High-speed FPGA-GPU processing for 3D-OCT imaging

Doppler optical coherence tomography for energy seal evaluation and comparison to visual evaluation

“…In the literature, B-scan postprocessing rates have been reported to be as fast as 40 FPS when GPU-FPGA integrated systems were used. 37 If these frame rates were implemented here, then the postprocessing times could be decreased from 2.5 min to 5 s. Based on these improvements to supporting technology, it is estimated that the overall acquisition and processing time for a single data point could be performed in under a minute. The primary biomarker used for energy seal evaluation here was the characteristic lemniscate that defined the natural state of the CAM vessel.…”

Transformation and dynamic visualization of images from computer through an FPGA in a matrix of LED

“…Around the transformations [11], extraction of characteristics [12] or any other type of massive matrix operation, object detection and recognition and route planning, Programmable Logic Devices (PLD) have been used, since they require low power and a small amount of space to work [13,14] in addition to having the ability to perform real-time implementations, working with static and moving images [15], which requires thousands of iterations per second, with maximum use of FPGA resources, both in the combinatorial part (LUT) as in its storage part (memory bank), making efficient the implementations made with this type of devices [16,17]. The acquisition of real-time images has become an everyday task, using conventional cameras and stereo cameras for capturing in 3D [18,19], in which applications with digital programmable devices have been made to perform high-speed processing, achieving solutions where a single FPGA is used that replaces systems of multiple architectures that required connections and conversions of information, slowing down actions on processes in jobs where short response times are required [20,21].…”