A neurovascular high-frequency optical coherence tomography system enables in situ cerebrovascular volumetric microscopy

Abstract

Intravascular imaging has emerged as a valuable tool for the treatment of coronary and peripheral artery disease; however, no solution is available for safe and reliable use in the tortuous vascular anatomy of the brain. Endovascular treatment of stroke is delivered under image guidance with insufficient resolution to adequately assess underlying arterial pathology and therapeutic devices. High-resolution imaging, enabling surgeons to visualize cerebral arteries' microstructure and micron-level features of neurovascular devices, would have a profound impact in the research, diagnosis, and treatment of cerebrovascular diseases. Here, we present a neurovascular high-frequency optical coherence tomography (HF-OCT) system, including an imaging console and an endoscopic probe designed to rapidly acquire volumetric microscopy data at a resolution approaching 10 microns in tortuous cerebrovascular anatomies. Using a combination of in vitro, ex vivo, and in vivo models, the feasibility of HF-OCT for cerebrovascular imaging was demonstrated.

Fig. 1. Principles of HF-OCT imaging in an intracranial vessel.

During a brief injection of contrast (blue color) through a 5 F distal access catheter, the Vis-M device is retracted while quickly rotating its internal optics resulting in a helical scanning pattern. By the means of a tightly spaced pattern (e.g., 80 µm) and an axial resolution approaching 10 µm, comprehensive volumetric microscopy of the arterial wall, neurovascular devices, and intraluminal objects is obtained.

Fig. 2. In vivo imaging in a flexed forelimb model of brachial porcine artery.

a The dotted line highlights the estimated path taken by the Vis-M device through the vessel tortuosity. b HF-OCT microscopy shows the external elastic lamina (arrowheads) and the individual layers of the vessel wall (arrows). A bright tunica intima is followed by a dark tunica media and a bright adventitia (inset). The asterisks denote the ostia of two side-branches, with diameters of ~0.2 and 0.7 mm, respectively. c The arrow indicates the eccentric position of the HF-OCT device in the arterial lumen. The image shows a uniform illumination and absence of NURD artifacts. Imaging in a flexed forelimb swine model was repeated in a total of n = 16 arteries from all animals included in this study (n = 8). The scale bar on DSA image is equal to 1 cm (a). HF-OCT images scale bars are equal to 1 mm (b and c), and 0.5 mm in the inset.

Fig. 3. HF-OCT volumetric rendering of a stented swine internal maxillary artery.

a HF-OCT three-dimensional cutaway rendering (top = distal; bottom = proximal). Flow-diverter malapposition (arrow) and clots of different sizes (purple color) are visible over the flow-diverter surface. b Cross-sectional HF-OCT imaging show a jailed branch (asterisk) and several thrombus formations over the flow-diverter surface (arrowheads). c Flow-diverter proximal edge, with incomplete apposition (3 o’clock) and several clots over the device surface (arrowheads). d Endoscopic view of HF-OCT volumetric rendering. Small, perforator-like side-branches jailed by a flow-diverting stent are visible. e The side-branch located on the left side of the image (arrow) is free of clots and the flow-diverting stent is well-apposed to the parent artery. f A second branch located on the right (indicated by the asterisk) shows flow-diverter struts that are embedded by a clot. Stented arteries data collection was repeated in a total of n = 16 swine internal maxillary arteries, including n = 16 flow-diverting stents, and n = 15 self-expanding intracranial stents. Scale bars on all HF-OCT images are equal to 1 mm. Three-dimensional rendering color scheme: red, artery wall; purple, clot; silver, metallic struts.

Fig. 4. HF-OCT imaging comparison with DSA.

a Digital subtraction angiography (DSA) of an internal maxillary artery stented using flow-diverting and neurovascular stents. The stented segment is indicated by the arrow, with the devices partially overlapping. Corresponding HF-OCT images show thrombus formations (arrows) not observed on fluoroscopy located over the struts of the flow-diverter at the level of a side branch b, and over the struts of the flow-diverter and intra-cranial stent in the overlapping segment c. d DSA of a second IMAX artery treated with a flow-diverting stent. The corresponding HF-OCT cross-sectional images depict device malapposition at different locations not seen on DSA, with maximum extensions of ~0.35 mm e and 0.45 mm f. HF-OCT comparison with DSA was repeated for all swine internal maxillary arteries (n = 16), following flow-diverting (n = 16), and self-expanding intracranial stent implantation (n = 15). Scale bars on DSA images are equal to 5.0 mm; HF-OCT scale bars to 1.0 mm.

