Process for obtaining three-dimensional images of an object and its application to the tomography of an organ

A process for obtaining three-dimensional images of an object from radiation emitted by a tracer contained in said object, said radiation being received by certain detection areas of a planar detector, designated by numbers n in the plane of said detector, the latter performing a rotary movement around the object, so that it constantly remains tangential to an imaginary cylinder of revolution (C), whose axis corresponds to an axis of the object, the angular positions of the planar detector being designated by angles .theta..sub.p between a reference plane (PR) passing through the axis of the object and a plane passing through the axis of the object and perpendicular to the plane of the detector, the object being assumed as broken down into elementary cells for the emission of radiation designated by their coordinates in a fixed reference mark (i, j, k), axis k of said mark being parallel to the axis of the object, the radiation received by certain detection areas being those which are emitted by the elementary cells of the object in a direction perpendicular to the plane of the detector, the detection areas supplying, for each angular position .theta..sub.p, electrical signals whose amplitudes have values dependent on the intensities of the radiation received, said amplitude values being stored in order to obtain the images of the object, in the sectional planes, of ordinates k perpendicular to the axis of the object, wherein, for obtaining images of the object in the sectional planes, it comprises:

• producing on the basis of the stored amplitude values, for each sectional plane of ordinates k, for all the angular positions .theta..sub.p, and for all the detection areas of number n receiving the radiation, correction values of said amplitude values, whereby these correction values are obtained on initialization, on the basis of simulated amplitude values, corresponding to a simulated image of the object in which it is assumed that the distribution of the tracer is uniform, and from stored correction factors relating to the attenuation (u) of the radiation in the object, as a function of coordinates (i, j) of each elementary cell and relating to the geometry of each detector;
• determining and storing for each sectional plane, for all the angular positions (.theta..sub.p) and for all the detection areas of numbers n which receive the radiation, the values of the respective differences between the values of the amplitudes of the signals supplied by these areas and the correction values, the images corresponding to the values of the amplitudes of the signals supplied by the detection areas being available for the sectional planes for which the values of the differences comply with a predetermined convergence criterion, then, when the values of the differences do not comply with the predetermined criterion, the process then comprises:
• producing other correction values of the signal amplitude values, for the sectional planes of ordinates k, for all the angular positions .theta..sub.p and for all the detection areas of numbers n which receive the radiation, said other correction values being obtained from values of stored differences, on the basis of said correction factors and on the basis of stored filtering coefficients (.lambda.) of said other correction values;
• performing for each sectional plane of ordinates k, for all the angular positions (.theta.p) and for all the detection areas receiving the radiation, additions between respectively the stored amplitude values and the other correction values, the values of the results of these additions being stored, in order to supply the corresponding images of the object in the sectional planes or ordinates k or to obtain new correction values on the basis of the results of said additions until these results converge.