Richard Freyer

Fast image reconstruction for optical absorption tomography in media with radially symmetric boundaries

In this paper we present a reconstruction algorithm to invert the linearized problem in optical absorption tomography for objects with radially symmetric boundaries. This is a relevant geometry for functional volume imaging of body regions that are sensitive to ionizing radiation, e.g., breast and testis. From the principles of diffuse light propagation in scattering media we derive the governing integral equations describing the effects of absorption variations on changes in the measurement data. Expansion of these equations into a Neumann series and truncation of higher-order terms yields the linearized forward imaging operator. For the proposed geometry we utilize an invariance property of this operator, which greatly reduces the problem dimensionality. This allows us to compute the inverse by singular value decomposition and consequently to apply regularization techniques based on the knowledge of the singular value spectrum. The inversion algorithm is highly efficient computing slice images as fast as convolution-backprojection algorithms in computed tomography (CT). To demonstrate the capacity of the inversion scheme we present reconstruction results for synthetic and phantom measurement data.

Volume image reconstruction for diffuse optical tomography

Optical tomography is a potential diagnostic method for visualizing optical properties of tissues in vivo. We present an optical tomography method that has been designed for imaging of the human testes, particularly for spectroscopic tumor differentiation. In this application we need to compute three-dimensional distributions of the optical contrast (absorption coefficient) in the tissue in real time. Thus we have given special care to elaborate an efficient inverse algorithm that takes the limitations of spatial resolution and data space point density into account. Our inverse solution is based on a linearization approach and a dedicated object space discretization. Furthermore, we introduce the concept of fuzzy voxels, which enables a reconstruction-inherent image smoothing.

Development of a digital signal processor-based new 12-lead synchronization electrocardiogram automatic analysis system

This paper presents a digital signal processor (DSP)-based new multichannel electrocardiogram (ECG) system for 12-lead synchronization ECG automatic analysis in real-time with high sampling rate at 1000 Hz and 12-bits precision. Using the hardware structure of double-CPU based on Microprocessor (MPU) 89C55 and DSP TMS320F206 combines the powerful control ability of MPU with DSPs fast computation ability. Fully utilizing the double-CPUs resource, the system can distribute the reasonable CPU-time for the real-time tasks of multichannel synchronization ECG sampling, digital filter, data storing, waveform automatic analysis and print at high sampling rate. The digital ECG system has the advantages of simple structure, sampling with high speed and precision, powerful real-time processing ability and good quality. The paper discusses the systems principle and the skilful hardware design, also gives the ECG processing using the fast simple integer-coefficient filter method and the automatic calculation algorithms of the ECG parameters such as heart rate, P-R interval, Q-T interval and deflexion angle of ECG-axis etc. The system had been successfully tested and used in the ECG automatic analysis instrument.

Clinical NIR spectroscopy and optical tomography of the testis

Optical tomography and NIR spectroscopy are potential methods to improve the diagnosis of testicular pathologies. To evaluate the methods clinically we developed a special measurement device with the capability of spatially resolved laser spectroscopy and optical tomography of the testis. Simple spectroscopy is primarily used to obtain global tissue optical properties of the testis and to find correlations of optical parameters with type and stage of certain pathologies. Optical tomography is applied to visualize spectral contrasts in limited tissue volumes, such as tumors. In the course of the study we will determine whether NIR techniques posses the required specifity and sensitivity to give additional quantitative information about tissue perfusion parameters and to serve for a tumor differentiation.

Optical tomography of rat brain

An experimental set-up for the optical tomography using photon density waves is described. The intensity of a NIR laser diode (825 nm) is modulated with a frequency of 110 MHz (modulating degree equal to 80%). The line scan of a rat brain at 64 linear steps and 64 angular steps contains the ac attenuation and the phase shift separately. Due to the impact of scattering in the medium these data are not the basis for a backprojection and deconvolution for the tomographic reconstruction itself. The first approach for a preprocessing consists in the determination of the geometry dependent modulation transfer function (MTF). The line scan is then corrected by the filter which varies with the position. An optical tomogram of a rat brain is presented. This procedure is suitable for biological objects up to a diameter of 30 mm approximately, for example small finger joints.

