Selection for increasing the intramuscular fat content (IMF) of pigs is a mechanism to improve the organoleptic quality of pork. In vivo estimation of IMF is based on ultrasound images recorded on the loin. Historically, the IMF of these images was estimated manually by a technician using the BioSoft Toolbox for Swine (GIMr) (Biotronics, Inc.). The disadvantage of this method is that the time between the ultrasound measurement and receipt of the estimated IMF takes several days, making it impossible to use the phenotype directly when selecting pigs. To overcome this situation, an automatic method was developed. The objective of this project was to evaluate this method developed to automate image processing to estimate IMF (GIMa). A dataset based on images collected for 1957 Duroc pigs and collected from 2021-2023 was used. The first step of this method consists of automatically delineating the area of interest using a Canny filter. Then, a linear regression model that includes several variables is used to estimate the GIMa. The automated method made it possible to estimate IMF for all pigs in the dataset, whereas the reference method could estimate IMF for only 73.5 % of the pigs. The correlation coefficient between GIMr and GIMa was 0.78, and the rank correlation of pigs between GIMr and GIMa was 0.89. This study thus confirmed the benefit of the automated method, which reduced imageprocessing time.