3D printing of the choroidal vessels and tumours based on optical coherence tomography

T he choroid has the highest blood flow of any structure in the human body with specific hemodynamic regulatory mechanisms that differ from those of the retinal circulation. It is modulated by a strong autonomic input and is largely insensitive to light stimulation and to differences in blood oxygenation (Kur et al. 2012). Detailed imaging of the choroidal layers has recently been facilitated by the introduction of new imaging modalities such as enhanced depth imaging optical coherence tomography (EDI-OCT) and swept source OCT (SSOCT) (Mrejen & Spaide 2013). One of the more important limitations of routine OCT lies with the two-dimensional (2D) display of static images on a computer screen, hampering characterization of clinically important structural features, such as depth and spatio-anatomical localization. This drawback has been circumvented by the advent of threedimensional (3D) imaging techniques that provide detailed and problemoriented information on both the retinal and choroidal compositions, including volume rendering (Spaide 2015). Similar to routine OCT, this technique is restricted to 2D display in computer screens. Recently, a new method has been reported in which printing of OCT data has been described in a patient with an epiretinal membrane (Choi et al. 2016). This study describes for the first time the use of a 3D printing technique, speckle-free 3D choroidal angiography and tumoropsy (Maloca et al., 2016), applied to3Dprintingof choroidalvessels and pigmented choroidal tumours. In this study, retrospective 1050 nm OCT volumes were collected from healthy eyes and eyes with pigmented choroidal tumours to evaluate choroidal vessel architecture and tumour 3D printing. Inclusion criteria were age >18 years, adequate media clarity for fundus imaging, good central fixation and visual acuity >20/20. Exclusion criteria were nystagmus, poor cooperation and dry eye syndrome. All subjects underwent a comprehensive baseline ophthalmologic examination to exclude any potential retinal or choroidal disorders. Written informed consent was obtained from all patients, and approval was attained from the local ethical committee in accordance with the Declaration of Helsinki and in compliance with data protection regulations. All retinal OCT volumes were acquired in nondilated pupils with a SSOCT device (DRI OCT Triton; Topcon, Tokyo, Japan). The SSOCT volume was captured in a 3D scan pattern over a 3 9 3 mm, 6.0 9 6.0 mm or 9 9 12 mm area, respectively, centred on the region of interest (ROI) with 256 B-scans and a scan density of 512 9 256 pixel. Image processing was performed with a previously published 3D speckle-noise removal method with structure preservation (Gyger Cyrill et al. 2014). For choroidal vessel lumen and tumour extraction, the hyporeflective choroidal vessels and hyperreflective tumour structures, respectively, were manually segmented by threshold filtering in the speckle-free OCT volume (IMAGEJ v1.467; ref – Rasband, W.S., IMAGEJ, US National Institutes of Health, Bethesda, MD, USA, https:// imagej.nih.gov/ij/, 1997–2016) by extracting lumen information from the scan volume. The 3D information of the processed choroid was saved as obj-file which was then enhanced by sealing gaps in the mesh or removing obvious artefacts. Ultimately, a 3D printable OCT model was obtained (Fig. 1). Some models were sent for 3D stereolithography printing in transparent resin or constructed from a hardened liquid (i.materialise, i.Materialise HQ, Leuven, Belgium). One model was submerged in a bath of carat gold (24K) to increase robustness and durability. Other models were printed in additive fused deposition modelling using a gypsum powder for testing combined vessel and tumour structure printing, respectively (3d-prototyp.com, Stans, Switzerland). Design specifications for 3D printing included minimum wall thickness of 1 mm, minimum details of 0.5–1 mm and a size of 130 9 200 9 10 mm. In addition, 3D prints of 300 9 300 9 23 and 210 9 390 9 23 mm have been made (Fig. 2). This corresponds to a magnification of up to 70–100 times. Analysis of 3D print models allows a detailed spatio-anatomical characterization of choroidal vessels and their