Image segmentation with constraint satisfaction synergetic potential network

By the interpretation of the segmentation process as an assignment of labels to objects dependent on spatial constraints, image segmentation can be described as a constraint satisfaction problem (CSP). Starting from this model, a new technique for the segmentation of medical images is presented: the constraint satisfaction synergetic potential network (CSSPN). In CSSPN the actually possible labels of an object are represented by singular points of synergetic potential systems. The fuzzy-algorithmic initialization model of the CSSPN allows a label-number-independent dimensioning of the network with n2 nodes. The parallel relaxation dynamics of the CSSPN controlled by interactions of the potential systems will bring selection or evolution of the input image by complete deterministic or stochastically perturbed equations of motion in the potential systems. Constraint functions are significant to the relaxation dynamics and to the result of segmentation within an object adjacency, information of the image model like the image semantics or the optimization strategy of network parameters are mapped onto the CSP with them. Experimental comparative analyses of the segmentation results demonstrate the efficiency of the technique and confirm that the CSSPN is a very promising method for image segmentation.

NIR spectroscopy for the diagnosis of testicular pathologies

NIR spectroscopy is a method principally capable of measuring tissue perfusion and oxygen saturation in subsurface and deeper tissue layers. In urology such perfusion related parameters are of some importance for the differentiation and evaluation of certain types of testicular pathologies. Among them are in the first place the differentiation between inflammations and torsion of the testis, both with similar symptoms, and the assessment of tissue viability in cases of torsion and necrosis, which is sometimes not sufficiently covered by sonography. Although NIR spectroscopy is the method of choice to measure blood oxygen saturation in tissue non- invasively the strong light scattering complicates spectroscopy quantification methods. Spatially-resolved spectroscopy (SRS) is one method to quantify absolute oxygen saturation and relative blood volume. To evaluate this method for in-vivo measurements we have developed a laser scanning device and evaluated quantification algorithms by help of numerical and experimental investigations. As first results suggest the method can in principle quantify absorption differences in tissue, oxygen saturation measurements, however, not work on the testis under the simplified assumptions made for other parts of the body.

Quantitative evaluation of emission-computed tomography images based on fuzzy segmentation

Single photon emission computed tomography (SPECT) performs real 3D functional imaging in nuclear medicine. Even though the visual interpretation of nuclear medicine images by experienced physicians predominates in diagnostics the quantitative image analysis can reduce the subjective observer influences and offers the possibility for the use of objective numerical criteria. An essential problem of the quantitative evaluation of SPECT images is the determination of functional volumes from reconstructed SPECT data. This is the prerequisite for the estimation of the absolute radio nuclide concentration as a high level quantitation. In this paper we present a combined technique for the determination of functional volumes in SPECT data including elements of voxel based clustering methods and edge based fuzzy segmentation.

Fast inversion scheme for the linearized problem in optical absorption tomography on objects with radially symmetric boundaries

We present a reconstruction scheme which solves the inverse linear problem in optical absorption tomography for radially symmetric objects. This is a relevant geometry for optical diagnosis in soft tissues, e.g. breast, testis and even head. The algorithm utilizes an invariance property of the linear imaging operator in homogeneously scattering media. The inverse problem is solved in the Fourier space of the angular component leading to a considerable dimension reduction which allows to compute the inverse in a direct way using singular value decomposition. There are two major advantages of this approach. First the inverse operator can be stored in computer memory and the computation of the inverse problem comprises only a few matrix multiplications. This makes the algorithm very fast and suitable for parallel execution. On the other hand we obtain the spectrum of the imaging operator that allows conclusions about reconstruction limits in the presence of noise and gives a termination criterion for image synthesis. To demonstrate the capabilities of this scheme reconstruction results from synthetic and phantom data are presented.

Detection of local inhomogeneities in scattering media using tomographic reconstruction techniques

Investigations on photon migration in strongly scattering media have shown that due to the practical bandlimitation of the medias transfer characteristic a complete reconstruction of size, shape, and optical parameters of embedded inhomogeneities is generally impossible. However, using a priori information can considerably reduce the solution space. In this article we present a method to determine the position of an inhomogeneity in a cylindrical scattering object by backprojecting the changes in measurement value. For the backprojection a set of trajectories is used approximating the maximum probability photon path. Reconstruction methods as known from computed tomography (CT) are used to obtain a visually improved slice image giving the opportunity to estimate size and shape of the inhomogeneities. Although in this article the procedure is described for time-integrated measurements it should be applicable to any other measurement technique. Above all the method is computationally very efficient.