The lymphatic system is the responsible to collect the lymph. When the lymphatic system is harmed, it may cause fluid to build up especially in arms and legs. This article focuses on the swelling of the lower-limbs due to lymphedema which are usually treated with compression garments therapy, such as compression stockings \cite{Ciudad2019}. The limited number of sizes/shapes of stockings reduces the efficiency of the stockings. The purpose of this study is to design them via a real-time physical simulation, which takes as inputs parameters defining the geometry of a given patient (patient-specific solution). For this purpose, a solution based in Reduced Order Model (ROM) is developed to consider the geometrical parameters of the lower-limbs \cite{Chinesta2023}. The ROM predicts the desired parameter, hydrostatic stresses, based on the high-fidelity solution reducing the computational cost significantly. Hydrostatic stress is a compressing pressure that squeeze lymphatic vessels (hypertension), which improves lymph circulation and increase the drainage capacity. The ROM enables to know stress distribution in the soft tissues for any leg without solving the high-fidelity model. The ROM require to solve a high number of models, in this case legs, in order to create a training dataset. Such pre-processing step (offline stage) is highly time-consuming. Starting from CT-scan of patients, the contour of the leg, muscle and bones are defined to build the 3D geometry of the leg. A semi-automatic method is proposed to carry out the image segmentation of the lower-limb based in an extrapolation of a limited set of geometrical parameters. The geometry is registered through a reduced number of variables applying the Free-Form Deformation (FFD) algorithm, which are the variables or inputs that define the ROM. The proposed method reduces the cost of time to identify and associate each pixel to an object (skin/fat, muscle or bone). This facilitates the definition of the geometry of a high number of patients lower-limbs and it enables to define the geometrical parameters for each patient.