Small animal imaging using bioluminescent sources has become increasingly important over recent years. The use of bioluminescent sources, such as cells tagged with light-emitting probes, allows detection of gene expression in living cells, which is a well-established technique for non-invasive studies at cellular and molecular levels. In vivo imaging of bioluminescent sources in the living animals (e.g. genetically engineered mice) offers the opportunity to evaluate pathologic progression in a much-compressed time frame and with a much-improved resolution, which is of increasing importance to understand the biologic basis for pathologic disease manifestations. Numerical simulation plays a critical role in this type of bioluminescence-based investigation, because some clinically important parameters cannot be obtained directly from real experiments. Moreover, results and parameters of numerical simulation can be instructive to real experiments.
The Monte Carlo approach is rigorous, flexible and powerful to study photon transport phenomena in turbid tissues, which has been used for both diagnostic and therapeutic applications of lasers and other optical sources. The Monte Carlo simulation is based on the random walks that photons make as they travel through tissues.
MOSE is a forward model for bioluminescent light propagation based on Monte Carlo method in order to simulate bioluminescent phenomena in the mouse imaging and to predict bioluminescent signals around the mouse. All the details of light propagation, such as photon generation, photon absorption, photon scattering, internal reflection and transmission of the photon packet at boundaries and so on, were described. In the simulation, relevant probability distributions are statistically sampled for step size and angular deflection per scattering event. After tracing many photons, such as 1000000 photons, bioluminescent signals detected by CCD camera can be quite well estimated. Moreover, photon absorption and transmittance distributions can be easily obtained. In MOSE, two fundamental assumptions are given as follows. One is that photons are treated as classical particles; the other is that the polarization and wave phenomenon are neglected. Moreover, the optical parameters of biological tissues are macroscopic optical properties.
Currently, there are many Monte Carlo programs for light transport simulation, such as MCML, MCNP, EGS4, and so on. Compared to these popular programs, our MOSE has two unique features. First, our underlying object model can be more complex than that allowed by other software. In the 2D/3D version of the MOSE, ellipses/ellipsoids are used as building blocks, with which it is more accurate and flexible to simulate the real tissues. Therefore, the tissue model of MOSE is more complex than that of other software. Also, bioluminescent sources inside the biological tissues are assumed in our program. Second, our photon generation model is physically more realistic than that of other programs.
To verify MOSE, three experimental comparisons of bioluminescent signals between MOSE and TracePro (a commercial software package for optical analysis of solid models copyrighted by Lambda Research Corporation) are given, which shows a good agreement between the simulated data of MOSE and those of TracePro.
Optical molecular imaging using near-infrared light is very useful to study the development and changes of disease in the biomedical field. Over the past twenty years, optical molecular imaging has attracted more and more attention and made progress with a series of breakthroughs.
The imaging technologies can be divided into two groups: the first is the two-dimensional (2D) planar imaging, and the second is the three-dimensional (3D) tomographic imaging, such as diffuse optical tomography (DOT), fluorescence molecular tomography (FMT), and bioluminescence tomography (BLT). The forward problem of tomographic imaging is to study the light propagation and the inverse problem is to reconstruct the optical properties of the inner tissues or the light sources. There are three distinct technology domains for optical tomography, which are the continuous wave (CW), the time-domain (TD) and the frequency-domain (FD). Each has distinct advantages and disadvantages, and the selection of the appropriate technology depends on the specific application. In order to realize high-fidelity in small-animal imaging, non-contact imaging approaches in free-space were introduced recently compared to the traditional method using light-guiding fibers. Although non-contact imaging has become mainstream, it needs to consider the procedure of light propagation in free-space and makes light propagation research in this medium more difficult.
Molecular Optical Simulation Environment (MOSE) is a simulation platform for optical molecular imaging research co-developed by Xidian University, Institute of Automation, Chinese Academy of Sciences, China and Virginia Tech–Wake Forest University School of Biomedical Engineering & Sciences, USA. MOSE is featured by implementing the simulation of near-infrared light propagation both in a medium with complicated shapes (such as a mouse) and in free-space. Up until now, MOSE has accomplished simulation of light propagation both in a medium and in free-space under CW, TD, and FD, thus it is a powerful tool to solve the forward problems in DOT, FMT, and BLT. This manual will help users learn how to use MOSE, and more detailed information will be introduced in the following sections. The solution of the inverse problem remains under investigation and will be added in a future version.
Feature list of MOSE v2.2
1. Support three kinds of forward simulations in optical molecular imaging: BLT, DOT, and FMT. The simulation algorithms are based on the MC method.
2. Support three kinds of simulation modes: CW, TD, and FD.
3. Support the description of the medium with a regular shape (ellipse, rectangle) under 2D, regular shape (ellipsoid, cylinder, cube) under 3D and irregular shape (the boundary is described by a triangle mesh; for example, data in PLY, OFF, SURF, MESH, and AM formats), and it’s helpful for the users to implement the MC simulation under a complex medium.
4. Support the simulation of light propagation in free-space to the panel detector (such as CCD) under CW. The algorithm is based on the theory of pinhole imaging and Lambert’s cosine law.
5. Support the function of mapping the 2D detection results from a multi-angle to the 3D surface of the complex medium.
6. Support Windows and Linux systems.
7. Support multi-thread optical simulation using the OpenMP API.
8. Absorption results can be saved as photon absorption or photon flux density.
9. Support threshold extraction of the medical image data. Users can extract the boundaries of different organs based on the threshold setting (the data can be used in the MC simulation).
10. Support simplification of the trianglar mesh, where it can reduce the mesh number effectively.
11. Support boundary extraction of the tetrahedral mesh.
12. A rich image display mainly includes:
1) The display properties of the tissues can be set independently, including color, transparency, solid/wireframe display, and whether or not to use a display.
2) There are two ways to show simulation results (absorption and transmittance results): Point rendering-based and surface rendering-based.
3) It can display multiple layers (parallel to the XY, YZ or XZ plane) of the absorption results simultaneously which is helpful for users to quickly analyze the simulation results.
13. Support the input and output of the parameter and result files under various types of simulation. All of the files are managed by the project file (.MSE).