records of 94 patients with fatty liver background liver nodules treated at Xi'an High tech Hospital from March 2021 to March 2023. T1WI, liposuppression T2WI, MRI diffusion-weighted imaging (DWI), dynamic enhanced imaging (DEC)-MRI scan data were collected from all patients before pathological puncture or surgery. Multimodal parameters and image texture feature parameters were analyzed. Using pathological diagnosis as the gold standard, determine the benign or malignant nature of liver nodules in the background of fatty liver. Analyze the influencing factors for malignancy of liver nodules in the context of fatty liver, and analyze the sensitivity and specificity of MRI-DWI parameters, DEC-MRI parameters, and image texture feature parameters for the single or combined diagnosis of benign or malignant liver nodules in fatty liver. Results Among 94 patients with background liver nodules in fatty liver, 40 patients were diagnosed as malignant nodules (+) and 54 patients were diagnosed as benign nodules (-) by pathology. Under diffusion sensitive gradient factor (b value) conditions of 100 s/mm2, 500 s/mm2, and 1 000 s/mm2, the apparent diffusion coefficient (ADC) of the malignant nodule group was (1.99±0.51)×10-3 mm2/s, (1.87±0.54)×10-3 mm2/s, and (1.65±0.49)×10-3 mm2/s, respectively, which were lower than those of the benign nodule group [(2.40±0.68)×10-3 mm2/s, (2.23±0.57)×10-3 mm2/s, and (2.09±0.53)×10-3 mm2/s, (P<0.05)]. The maximum slope increase (MSI), maximum slope decrease (MSD), and peak value (PV) of the malignant nodule group were (301.08±59.06), (92.63±24.08), and (512.36±72.19), respectively, which were lower than those of the benign nodule group [(369.16±58.14), (106.35±27.41), and (581.34±80.27), (P<0.05)]. The S(5,5) Sum Varnc/×102, S(3,3) Sum Varnc/×102, S(2,3) Sum Varnc/×102, and S(0,1) Sum Varnc/×102 of the malignant nodule group were higher than those of the benign nodule group (P<0.05). ADC value (b=1 000 s/mm2, OR=0.282, 95%CI: 0.104~0.763), MSI (OR=0.271, 95%CI: 0.100~0.735), MSD (OR=0.265, 95%CI: 0.098~0.717), PV (OR=0.318, 95%CI: 0.117~0.860), S(5,5) Sum Varnc/×102 (OR=3.823, 95%CI: 1.411~10.356) were identified as influencing factors for malignancy of liver nodules in the background of fatty liver (P<0.05). The sensitivity of MRI-DWI parameters, DEC-MRI parameters, and image texture feature parameters for single and combined diagnosis of benign or malignant liver nodules in fatty liver were 0.624, 0.697, 0.683, and 0.731, respectively, while the specificity was 0.713, 0.772, 0.659, and 0.800, and the area under the curve was 0.720, 0.751, 0.713, and 0.832, respectively. Conclusion The combination of MRI multimodal parameters and MRI image texture parameters has a high accuracy in diagnosing small liver nodules in the background of fatty liver, and could be used to assist in the diagnosis of benign and malignant liver nodules in the background of fatty liver.

Key words: Fatty liver, Small nodules, Magnetic resonance imaging, Diffusion weighted imaging, Dynamic enhanced imaging, Image texture parameters