The effects of gray scale image processing on digital mammography interpretation performance

Early signs of cancer present in the fine detail of mammograms

The gist of abnormality can be rapidly extracted by medical experts from global information in medical images, such as mammograms, to identify abnormal mammograms with above-chance accuracy-even before any abnormalities are localizable. The current study evaluated the effect of different high-pass filters on expert radiologists' performance in detecting the gist of abnormality in mammograms, especially those acquired prior to any visibly actionable lesions. Thirty-four expert radiologists viewed unaltered and high-pass filtered versions of normal and abnormal mammograms. Abnormal mammograms consisted of obvious abnormalities, subtle abnormalities, and currently normal mammograms from women who would go to develop cancer in 2-3 years. Four levels of high-pass filtering were tested (0.5, 1, 1.5, and 2 cycles per degree (cpd) after brightening and contrast normalizing to the unfiltered mammograms. Overall performance for 0.5 and 1.5 did not change compared to unfiltered but was reduced for 1 and 2 cpd. Critically, filtering that eliminated frequencies below 0.5 and 1.5 cpd significantly boosted performance on mammograms acquired years prior appearance of localizable abnormalities. Filtering at 0.5 did not change the radiologist's decision criteria compared to unfiltered mammograms whereas other filters resulted in more conservative ratings. The findings bring us closer to identifying the characteristics of the gist of the abnormal that affords radiologists detection of the earliest signs of cancer. A 0.5 cpd high-pass filter significantly boosts subtle, global signals of future cancerous abnormalities, potentially providing an image enhancement strategy for rapid assessment of impending cancer risk.

Radiation Dose Reduction in Digital Mammography by Deep-Learning Algorithm Image Reconstruction: A Preliminary Study

Purpose

To develop a denoising convolutional neural network-based image processing technique and investigate its efficacy in diagnosing breast cancer using low-dose mammography imaging.

Materials and Methods

A total of 6 breast radiologists were included in this prospective study. All radiologists independently evaluated low-dose images for lesion detection and rated them for diagnostic quality using a qualitative scale. After application of the denoising network, the same radiologists evaluated lesion detectability and image quality. For clinical application, a consensus on lesion type and localization on preoperative mammographic examinations of breast cancer patients was reached after discussion. Thereafter, coded low-dose, reconstructed full-dose, and full-dose images were presented and assessed in a random order.

Results

Lesions on 40% reconstructed full-dose images were better perceived when compared with low-dose images of mastectomy specimens as a reference. In clinical application, as compared to 40% reconstructed images, higher values were given on full-dose images for resolution (p < 0.001); diagnostic quality for calcifications (p < 0.001); and for masses, asymmetry, or architectural distortion (p = 0.037). The 40% reconstructed images showed comparable values to 100% full-dose images for overall quality (p = 0.547), lesion visibility (p = 0.120), and contrast (p = 0.083), without significant differences.

Conclusion

Effective denoising and image reconstruction processing techniques can enable breast cancer diagnosis with substantial radiation dose reduction.

How does image quality affect radiologists' perceived ability for image interpretation and lesion detection in digital mammography?

OBJECTIVES

To study how radiologists' perceived ability to interpret digital mammography (DM) images is affected by decreases in image quality.

METHODS

One view from 45 DM cases (including 30 cancers) was degraded to six levels each of two acquisition-related issues (lower spatial resolution and increased quantum noise) and three post-processing-related issues (lower and higher contrast and increased correlated noise) seen during clinical evaluation of DM systems. The images were shown to fifteen breast screening radiologists from five countries. Aware of lesion location, the radiologists selected the most-degraded mammogram (indexed from 1 (reference) to 7 (most degraded)) they still felt was acceptable for interpretation. The median selected index, per degradation type, was calculated separately for calcification and soft tissue (including normal) cases. Using the two-sided, non-parametric Mann-Whitney test, the median indices for each case and degradation type were compared.

