Objective:To study the value of sound touch elastography (STE) linear combined with ultrasound score (US) in the diagnosis of chronic hepatitis B (CHB) liver fibrosis, and to investigate whether their combination can improve the diagnostic efficiency of subdividing the degree of CHB liver fibrosis. Furthermore, a comparison with STE linear combined with the serological model was performed to seek the optimal linear combination model.Methods:A total of 313 subjects were enrolled from September 2018 to December 2021 in Shenzhen Third People′s Hospital Affiliated to Guangdong Medical University, including 259 patients with CHB who had completed liver biopsy and 54 healthy volunteers. CHB patients were divided into liver fibrosis group (F1-F4 group) according to METAVIR classification standard, and healthy volunteers were used as the control group. All subjects underwent liver ultrasound examination, STE and blood biochemical indexes of liver function. The US was performed according to the liver ultrasound examination, and the liver stiffness measurement (LSM) was measured by STE, aspartate aminotransferase and platelet ratio index (APRI) was calculated by blood biochemical index. Fisher discriminant analysis was used to establish the linear combination (LC) diagnostic marker of US and LSM, and the linear combination (LC2) diagnostic marker of LSM and APRI, successively. Spearman rank correlation coefficient was used to analyze the correlations between US, LSM, APRI, LC2, LC and pathological results. The ROC curves of US, LSM, APRI, LC2 and LC for diagnosing CHB liver fibrosis were plotted, and the diagnostic efficiency of above diagnostic markers was evaluated according to the accuracy, sensitivity, specificity and area under the ROC curve (AUC).Results:The formula for the linear combination of US and LSM was LC=0.986 0×US+ 0.166 7×LSM, and LC was highly positively correlated with pathological findings ( rs=0.851, P<0.001), higher than US, LSM, LC2 and APRI ( rs=0.825, 0.775, 0.802, 0.586, all P<0.001). LC showed the best diagnostic efficiency. The AUCs for diagnosing ≥F1, ≥F2, ≥F3 liver fibrosis and =F4 cirrhosis were 0.945, 0.911, 0.954, 0.955, respectively, which superior to the AUCs of US (0.913, 0.879, 0.934 and 0.916, respectively), the AUCs of LSM (0.860, 0.871, 0.934 and 0.952, respectively) and the AUCs of LC2(0.899, 0.883, 0.941, 0.946, respectively). Compared with US, the AUC of LC diagnosis of ≥F1, ≥F2, ≥F3 liver fibrosis and =F4 cirrhosis increased by 3.2%, 3.2%, 2.0% and 3.9%, respectively, with all significant differences ( P<0.05). Compared with LSM, the AUC of LC increased by 8.5%, 4.0%, 2.0% and 0.3%, respectively, with significant difference ( P<0.05) except for stage =F4 cirrhosis.Compared with LC2, the AUC of LC increased by 4.6%, 2.8%, 1.3% and 0.9%, respectively, and there were significant differences in the diagnosis of ≥F1 and ≥F2 liver fibrosis ( P<0.05). Moreover, the overall efficiency of LC2 was not significantly improved than LSM, the difference was not significant ( P>0.05). Conclusions:US, LSM, LC2 and LC can be used to diagnose the degree of CHB liver fibrosis, but LC is better than US or LSM and LC2 alone, especially in the subdivision of mild liver fibrosis, which is a promising new diagnostic marker to subdivide the degree of CHB liver fibrosis.