Lymphedema is a prevalent condition affecting approximately 250 million individuals worldwide [1]. Upper extremity lymphedema (UEL) is a chronic disorder caused primarily by lymph node dissection during breast cancer treatment [2]. Although the diagnosis of UEL is essential for efficient treatment, there have been few clinical examinations giving meaningful information about the underlying abnormalities. Indocyanine green lymphography (ICG-L) has emerged as a valuable tool for real-time visualization of the functional aspects of the lymphatic system, allowing for the precise identification of lymphatic pathways, and dermal backflow [[3], [4], [5], [6]]. This imaging technique has become an essential diagnostic modality for UEL.

Recently, more and more plastic surgeons have been familiar with supermicrosurgery and several articles reported the efficacy of lymphaticovenular anastomosis (LVA) and vascularized lymph node transfer (VLNT) [7,8]. Although surgical treatments are often applied for patients with moderate or severe dermal backflow patterns on ICG-L and resistant to conservative therapies, there are no definitive criteria for choosing the appropriate candidate for LVA or VLNT regarding fluid distribution [9]. Thus, an efficient image inspection for fluid distribution is urgently needed to guide the selection of appropriate surgical treatments, leading to improved outcomes and higher patient satisfaction [10].

In the past decade, non-contrast magnetic resonance lymphography (NMRL) has attracted broader attention as image inspections which have the potential for objective evaluation of fluid accumulation in lymphedema [11,12]. The three-dimensional (3D) NMRL sequence gives detailed information about the overview of fluid distribution on upper limbs and its images are very easy to understand even for physicians unfamiliar with radiographic images [13] However, there is limited information about characteristic 3D NMRL findings in relation to the pathophysiologic condition of lymph circulation. Because there are several reports about the effectiveness of ICG-L in evaluating the severity of lymphatic impairment, we choose ICG-L as the ideal control modality to evaluate NMRL images [14]. The purpose of this study was to investigate the association between characteristic NMRL findings and ICG-L findings. In particular, we wish to study the correlation between the NMRL staging and the ICG-L staging.