RESULTS

Radiologists were not tolerant to increases (medians: 1.5 (calcifications) and 2 (soft tissue)) or decreases (median: 2, for both types) in contrast, but were more tolerant to correlated noise (median: 3, for both types). Increases in quantum noise were tolerated more for calcifications than for soft tissue cases (medians: 3 vs. 4, p = 0.02). Spatial resolution losses were considered less acceptable for calcification detection than for soft tissue cases (medians: 3.5 vs. 5, p = 0.001).

CONCLUSIONS

Perceived ability of radiologists for image interpretation in DM was affected not only by image acquisition-related issues but also by image post-processing issues, and some of those issues affected calcification cases more than soft tissue cases.

KEY POINTS

• Lower spatial resolution and increased quantum noise affected the radiologists' perceived ability to interpret calcification cases more than soft tissue lesion or normal cases. • Post-acquisition image processing-related effects, not only image acquisition-related effects, also impact the perceived ability of radiologists to interpret images and detect lesions. • In addition to current practices, post-acquisition image processing-related effects need to also be considered during the testing and evaluation of digital mammography systems.

Validation of a candidate instrument to assess image quality in digital mammography using ROC analysis

PURPOSE

To validate a candidate instrument, to be used by different professionals to assess image quality in digital mammography (DM), against detection performance results.

METHODS

A receiver operating characteristics (ROC) study was conducted to assess the detection performance in DM images with four different image quality levels due to different quality issues. Fourteen expert breast radiologists from five countries assessed a set of 80 DM cases, containing 60 lesions (40 cancers, 20 benign findings) and 20 normal cases. A visual grading analysis (VGA) study using a previously-described candidate instrument was conducted to evaluate a subset of 25 of the images used in the ROC study. Eight radiologists that had participated in the ROC study, and seven expert breast-imaging physicists, evaluated this subset. The VGA score (VGAS) and the ROC and visual grading characteristics (VGC) areas under the curve (AUCROC and AUCVGC) were compared.

RESULTS

No large differences in image quality among the four levels were detected by either ROC or VGA studies. However, the ranking of the four levels was consistent: level 1 (partial AUCROC: 0.070, VGAS: 6.77) performed better than levels 2 (0.066, 6.15), 3 (0.061, 5.82), and 4 (0.062, 5.37). Similarity between radiologists' and physicists' assessments was found (average VGAS difference of 10 %).

CONCLUSIONS

The results from the candidate instrument were found to correlate with those from ROC analysis, when used by either observer group. Therefore, it may be used by different professionals, such as radiologists, radiographers, and physicists, to assess clinically-relevant image quality variations in DM.

Key technologies of intelligent transportation based on image recognition

With the development of economy, the research of urban intelligent transportation system is becoming more and more important. The research and development of plate number recognition system is an important factor to realize the intelligence and modernization of transportation system. It uses each car to have a unique plate number and recognizes the vehicle number through the vehicle image captured by the camera. On the basis of image recognition, this article takes plate number image as the research object and discusses the key technologies of plate number recognition system. First, this article uses image preprocessing technology to process images to improve image quality. Second, the plate number location algorithm based on the connected region search is analyzed. According to the characteristics of the plate number itself, the regional features of the plate number are extracted to locate the plate number accurately. Then, an improved vertical projection-based plate number character segmentation method is proposed to segment plate number characters. Finally, combined with character characteristics, the template matching method is used to recognize plate number characters. The simulation results show that, on the basis of image recognition, this article studies the key technologies of plate number recognition system, which effectively improves the performance of the system and makes the recognition of plate number more effective and accurate.