Fig. 5. Cross-sectional HF-OCT images compared to corresponding CBCT slices.

a A side-branch thrombosis is visible on the HF-OCT image (arrow). b Flow-diverter malapposition with a maximum severity of ~400 µm is visible on the HF-OCT image between 1 and 8 o’clock. Small thrombus formations over the flow-diverter struts, with a thickness between ~30 and 220 µm, are indicated by the arrowheads. The presence of thrombus and device malapposition are often undetected on the corresponding cone beam CT images. c Thrombosis on the ostium of a large side-branch. HF-OCT comparison with CBCT was repeated for all swine internal maxillary arteries (n = 16), following flow-diverting (n = 16), and self-expanding intracranial stent implantation (n = 15). Scale bars on HF-OCT images are equal to 1 mm. The star symbol (*) on CT images denotes the location of side-branches.

Fig. 6. HF-OCT volumetric microscopy of an intracranial stent partially overlapping with the distal end of a flow diverting stent.

a Three-dimensional rendering showing significant thrombus accumulation. b HF-OCT microscopy depicts the vessel wall microstructure including the individual vessel layers comprising a bright tunica intima, a low-scattering tunica media, and the tunica adventitia (arrows). The internal elastic lamina (IEL) and the external elastic lamina (EEL) are indicated by the green arrows. c A thrombus dislodged from the surface of the device floating inside a large branch is visible and denoted by the asterisk. d Thrombus formations ranging between 100 and 200 µm in thickness are indicated by the asterisks and distributed over the FDS surface. e A semi-occlusive clot formation in correspondence of a significant malapposition (>500 µm) is visible over the proximal end of the FDS. HF-OCT imaging was repeated in a total of n = 15 swine internal maxillary arteries with an intracranial stent partially overlapping with a flow-diverting stent. Scale bars are equal to 1.0 mm. Three-dimensional endoscopic rendering color scheme: red, artery wall; purple, clot; silver, flow-diverting stent struts; gray, neurovascular stent struts.

Fig. 7. Intracranial plaques from an ex vivo segment of the MCA artery.

a HF-OCT imaging of a fibrotic plaque and corresponding trichrome b and Movat’s staining c. Fibrotic tissue is characterized by HF-OCT as a region of homogenous signal, resulting from elevated backscattering and low optical attenuation coefficients. d HF-OCT imaging of a necrotic core plaque and corresponding H&E e and Movat’s staining f. A necrotic core plaque is characterized on HF-OCT images as an area with poorly delineated borders followed by an elevated optical attenuation coefficient within an atherosclerotic plaque. The asterisk indicates a vessel wall dissection. From n = 10 cadaveric specimens of intracranial arteries, a total of n = 3 representative atherosclerotic plaques were identified, processed for histopathology analysis, and compared to HF-OCT imaging. Scale bars are equal to 1.0 mm.

Fig. 8. Intracranial fibrocalcific plaques in a segment of an intradural vertebral artery.

a Two fibrocalcific plaques are visible in the image. A smaller plaque with a calcium thickness between 100 and 300 µm is located at 11 o’clock. A larger plaque with a calcium maximum thickness of 900 µm is located in the bottom-left quadrant of the image. b Magnification showing fine details of the plaque microstructure, including inner and outer boundaries. On HF-OCT, calcific tissue is characterized by a sharply demarcated area with a weaker and heterogeneous signal, resulting from a low optical backscattering and low absorption coefficients. Characterization of HF-OCT versus histopathology for imaging of fibrocalcific plaques was obtained from one intracranial artery specimen. Scale bars are equal to 1 mm.