Volumetric analysis of non-contrast Magnetic Resonance Lymphangiography in patients affected by lower extremities primary lymphedema

Aim: Since the response to conservative and surgical treatments in lymphedema is based only on clinical evaluation, we want to propose a new technique for the volume calculation of lower extremities affected by lymphedema. Materials and methods: Ten female patients affected by primary LE of the lower limbs (mean age 33 ± 5 years; age range 16–42) underwent NCMRL in September 2019. Acquisition protocol included a STIR sequence (FOV 460 × 504 mm; matrix 315 × 384 pixel; voxel size 1.1 × 1.1 × 1.0 mm; TR 3000 ms; TE 254 ms; TI 160 ms). Two readers independently performed a manual segmentation with ITK-SNAP open-source software to assess the volume of the lower limbs. The agreement for volumetric segmentation between the two operators was assessed through intraclass correlation coefficient (ICC). Results: Mean segmentation time was 5 ± 0.3 min. ICC demonstrated excellent intra-examiner agreements for both readers 1 (ICC = 0.9991, p < 0.0001) and 2 (ICC = 0.9989, p < 0.0001). Inter-reader agreement was excellent (ICC = 0.9991, p < 0.0001). Conclusions: Manual segmentation with ITK-SNAP of NCMRL examinations in patients affected by primary lower limb LE represents a reproducible procedure for an objective evaluation of volumes, useful for follow-up.