Artificial Intelligence for Mammography and Digital Breast Tomosynthesis: Current Concepts and Future Perspectives

Although computer-aided diagnosis (CAD) is widely used in mammography, conventional CAD programs that use prompts to indicate potential cancers on the mammograms have not led to an improvement in diagnostic accuracy. Because of the advances in machine learning, especially with use of deep (multilayered) convolutional neural networks, artificial intelligence has undergone a transformation that has improved the quality of the predictions of the models. Recently, such deep learning algorithms have been applied to mammography and digital breast tomosynthesis (DBT). In this review, the authors explain how deep learning works in the context of mammography and DBT and define the important technical challenges. Subsequently, they discuss the current status and future perspectives of artificial intelligence-based clinical applications for mammography, DBT, and radiomics. Available algorithms are advanced and approach the performance of radiologists-especially for cancer detection and risk prediction at mammography. However, clinical validation is largely lacking, and it is not clear how the power of deep learning should be used to optimize practice. Further development of deep learning models is necessary for DBT, and this requires collection of larger databases. It is expected that deep learning will eventually have an important role in DBT, including the generation of synthetic images.

Consistency of visual assessments of mammographic breast density from vendor-specific "for presentation" images

Discussions of percent breast density (PD) and breast cancer risk implicitly assume that visual assessments of PD are comparable between vendors despite differences in technology and display algorithms. This study examines the extent to which visual assessments of PD differ between mammograms acquired from two vendors. Pairs of "for presentation" digital mammography images were obtained from two mammography units for 146 women who had a screening mammogram on one vendor unit followed by a diagnostic mammogram on a different vendor unit. Four radiologists independently visually assessed PD from single left mediolateral oblique view images from the two vendors. Analysis of variance, intra-class correlation coefficients (ICC), scatter plots, and Bland-Altman plots were used to evaluate PD assessments between vendors. The mean radiologist PD for each image was used as a consensus PD measure. Overall agreement of the PD assessments was excellent between the two vendors with an ICC of 0.95 (95% confidence interval: 0.93 to 0.97). Bland-Altman plots demonstrated narrow upper and lower limits of agreement between the vendors with only a small bias (2.3 percentage points). The results of this study support the assumption that visual assessment of PD is consistent across mammography vendors despite vendor-specific appearances of "for presentation" images.

Effect of image quality on calcification detection in digital mammography

PURPOSE

This study aims to investigate if microcalcification detection varies significantly when mammographic images are acquired using different image qualities, including: different detectors, dose levels, and different image processing algorithms. An additional aim was to determine how the standard European method of measuring image quality using threshold gold thickness measured with a CDMAM phantom and the associated limits in current EU guidelines relate to calcification detection.

METHODS

One hundred and sixty two normal breast images were acquired on an amorphous selenium direct digital (DR) system. Microcalcification clusters extracted from magnified images of slices of mastectomies were electronically inserted into half of the images. The calcification clusters had a subtle appearance. All images were adjusted using a validated mathematical method to simulate the appearance of images from a computed radiography (CR) imaging system at the same dose, from both systems at half this dose, and from the DR system at quarter this dose. The original 162 images were processed with both Hologic and Agfa (Musica-2) image processing. All other image qualities were processed with Agfa (Musica-2) image processing only. Seven experienced observers marked and rated any identified suspicious regions. Free response operating characteristic (FROC) and ROC analyses were performed on the data. The lesion sensitivity at a nonlesion localization fraction (NLF) of 0.1 was also calculated. Images of the CDMAM mammographic test phantom were acquired using the automatic setting on the DR system. These images were modified to the additional image qualities used in the observer study. The images were analyzed using automated software. In order to assess the relationship between threshold gold thickness and calcification detection a power law was fitted to the data.

RESULTS

There was a significant reduction in calcification detection using CR compared with DR: the alternative FROC (AFROC) area decreased from 0.84 to 0.63 and the ROC area decreased from 0.91 to 0.79 (p < 0.0001). This corresponded to a 30% drop in lesion sensitivity at a NLF equal to 0.1. Detection was also sensitive to the dose used. There was no significant difference in detection between the two image processing algorithms used (p > 0.05). It was additionally found that lower threshold gold thickness from CDMAM analysis implied better cluster detection. The measured threshold gold thickness passed the acceptable limit set in the EU standards for all image qualities except half dose CR. However, calcification detection varied significantly between image qualities. This suggests that the current EU guidelines may need revising.

CONCLUSIONS

Microcalcification detection was found to be sensitive to detector and dose used. Standard measurements of image quality were a good predictor of microcalcification cluster detection.

