The present invention relates to a device for predicting liver fibrosis by using CT image analysis and a method therefor. A method for predicting liver fibrosis by using CT image analysis according to the present invention comprises: a first step of receiving, by a liver fibrosis prediction device, a liver CT image of an analysis target from a user； a second step of receiving a selected area of interest regarding a layer selected by the user from multiple layers of the liver CT image； a third step of analyzing the degree of corrugation by calculating average values of coordinates of lines regarding a liver surface included in the area of interest, principal component analysis (PCA) being used to expand the area of interest in the horizontal direction so as to analyze the degree of corrugation of the corresponding layer； a fourth step of moving to the next layer of the selected layer, if the area of analysis is out of a preconfigured area, and analyzing the degree of corrugation of the corresponding layer with regard to the area of interest of the corresponding layer according to the same method as in the third step； a fifth step of calculating the average and standard deviation of values resulting from analyzing the degree of corrugation of respective layers, after analysis of all layers is completed； and a sixth step of predicting the degree of liver fibrosis from the calculated values and outputting the result of prediction. As such, according to the present invention, an abdominal CT is analyzed, the degree of corrugation of a liver surface is quantified, and the degree of liver fibrosis is accordingly predicted, thereby ensuring more accurate diagnosis.La présente invention concerne un dispositif de prédiction de fibrose hépatique à l'aide d'une analyse d'image de tomodensitométrie et une méthode associée. Une méthode de prédiction de fibrose hépatique à l'aide d'une analyse d'image de tomodensitométrie selon la présente invention comprend : une première étape c
