Rapid Quantification of Microvessels of Three-Dimensional Blood-Brain Barrier Model Using Optical Coherence Tomography and Deep Learning Algorithm

Abstract

The blood-brain barrier (BBB) is a selective barrier that controls the transport between the blood and neural tissue features and maintains brain homeostasis to protect the central nervous system (CNS). In vitro models can be useful to understand the role of the BBB in disease and assess the effects of drug delivery. Recently, we reported a 3D BBB model with perfusable microvasculature in a Transwell insert. It replicates several key features of the native BBB, as it showed size-selective permeability of different molecular weights of dextran, activity of the P-glycoprotein efflux pump, and functionality of receptor-mediated transcytosis (RMT), which is the most investigated pathway for the transportation of macromolecules through endothelial cells of the BBB. For quality control and permeability evaluation in commercial use, visualization and quantification of the 3D vascular lumen structures is absolutely crucial. Here, for the first time, we report a rapid, non-invasive optical coherence tomography (OCT)-based approach to quantify the microvessel network in the 3D in vitro BBB model. Briefly, we successfully obtained the 3D OCT images of the BBB model and further processed the images using three strategies: morphological imaging processing (MIP), random forest machine learning using the Trainable Weka Segmentation plugin (RF-TWS), and deep learning using pix2pix cGAN. The performance of these methods was evaluated by comparing their output images with manually selected ground truth images. It suggested that deep learning performed well on object identification of OCT images and its computation results of vessel counts and surface areas were close to the ground truth results. This study not only facilitates the permeability evaluation of the BBB model but also offers a rapid, non-invasive observational and quantitative approach for the increasing number of other 3D in vitro models.

Keywords: 3D BBB model; OCT image processing; vessel quantification.

Figure 1

The workflow of non-invasive OCT-based approach to visualize lumen structures of 3D BBB in vitro model. The x-z cross-sectional images obtained from OCT were converted to z-stacks of x-y images. These converted images were analyzed with morphological imaging processing (MIP), RF machine learning using TWS plugin (RF-TWS), and deep learning using pix2pix cGAN and then the output images were used for visualization and quantification of the 3D vessel network of BBB model.

Figure 2

Orthogonal views (xy, xz, and yz) of the converted OCT images showing 3D structure of microvessels in the BBB model. (a). Top, front, and side views of a representative OCT image located at 160 µm depth from bottom side of the tissue structure. The lengths of the x-axis, y-axis, and z-axis are 6000, 6000, and 900 µm, respectively (z was magnified 3 times), and vessel lumen structures are indicated with red arrows. (b). The high-magnification images of the red boxed area of a and open and closed vascular structures are indicated with red and green arrows, respectively. Scale bar of the global and magnification image is 1 mm and 0.1 mm, respectively.

Figure 3

Input and output images identified by manual selection (ground truth), morphological imaging processing (MIP), Random forest machine learning using Trainable Weka Segmentation plugin (RF-TWS), and deep learning using pix2pix cGAN (pix2pix cGAN) of surface (0~50 µm), middle (50~200 µm), and deep layer (200~310 µm) images.

Figure 4

(a). Cohen’s kappa coefficient (Equation (1)) of each frame with increasing depths (from left to right) of three datasets; each frame was 10 µm depth. (b). Averaged Cohen’s κ of three approaches calculated from test datasets 1, 2, and 3; error bar represents maximum and minimum value.

Figure 5

Vascular structures generated by the binary images of ground truth (upper left) and deep learning using pix2pix cGAN (upper right) of test dataset 1 (the dimension of x, y, and z-axis are 600, 600, and 20 pixels, respectively) (a), and averaged computation results (n = 3) of each 3D BBB model (b): total vessel number (left), total surface area of open vascular structures (middle) and mean diameter of each vessel (right) calculated from images of ground truth (GT) and deep learning (DL); error bars represent standard deviation of three datasets. p-values for paired t-tests for equivalence between GT and DL are 0.97 for the number of structures, 0.59 for the total surface area, and 0.01 for the diameters.