Comparison of visual grading and free-response ROC analyses for assessment of image-processing algorithms in digital mammography

OBJECTIVE

To compare two methods for assessment of image-processing algorithms in digital mammography: free-response receiver operating characteristic (FROC) for the specific task of microcalcification detection and visual grading analysis (VGA).

METHODS

The FROC study was conducted prior to the VGA study reported here. 200 raw data files of low breast density (Breast Imaging-Reporting and Data System I-II) mammograms (Novation DR, Siemens, Germany)-100 of which abnormal-were processed by four image-processing algorithms: Raffaello (IMS, Bologna, Italy), Sigmoid (Sectra, Linköping, Sweden), and OpView v. 2 and v. 1 (Siemens, Erlangen, Germany). Four radiologists assessed the mammograms for the detection of microcalcifications. 8 months after the FROC study, a subset (200) of the 800 images was reinterpreted by the same radiologists, using the VGA methodology in a side-by-side approach. The VGA grading was based on noise, saturation, contrast, sharpness and confidence with the image in terms of normal structures. Ordinal logistic regression was applied; OpView v. 1 was the reference processing algorithm.

RESULTS

In the FROC study all algorithms performed better than OpView v. 1. From the current VGA study and for confidence with the image, Sigmoid and Raffaello were significantly worse (p<0.001) than OpView v. 1; OpView v. 2 was significantly better (p=0.01). For the image quality criteria, results were mixed; Raffaello and Sigmoid for example were better than OpView v. 1 for sharpness and contrast (although not always significantly).

CONCLUSION

VGA and FROC discordant results should be attributed to the different clinical task addressed.

ADVANCES IN KNOWLEDGE

The method to use for image-processing assessment depends on the clinical task tested.

Increase in perceived case suspiciousness due to local contrast optimisation in digital screening mammography

Objectives

To determine the influence of local contrast optimisation on diagnostic accuracy and perceived suspiciousness of digital screening mammograms.

Methods

Data were collected from a screening region in the Netherlands and consisted of 263 digital screening cases (153 recalled,110 normal). Each case was available twice, once processed with a tissue equalisation (TE) algorithm and once with local contrast optimisation (PV). All cases had digitised previous mammograms. For both algorithms, the probability of malignancy of each finding was scored independently by six screening radiologists. Perceived case suspiciousness was defined as the highest probability of malignancy of all findings of a radiologist within a case. Differences in diagnostic accuracy of the processing algorithms were analysed by comparing the areas under the receiver operating characteristic curves (A_z). Differences in perceived case suspiciousness were analysed using sign tests.

Results

There was no significant difference in A_z (TE: 0.909, PV 0.917, P = 0.46). For all radiologists, perceived case suspiciousness using PV was higher than using TE more often than vice versa (ratio: 1.14–2.12). This was significant (P <0.0083) for four radiologists.

Conclusions

Optimisation of local contrast by image processing may increase perceived case suspiciousness, while diagnostic accuracy may remain similar.

Key Points

• Variations among different image processing algorithms for digital screening mammography are large.

• Current algorithms still aim for optimal local contrast with a low dynamic range.

• Although optimisation of contrast may increase sensitivity, diagnostic accuracy is probably unchanged.

• Increased local contrast may render both normal and abnormal structures more conspicuous.

The effect of Premium View post-processing software on digital mammographic reporting

