Yapay Zeka Hızlı Bakış & Çıkarımlar
AI InsightGeneralized image reconstruction algorithms are using mathematical inverse problem solution methods which might be back-projection method as an example of algebraic reconstruction technique (ART), regularization methods as an example of Tikhonov-Morozov discrepancy method, or sub-space methods such as conjugated gradient (CG) methods.
📌 3 Maddede Bu Makale
Abstract
Diffuse optic tomography (DOT) techniques are part of the classification of molecular biomedical optic imaging modality. DOT has the device instrumentation and mathematical image reconstruction parties. In device instrumentation part, electronic, optic and mechanic combinations are prepared for laser data acquisition processes. Basically, DOT devices have light source and photo detector units. Depend on the source and detector placement on imaging tissue surface, device geometry may be transmission through, back-reflected, cylindirical ring or spherical. Laser sources are illumination devices. Laser photons with specific wavelength are sent through tissue from tissue surface. Depend on the molecules’ biochemical structure, laser wavelength can be selected application specific. Molecules have different absorption coefficients depend on the laser wavelength. Detector units can be semiconductor PIN photodiodes, CCD or CMOS imagers. Different approaches can be used for geometrical DOT source-detector placements such that transmission through, back-reflected, cylindirical ring or spherical models. For instance, multi-sources and detectors might be placed like chessboard shape for back-reflected tissue imaging geometry. DOT imaging modality is also divided into three major branches depend on the run mode principle. Continuous Wave (CW), Frequency Domain (FD), and Time Resolved (TR) techniques are using different laser sources. CW technique is using steady state laser source. FD technique is using wide frequency range. TR technique is using picosecond (ps) or femtosecond (fs) pulsed laser source. All of these techniques are trying to investigate tissue molecule concentrations and spatial distributions by using acquired data in image reconstruction algorithm. Generalized image reconstruction algorithms are using mathematical inverse problem solution methods which might be back-projection method as an example of algebraic reconstruction technique (ART), regularization methods as an example of Tikhonov-Morozov discrepancy method, or sub-space methods such as conjugated gradient (CG) methods.