A new philosophy of medical imaging

In general, the traditional approach to medical imaging is based on the solution of the inverse problem of deducing the characteristics of tissues within the body from the received field resulting from probing radiation. Ambiguities and lack of complete data, and physical limitations such as diffraction, field non-uniformity and so on, prevent the image from being an exact representation of what would be seen if the imaged part of the patient were to be exposed to direct vision or drawn by an artist. Much more exact representation could be produced, however, if the philosophy of imaging were to be changed to involve the solution of the forward problem in which the received field is iteratively compared with that calculated to be produced by a computer-simulated model of the object from a knowledge base of anatomy, pathology, histology, physical properties of tissues etc. As a result of this converging process of comparison and iterative minimisation of the difference signal, a model of the object would be simulated consistent with the received field resulting from probing radiation as well as with all the relevant information form a knowledge base containing general knowledge and also including the personal records of the particular patient. Using such a model, any required cross-section or three-dimensional pictorial representation of the region of interest in the investigated object could be displayed in the form of a pictorial image. Thus the proposed philosophy would make medical imaging much less subjective and increase its information value, because inherently it includes all relevant biomedical data in the displayed image.