The aim of this study was to identify the effect of the installation of Premium View post-processing software on our mammographic reporting performance, in particular the effects on our recall rate, biopsy rate and cancer detection rate. The case notes and imaging of all patients discussed at the weekly indeterminate imaging multidisciplinary team meeting were reviewed retrospectively before, immediately after and at a delayed interval following the installation of Premium View post-processing software. Factors recorded included the mammographic abnormality, further investigations and final histology. The indeterminate mammogram rate increased significantly from a baseline of 5.7% (before Premium View) to 8.7% in the time period immediately after the installation of Premium View (p=0.002). The stereotactic biopsy rate also increased from 0.8% to 2.4% (p=0.001), with a significant increase in the overall cancer detection rate from 3.4% to 4.4% (p=0.02). In the follow-up period several months after the installation of Premium View, the indeterminate mammogram rate returned to a level similar to that before Premium View (6%; p=0.7). The stereotactic biopsy rate remained significantly higher at 1.6% (p=0.07), as did the overall cancer detection rate of 5.0% (p=0.003). In conclusion, the use of Premium View may lead to higher cancer detection rates, at the expense of an initial increase in recall rate. Although prospective studies are suggested, this result is of interest in light of the proposed installation of digital mammography across the NHS Breast Screening Programme.

Evaluation of clinical image processing algorithms used in digital mammography

Screening is the only proven approach to reduce the mortality of breast cancer, but significant numbers of breast cancers remain undetected even when all quality assurance guidelines are implemented. With the increasing adoption of digital mammography systems, image processing may be a key factor in the imaging chain. Although to our knowledge statistically significant effects of manufacturer-recommended image processings have not been previously demonstrated, the subjective experience of our radiologists, that the apparent image quality can vary considerably between different algorithms, motivated this study. This article addresses the impact of five such algorithms on the detection of clusters of microcalcifications. A database of unprocessed (raw) images of 200 normal digital mammograms, acquired with the Siemens Novation DR, was collected retrospectively. Realistic simulated microcalcification clusters were inserted in half of the unprocessed images. All unprocessed images were subsequently processed with five manufacturer-recommended image processing algorithms (Agfa Musica 1, IMS Raffaello Mammo 1.2, Sectra Mamea AB Sigmoid, Siemens OPVIEW v2, and Siemens OPVIEW v1). Four breast imaging radiologists were asked to locate and score the clusters in each image on a five point rating scale. The free-response data were analyzed by the jackknife free-response receiver operating characteristic (JAFROC) method and, for comparison, also with the receiver operating characteristic (ROC) method. JAFROC analysis revealed highly significant differences between the image processings (F = 8.51, p < 0.0001), suggesting that image processing strongly impacts the detectability of clusters. Siemens OPVIEW2 and Siemens OPVIEW1 yielded the highest and lowest performances, respectively. ROC analysis of the data also revealed significant differences between the processing but at lower significance (F = 3.47, p = 0.0305) than JAFROC. Both statistical analysis methods revealed that the same six pairs of modalities were significantly different, but the JAFROC confidence intervals were about 32% smaller than ROC confidence intervals. This study shows that image processing has a significant impact on the detection of microcalcifications in digital mammograms. Objective measurements, such as described here, should be used by the manufacturers to select the optimal image processing algorithm.

Technical developments in mammography

The art, science, and technology of mammography have developed steadily over the past 35 y. Mammography is a central tool for diagnosis of symptoms of breast cancer. In addition, periodic screening of asymptomatic women in certain age groups has been clearly demonstrated to contribute to reduction of mortality from breast cancer. Technical improvements have allowed the examination to be carried out at substantially lower radiation dose than was necessary to obtain a good image in the 1970's, while at the same time providing greatly improved contrast, spatial resolution, dynamic range and tissue coverage. Digital mammography overcomes many of the technical limitations inherent in screen-film mammography and has been shown to offer increased accuracy for women under 50 and those with dense breasts. The radiation risk associated with mammography cannot be ignored, however, modern analysis suggests that it is very low, especially compared to the benefits from the exam. Nevertheless, imaging should be conducted with careful attention to efficient use of the radiation. New techniques, currently under development and evaluation, promise to add further to the value of mammography.

Optimal gradation processing parameter for soft-copy reading of digital mammogram: Comparison between the parameter recommended for hard-copy and other parameters

PURPOSE

To investigate optimal gradation processing parameter for soft-copy reading comparing the parameter recommended for hard-copy with other parameters.

