The current strategies for the transplantation of meniscus should be strengthened to tackle the faced limitations of current methods in the clinics. One of the limitations is that current implants are not patient-specific. There is, therefore, a pressing need in the clinics to develop patient-specific implants. The aim of this study was to demonstrate a semi-automatic way of segmenting meniscus tissues from patients‘ volumetric knee magnetic resonance imaging (MRI) datasets in order to obtain patient-specific 3D models for 3D printing of patient-specific constructs. High-quality MRI volumetric images were acquired from five healthy male human subjects. The advanced segmentation software, RheumaSCORE, was used for semi-automatic MRI image segmentation of the meniscus tissues. Our methodology allows a full 3D segmentation of the menisci with only minimal interaction on 2D slices. The obtained 3D models were used for the fabrication of tissue engineering scaffolds from polycaprolactone with different internal architectures. The fabricated scaffolds were characterized by micro-computed tomography (µ-CT), scanning electron microscopy (SEM), and mechanical testing. This study demonstrated the 3D fabrication of patient-specific scaffolds with a 3D printer using the reconstructed 3D models obtained by an advanced segmentation method of menisci from knee MRI. This is a step towards a personalized tissue engineering therapy model for the knee meniscus.