Diffuse fluorescence tomography for early cancer diagnosis
About 6 million new cases of cancer are registered annually in the world. One of the key factors in the fight against this terrible disease is the diagnosis of tumors in the early stages. The use of optical biomedical diagnostics methods in combination with graphic data processing technologies makes it possible to find malignant neoplasms 100 times faster.
Diffuse fluorescence tomography (DFT) is one of the new modern methods of optical diagnosis of tumors. Special fluorescent markers (complex organic molecules) are introduced into the body, which are attached to malignant cells. The illumination of tissues at a certain wavelength causes the fluorescence of markers, and its registration makes it possible to determine the location of the tumor.
The main difficulty of the DFT method is that the light passing through biological tissues is strongly scattered. Therefore, it is impossible to directly see the outlines of the glowing area, especially if it is located at a considerable depth.
As a result of experiments with different locations of illumination and detectors, specialists of the Institute of Applied Physics of the Russian Academy of Sciences have developed special algorithms for reconstructing the three-dimensional distribution of fluorophores in tissues, which make it possible to accurately determine the location and geometry of the tumor.
The Biophotonics Laboratory of the Institute of Applied Physics of the Russian Academy of Sciences (see the list of structural divisions of the Institute) was established more than 15 years ago, and is currently developing methods and creating devices for optical biomedical diagnostics. One of the key areas of ongoing research is the development of optical diffusion tomography and its fluorescent modification.
Alexey Katichev, a junior researcher at the Institute, notes: "In our research, we used the Monte Carlo method (MMC), a method for numerical simulation of radiation propagation in the medium. For all its advantages, this method requires huge computing resources of the system: modeling a typical situation requires calculating about a billion random trajectories! Initially, it took us a considerable amount of time to conduct one experiment – up to several hours. It was unacceptable. The transfer of computing to the architecture of NVIDIA CUDA GPUs gave more than a hundredfold increase in performance. The average time to get the result decreased from two and a half hours to 1.5 minutes. Reducing the calculation time allowed to increase the number of trajectories and, as a result, significantly improve the accuracy of the results."
NVIDIA CUDA is a hardware and software architecture that allows the use of GPU for general purpose computing. The CUDA architecture is provided with broad support for popular languages and application programming interfaces. Thousands of software developers, scientists and researchers widely use CUDA in various fields, including video processing, astrophysics, computational biology and chemistry, modeling of fluid dynamics, electromagnetic interactions, restoration of images obtained by computed tomography, seismic analysis, ray tracing and much more.
The application of this modeling algorithm is not limited to the tasks of diffuse fluorescence tomography. In the future, it is expected to be used in the planning of radiation therapy. It is known that this method is actively used in the treatment of oncological diseases, but it has a number of side effects – with radiation exposure, the irradiation zone is not determined accurately enough, so there is a risk of damaging healthy organs. If you learn to accurately simulate the passage of radiation through the whole body, the procedure will become more effective and safe. Solving such a problem requires huge computing power, which only graphics processors can provide in an acceptable time.
Detailed information about the project can be obtained from its curator Mikhail Kirillin (kirillin{O}ufp.appl.sci-nnov.ru ) or the lead programmer Ilya Fix (FiksII{O}yandex.ru ).
Portal "Eternal youth" http://vechnayamolodost.ru based on NVIDIA Corporation materials
08.02.2011