MATERIALS AND METHODS

Digital mammograms using a Computed Radiography system were evaluated. The gradation processing parameter recommended by the manufacturer was GA (1.2). We prepared seven parameters that changed the degree of contrast: GA (0.6, 0.8, 1.0, 1.2, 1.4, 1.6, and 1.8). Images of an anthropomorphic breast phantom were displayed on a 5-megapixel liquid crystal display monitor using each parameter. Three readers independently assessed each image, and scored for the following items: intramammary contrast, extramammary contrast, sharpness, and graininess. Total score was calculated in each parameter. We also displayed normal mammograms of nine cases: three with a scattered fibroglandular density, three with a heterogeneously dense breast, and three with an extremely dense breast. These were displayed using GA (1.2) or using parameters with a higher total score than GA (1.2) in the phantom test. Three readers assessed each mammogram as in the phantom test.

RESULTS

In phantom test, GA (1.4), GA (1.6), and GA (1.8) obtained higher scores than GA (1.2). In normal cases tested using these parameters, GA (1.4) obtained the highest score. This was significantly higher than that of GA (1.2) (P=0.004). The score obtained for GA (1.4) was the highest in cases with extremely dense and heterogeneously dense breast tissue, though there was no statistically significant difference.

CONCLUSION

Soft-copy image quality was improved by gradient processing using higher contrast parameter than that routinely used in hard-copy, especially in dense breast cases.

A quantitative method for evaluating the detectability of lesions in digital mammography

This study presents a quantitative method for evaluating the detectability of microcalcifications in digital mammography. Four hundred and twenty microcalcifications (with various morphology, size and contrast), simulated with a previously validated method, were used for the creation of image datasets. Lesions were inserted into 163 regions of interests of 59 selected raw digital mammograms with various anatomical backgrounds and acquired with a Siemens Novation DR. After processing, these composite images were scored by experienced radiologists, who located multiple simulated lesions and rated them under conditions of free-search. For statistical analysis, free-response receiver-operating characteristic curves are plotted; the use of jackknife free-response receiver-operating characteristic method has also been investigated. The main advantage of this methodology is that the exact number of inserted microcalcifications is well known and that the lesions are fully characterised in terms of pathology, size, morphology and peak contrast. A first application has been the evaluation of the effect of anatomical background on microcalcifications detection. Preliminary findings in this study indicate that this method may be a promising tool to evaluate factors that have an influence on the detectability of lesions, such as the clinical processing or the viewing conditions.

Clinical image quality criteria for full field digital mammography: a first practical application

In order to quantify the clinical quality of full-field digital mammography, a set of image quality parameters is developed. The set consisted of 12 image quality criteria and 8 physical characteristics of the image. The first set interrogates the visibility of anatomical structures and typical characteristics of a digital image, such as noise and saturation of dark and white areas. The second set of criteria evaluates contrast, sharpness and confidence with the representation of masses, microcalcifications and the image. The use of these criteria is reported in a retrospective study, in which the impact of dose on the radiological quality of digital mammograms is evaluated. Fifty patients acquired in a low-dose mode were retrieved and compared with 50 patients acquired in a dose mode that was set 41% higher. The dose affects, more than expected, contrast and sharpness of the image, whereas the visibility of the anatomical structures remains unchanged. With these parameters, quantification of the image quality is possible; however, because of subjectivity of the parameters, only intra-observer comparison and evaluation of the individual parameters rather than the overall results are advised. Together with physical tests of image quality, critical radiological evaluation of the quality should be included in the acceptance process of digital mammography.

Issues to consider in converting to digital mammography

This article outlines the reasons that many radiology practices are converting to digital mammography. In addition, it provides basic information about the issues that must be considered in making the transformation. These issues include technical matters regarding image display, storage, and retrieval as well as clinical and ergonomic considerations.

Image quality measurements and metrics in full field digital mammography: an overview

This paper gives an overview of test procedures developed to assess the performance of full field digital mammography systems. We make a distinction between tests of the individual components of the imaging chain and global system tests. Most tests are not yet fully standardised. Where possible, we illustrate the test methodologies on a selenium flat-panel system.