GigE Vision for 3D Medical Research
Rheumatoid Arthritis is a widespread autoimmune disease that attacks many tissues and organs, but principally affects synovial joints (hands, feet, knees and elbows) causing chronic inflammation as well as deformation. Rheumatoid arthritis affects millions of people of all ages worldwide. Although there is no known cure for the disease to date, medications to relieve symptoms and prevent joint destruction and deformity are widely available.
threeRivers 3D, a 3D imaging solutions company based near Pittsburgh PA, was commissioned by a major medical research center to develop a 3D custom hand scanner. The purpose of this custom device is to scan patients’ hands to monitor the progression of rheumatoid arthritis by tracking volume changes and joint inflammation over a period of time.
The system consists of a 80cm (H) x 60cm (W) x 40cm (D) metal frame with four posts, two high speed 3D laser scanners that project patterns on the patients hand for the capture of image data, a color Prosilica GC1290C camera fitted with a standard 10mm Edmund Optics fixed lens to monitor joint deformation, and a thermal camera that measures inflammation in the joints. Each 3D laser scanner contains a monochrome Prosilica GC1290 camera fitted with standard 8mm Edmunds Optics fixed lenses to capture the projected laser patterns. The system is also equipped with a thermal reference device located near the palm rest. The device measures the ambient temperature to calibrate the system for accurate measurement of the temperature of the patient’s hands.
The Prosilica GC1290 and GC1290C by Allied Vision Technologies (AVT) are compact 1.3 Megapixel cameras, which output uncompressed real-time images over a Gigabit Ethernet interface. It features the Sony ICX445 ExView HAD CCD sensor for high sensitivity and excellent image quality. threeRivers 3D selected the Prosilica GC1290 series for its compact size allowing flexible mounting orientations, its high dynamic range and spectral response facilitating the use of eye-safe lasers for the capture of 3D images in ambient room lighting, as well as advanced triggering options for synchronized scan data capturing. Several system peripherals were also driven through the camera’s RS232 port for increased functionality. In addition, the GigE Vision output provides a fast and reliable platform for the transfer of image data.
3D Rendition of Patients’ Hands
Both the Prosilica GC1290C and the thermal camera are mounted at the top of the frame, looking down at the target (i.e. the palm rest where the patient will place his/her hand). The two laser scanners and their internal Prosilica GC1290s are set up on the side of the structure at a 45 degree angle.
Each scanner projects a series of 50 different patterns that will help capture image data to create a 3D image of the patient’s hand. Each 3D rendition is captured under five seconds to ensure optimal quality and minimize issues such as blurring due to patient movement.
Once captured the 50 images are post processed by the system’s computer to create a 3D rendition of the patient’s hand based on a sophisticated analysis of the shape of patterns and pattern variations captured.
The software was developed using AVT’s PvAPI Software Development Kit. Thanks to its extensive Dynamic-link Library (DLL), the AVT PvAPI SDK can be developed directly into the application, which means that the system scanners can be deployed "driverless" simplifying both the initial installation and long-term maintenance of client software. In addition the AVT PvAPI SDK is available for a wide variety of operating systems such as Windows, OS X, QNX and Linux. This allowed threeRivers 3D to use a common code repository to deploy scanners on all major operating systems.
The Future for 3D Imaging
Currently the majority of 3D processing is done using manual techniques. With the rapid growth of 3D imaging for industrial uses, threeRivers 3D is currently developing automated 3D custom scanners for the manufacturing, quality control and scientific industries capable of capturing and analyzing data from thousands of 3D scans automatically, greatly reducing labor and increasing throughput and accuracy.