On image quality metrics and the usefulness of grids in digital mammography

History of Mammography: Analysis of Breast Imaging Diagnostic Achievements over the Last Century

Breast cancer is the most common forms of cancer and a leading cause of mortality in women. Early and correct diagnosis is, therefore, essential to save lives. The development of diagnostic imaging applied to the breast has been impressive in recent years and the most used diagnostic test in the world is mammography, a low-dose X-ray technique used for imaging the breast. In the first half of the 20th century, the diagnosis was in practice only clinical, with consequent diagnostic delay and an unfavorable prognosis in the short term. The rise of organized mammography screening has led to a remarkable reduction in mortality through the early detection of breast malignancies. This historical review aims to offer a complete panorama of the development of mammography and breast imaging during the last century. Through this study, we want to understand the foundations of the pillar of radiology applied to the breast through to the most modern applications such as contrast-enhanced mammography (CEM), artificial intelligence, and radiomics. Understanding the history of the development of diagnostic imaging applied to the breast can help us understand how to better direct our efforts toward an increasingly personalized and effective diagnostic approach. The ultimate goal of imaging applied to the detection of breast malignancies should be to reduce mortality from this type of disease as much as possible. With this paper, we want to provide detailed documentation of the main steps in the evolution of breast imaging for the diagnosis of breast neoplasms; we also want to open up new scenarios where the possible current and future applications of imaging are aimed at being more precise and personalized.

Image quality assessment metric for frame accumulated image

The medical image quality determines the accuracy of diagnosis, and the gray-scale resolution is an important parameter to measure image quality. But current objective metrics are not very suitable for assessing medical images obtained by frame accumulation technology. Little attention was paid to the gray-scale resolution, basically based on spatial resolution and limited to the 256 level gray scale of the existing display device. Thus, this paper proposes a metric, "mean signal-to-noise ratio" (MSNR) based on signal-to-noise in order to be more reasonable to evaluate frame accumulated medical image quality. We demonstrate its potential application through a series of images under a constant illumination signal. Here, the mean image of enough images was regarded as the reference image. Several groups of images by different frame accumulation and their MSNR were calculated. The results of the experiment show that, compared with other quality assessment methods, the metric is simpler, more effective, and more suitable for assessing frame accumulated images that surpass the gray scale and precision of the original image.

Bayesian framework inspired no-reference region-of-interest quality measure for brain MRI images

We describe a postacquisition, attribute-based quality assessment method for brain magnetic resonance imaging (MRI) images. It is based on the application of Bayes theory to the relationship between entropy and image quality attributes. The entropy feature image of a slice is segmented into low- and high-entropy regions. For each entropy region, there are three separate observations of contrast, standard deviation, and sharpness quality attributes. A quality index for a quality attribute is the posterior probability of an entropy region given any corresponding region in a feature image where quality attribute is observed. Prior belief in each entropy region is determined from normalized total clique potential (TCP) energy of the slice. For TCP below the predefined threshold, the prior probability for a region is determined by deviation of its percentage composition in the slice from a standard normal distribution built from 250 MRI volume data provided by Alzheimer's Disease Neuroimaging Initiative. For TCP above the threshold, the prior is computed using a mathematical model that describes the TCP-noise level relationship in brain MRI images. Our proposed method assesses the image quality of each entropy region and the global image. Experimental results demonstrate good correlation with subjective opinions of radiologists for different types and levels of quality distortions.