Feature-Guided Image Fusion of Intrasurgical Optical Coherence Tomography and Digital Surgical Microscopy
Eye surgery is a high-impact intervention necessary to preserve the sight of thousands of patients, with retinal detachment surgeries alone constituting approximately 28,000 cases per year in the United States alone. Microscope-integrated optical coherence tomography (MIOCT), which combines standard two-dimensional (2D) color microscopy with three-dimensional (3D) optical coherence tomography imaging (OCT), has emerged as a promising intrasurgical imaging technique to improve real-time guidance for major eye surgeries. However, in the current embodiment, visualizations for each mode are displayed separately, requiring surgeons to mentally align and compare features across both images. There is a need for new methods to combine information from both modalities into a single reconstruction in order to provide surgical guidance superior to that of separate visualizations, which should lead to better outcomes.
Duke inventors have developed a method for fusing data from 2D color microscopy and 3D OCT, yielding a more efficient visualization of cues from each. This is intended to be used by ophthalmic surgeons for image-guided eye surgeries, as well as for non-surgical applications where 2D color imaging data and 3D imaging data can be combined to enhance visualization for clinicians. Specifically, the method works by receiving 3D imaging data from OCT and 2D color imaging data of a region of interest. The 3D imaging data is segmented using either manual or automatic methods to identify anatomical features in the region of interest. Next, an image is generated by using a rendering process based on the volumetric distribution of color emission and density to fuse the 2D color data and the 3D imaging data according to the surfaces, the corresponding volumes, and identities of the anatomical features. Ultimately, this process leads to a single visualization that integrates information from each modality. This technology has been demonstrated using images previously acquired during surgery.
This technology could also be applied to many other multimodal imaging combinations, including, but not limited to, structural 3D sweptsource OCT (SSOCT), limited to time-domain OCT (TD-OCT), spectral-domain OCT (SDOCT), OCT angiography (OCT-A), color fundus 2D data, fluorescence imaging, hyperspectral imaging, scanner laser ophthalmoscopy (SLO), adaptive optics SLO (AO-SLO), confocal microscopy, polarization-sensitive-OCT (PS-OCT), spectral OCT (SOCT), and ultrasound.
- Integrates information from both 2D color microscopy and 3D OCT imaging in a single, easy to understand visualization
- Fused visualization of data will provide surgical guidance superior to that of separate visualizations
- Method does not compromise the 3D context provided by OCT data