Search

A method to quantitatively evaluate the performances of a radiographic detection system consists in measuring the contrast, noise and modulation transfer functions. These functions have been evaluated for a digital radiographic system based on a single photon counting pixel detector. The X-ray detector is a Silicon sensor with one side segmented in a matrix of 256 by 256 square contacts with a pitch of 55 [micro]m. The active area is about 2 [cm.sup.2] The sensor is connected to the Medipix2 read-out chip by bump-bonding. As X-ray source we have used a tube for general radiography. To reproduce the conditions of a radiographic examination, a 4 cm thick lucite block positioned above the detector has been used to simulate a tissue sample. To study the Contrast Transfer Function we have measured the contrast of an 1 mm thick lead slab with respect to the background. To evaluate the scattering contribution from the lucite, the measurements have been performed with and without a collimator placed at the beam exit. To assess the efficiency and noise transfer properties, we have measured the Detective Quantum Efficiency (DQE) of the detector as a function of the tube voltage. The Modulation Transfer Function has been measured applying the slit method for different conditions of tube voltage and energy threshold. Index Terms--Contrast transfer function (CTF), detective quantum efficiency (DQE), digital radiography, medical imaging equipment, modulation transfer function (MTF), solid-state detectors.

The mammography is the most effective procedure for an early diagnosis of the breast cancer. In this paper, an algorithm for detecting masses in mammographic images will be presented. The database consists of 3762 digital images acquired in several hospitals belonging to the MAGIC-5 collaboration (Medical Applications on a Grid Infrastructure Connection). A reduction of the whole image's area under investigation is achieved through a segmentation process, by means of a ROI Hunter algorithm, without loss of meaningful information. In the following classification step, feature extraction plays a fundamental role: some features give geometrical information, other ones provide shape parameters. Once the features are computed for each ROI, they are used as inputs to a supervised neural network with momentum. The output neuron provides the probability that the ROI is pathological or not. Results are provided in terms of ROC and FROC curves: the area under the ROC curve was found to be Az = 0.862 [+ or -] 0.007, and we get a 2.8 FP/Image at a sensitivity of 82%. This software is included in the CAD station actually working in the hospitals belonging to the MAGIC-5 Collaboration. Index Terms--Breast cancer, image processing, mammography, neural network.