Catálogo de publicaciones - libros

This paper describes the way the signal coming from a radiation detector is conditioned and processed to produce images useful for medical applications. First of all, the small signal produce by the radiation is processed by analogue electronics specifically designed to produce a good signal-over-noise ratio. The optimised analogue signal produced at this stage can then be processed and transformed into digital information that is eventually stored in a computer, where it can be further processed as required. After an introduction to the general requirements of the processing electronics, we will review the basic building blocks that process the ‘tiny’ analogue signal coming from a radiation detector. We will in particular analyse how it is possible to optimise the signal-over-noise ratio of the electronics. Some exercises, developed in the tutorial, will help to understand this fundamental part. The blocks needed to process the analogue signal and transform it into a digital code will be described. The description of electronics systems used for medical imaging systems will conclude the lecture.

Positron (PET) and Single photon tomography (SPECT) are both methods used to acquire data in nuclear medicine from the detection of gamma photons. While SPECT is normally performed by rotating a device such as one or more gamma cameras around a patient, conventional PET devices are normally composed of rings of individual detectors, such as small BGO crystals. The total number of such detectors can be very high, and requires clever electronic design to reduce the total number of components. In both PET and SPECT the data recorded must be reconstructed to form sets of transaxial slices, after various types of correction to remove sources of degradation such as scatter etc. Filtered backprojection as originally employed in both PET and SPECT is now being widely superseded by iterative techniques such as OSEM, which are increasing being operated in true 3D.

The use of computers in radioisotope imaging is now well established for data acquisition, data correction, image reconstruction, image display and manipulation, data storage, system control and multimodal imaging and registration. This paper reviews these areas highlighting some of the applications and demands on computing facilities.

In this paper we describe an overarching framework that allows us to interpret the information from an image. Starting with the imaging device, consideration will be given to the measurement of the quality of the raw data detected by the instrument. This uses Bayesian signal detection theory to combine the large area transfer characteristic, the modulation transfer function and the noise power spectrum in a single measure of quality.Then we will discuss how to assess the quality of the displayed image by measuring human performance directly. The most complete description of observer performance is provided by Receiver Operating Characteristic (ROC) analysis, which estimates all the combinations of sensitivity and specificity available from an imaging procedure. How subjective measures of image quality can be combined with the objective assessment of performance will be investigated. Finally, we will show how quality can be extended to the judgement of the influence of imaging technology on the clinical management of patients, including the quality of life of the patient.

There has been considerable progress in nuclear medicine instrumentation for both Positron Emission Tomography (PET) and Single Photon Emission Computerised Tomography (SPECT). The purpose of this article is to indicate areas of advance and progress and to makesome predictions of what nuclear medicine acquisition and processing systems might look like in future. This subject includes new detectors and detectors systems, progress in tomographic reconstruction, in the correction of errors associated with the data such as uniformity, scatter, attenuation, motion and the partial volume effect.

The research for the identification and development of new drugs represents a very complex process implying long times and massive investments. This process was not able to parallel the rate of discoveries made in the field of genomic and molecular biology and a gap created between demand of new drugs and the ability of pharmaceutical companies to select good candidates. Positron Emission Tomography, among the different Molecular Imaging modalities, could represent a new tool for the early assessment and screening of new drug candidates and, due to its physical performances and the characteristics of positron-labeled tracers, gain the role of “Biomarker” accepted by the Companies and the Regulatory Bodies of Drug Agencies. To fulfil this task PET has to exploit all of its special features such as data absolute quantification and modelling, high spatial resolution and dynamic imaging. Relevant efforts need to be directed to the careful design and validation of experimental protocols with the main goal of achieving consistency in multi- centric trials.

An outline is given of the underlying physical principles that govern the selection and use of systems for X-ray mammography. Particular attention is paid to screen-film mammography as some aspects of digital mammography are considered in another lecture. The size and composition of the compressed female breast and of calcifications are described and the magnitude of photon interaction processes in breast tissues discussed. The physical performance measures contrast, unsharpness, dose, noise and dynamic range are outlined and used in a treatment of the various components of the mammographic system. The selection of photon energy is a compromise between contrast and/or signal-to-noise ratio on the one hand, and breast dose on the other. For screen-film imaging the contrast achieved is considered to be the most important image measure and the performances of different mammographic target/filter combinations (including Mo/Mo, Mo/Rh, Rh/Rh and W/Rh) are compared on this basis. For digital imaging, the signal-tonoise ratio is the most important image measure, and the optimal X-ray spectra are then different to those for screen-film mammography. The relationship between image unsharpness and focal spot size and image magnification is explored. The importance of breast compression is stressed and the advantages of compression listed. The contrast in the image is degraded by scattered photons recorded by the image receptor and the magnitude of this effect and the reduction achievable using mammographic anti-scatter grids considered. The performance of mammographic screen-film receptors is described and analyzed, paying attention to unsharpness, noise and receptor DQE.

After its clinical introduction in 1973, computed tomography developed from an x-ray modality for axial imaging in neuroradiology into a versatile three dimensional imaging modality for a wide range of applications in for example oncology, vascular radiology, cardiology, traumatology and even in interventional radiology. Computed tomography is applied for diagnosis, follow-up studies and screening of healthy subpopulations with specific risk factors. This chapter provides a general introduction in computed tomography, covering a short history of computed tomography, technology, image quality, dosimetry, room shielding, quality control and quality criteria.

The principles of image reconstruction in positron emission tomography will be presented in the lecture. The filtered backprojection algorithm will be explained in detail for 2D reconstruction. The generalization of the algorithm in 3D will be described. A brief introduction to iterative reconstruction methods will be given, and their advantages and disadvantages against the filtered backprojection algorithm will be discussed.

In mammography to obtain the best image quality at the lower possible dose to minimize the risk, implementation of an effective quality control protocol is of the utmost importance. This paper will mainly review quality control (QC) procedures for screen-film mammography and will briefly present how screen-film and digital differ. Finally, the possibility to automate some tests in digital mammography will be introduced.