[Development of Auto Dense-breast Classification on Mammography Images Using Image Similarity]

PURPOSE

In Japan, radiologists perform qualitative visual classification to define four categories of mammary gland density. However, an objective estimation of mammary gland density is necessary. To address this, we developed an automatic classification software using image similarity.

METHODS

We prepared 741 cases of mediolateral oblique images (MLO) for evaluation, and they were diagnosed as normal among the mammography images taken at our hospital. Image matching was performed using the evaluation images and an image database for breast density determination. In this study, the image similarity used zero normalized cross-correlation (ZNCC) as an index. In addition, if the breast thickness is less than 30 mm and each breast density category ZNNC has the same value, the category is evaluated on the fat side. We compared the results of qualitative visual classification and automatic classification methods to assess consistency.

RESULTS

The agreement with the subjective breast composition classification was 78.5%, and the weighted kappa coefficient was 0.98. One mismatched case was evaluated on the higher density side with the same ZNCC value between categories and a breast thickness greater than 30 mm.

CONCLUSION

Image similarity provides an excellent estimation of quantification of breast density. This system could contribute to improving the efficiency of the mammography screening system.

Comparison of lesion detection and conspicuity between narrow-angle and wide-angle digital breast tomosynthesis for dense and non-dense breasts

Digital breast tomosynthesis (DBT) has been shown to improve both sensitivity and specificity for breast cancer detection compared to full-field digital mammography. However, its performance could be limited for patients with dense breasts. Clinical DBT systems vary in their system designs, one of which is the acquisition angular range (AR), which leads to varied performance for different imaging tasks. In this study, we aim to compare DBT systems with different AR. We used a previously validated cascaded linear system model to investigate the dependence of in-plane breast structural noise (BSN) and detectability of masses on AR. We conducted a pilot clinical study to compare the lesion conspicuity between clinical DBT systems with the narrowest and the widest AR. Patients called back for diagnostic imaging on suspicious findings were imaged with both narrow-angle (NA) and wide-angle (WA) DBT. We analyzed the BSN for clinical images using noise power spectrum (NPS) analysis. A 5-point Likert scale was used in the reader study to compare the lesion conspicuity. Our theoretical calculation results show that increasing AR leads to reduced BSN and improved mass detectability. The NPS analysis on clinical images shows the lowest BSN for WA DBT. The WA DBT provides better lesion conspicuity for masses and asymmetries and shows a greater advantage for non-microcalcification lesions in dense breasts. The NA DBT provides better characterizations for microcalcifications. The WA DBT can downgrade false-positive findings seen on NA DBT. In conclusion, WA DBT could improve the detection of masses and asymmetries for patients with dense breasts.

Water and lipid content of breast tissue measured by six-wavelength time-domain diffuse optical spectroscopy

SIGNIFICANCE

The water and lipid content of normal breast tissue showed mammary gland characteristics with less influence from the chest wall using six-wavelength time-domain diffuse optical spectroscopy (TD-DOS) in a reflectance geometry.

AIM

To determine the depth sensitivity of a six-wavelength TD-DOS system and evaluate whether the optical parameters in normal breast tissue can distinguish dense breasts from non-dense breasts.

APPROACH

Measurements were performed in normal breast tissue of 37 breast cancer patients. We employed a six-wavelength TD-DOS system to measure the water and lipid content in addition to the hemoglobin concentration. The breast density in mammography and optical parameters were then compared.

RESULTS

The depth sensitivity of the system for water and lipid content was estimated to be ∼15 mm. Our findings suggest that the influence of the chest wall on the water content is weaker than that on the total hemoglobin concentration. In data with evaluation conditions, the water content was significantly higher (p < 0.001) and the lipid content was significantly lower (p < 0.001) in dense breast tissue. The water and lipid content exhibited a high sensitivity and specificity to distinguish dense from non-dense breasts in receiver-operating-characteristic curve analysis.

CONCLUSIONS

With less influence from the chest wall, the water and lipid content of normal breast tissue measured by a reflectance six-wavelength TD-DOS system, together with ultrasonography, can be applied to distinguish dense from non-dense breasts.

[Automatic Quantification of Breast Density from Mammography Using Deep Learning]

BACKGROUND

In the field of breast screening using mammography, announcing to the examinees whether they are dense or not has not been deprecated in Japan. One of the reasons is a shortage of objectivity estimating their dense breast. Our aim is to build a system with deep learning algorithm to calculate and quantify objective breast density automatically.

MATERIAL AND METHOD

Mammography images taken in our institute that were diagnosed as category 1 were collected. Each processed image was transformed into eight-bit grayscale, with the size of 2294 pixels by 1914 pixels. The "base pixel value" was calculated from the fatty area within the breast for each image. The "relative density" was calculated by dividing each pixel value by the base pixel value. Semantic segmentation algorithm was used to automatically segment the area of breast tissue within the mammography image, which was resized to 144 pixels by 120 pixels. By aggregating the relative density within the breast tissue area, the "breast density" was obtained automatically.

RESULT

From each but one mammography image, the breast density was successfully calculated automatically. By defining a dense breast as the breast density being greater than or equal to 30%, the evaluation of the dense breast was consistent with that by a computer and human (76.6%).

CONCLUSION

Deep learning provides an excellent estimation of quantification of breast density. This system could contribute to improve the efficiency of mammography screening system.

The Added Value of Supplemental Breast Ultrasound Screening for Women With Dense Breasts: A Single Center Canadian Experience

PURPOSE

We evaluated the contribution to cancer detection of supplemental breast ultrasound screening in women with dense breasts based on a single center experience by comparing our results with similar programs elsewhere.

METHODS

We performed a retrospective review of handheld sonographer-performed screening ultrasound exams at our academic breast imaging center, from January 1st to December 31st, 2019. Breast density, breast cancer risk factors, BI-RADS assessment, and lesion pathology were reviewed and tallied, followed by derivation of the biopsy rate, breast cancer detection rate, PPV3 and average tumor size. These values were compared to published results of breast screening programs elsewhere.

RESULTS

695 screening breast ultrasounds for women with dense breasts and negative mammograms were performed in 2019. The biopsy rate was 1.3%, breast cancer detection rate was 7 in 1000, PPV3 was 42%, and the average tumor size was 9.0 ± 1.4 mm.

CONCLUSIONS

The first-year data of the breast screening ultrasound program at our practice are promising, demonstrating comparable cancer detection rate, higher PPV3, and similar biopsy rate in those with dense breasts compared with similar programs elsewhere. Longitudinal analysis and larger sample size are required for validation. Comparison of incidence and prevalence screening data is also warranted to elucidate the true value of this program.

Contrast-enhanced spectral mammography without and with a delayed image for diagnosing malignancy among mass lesions in dense breast

Aim of the study

To analyse the diagnostic performance of contrast-enhanced spectral mammography (CESM) based on morphologic and enhancement patterns of mass lesions in dense breast using different protocols: CESM without delayed image and CESM with delayed image.

Material and methods

A total of 151 informed women with suspicious for malignancy mass lesions in dense breast were included in this study. All of them underwent CESM using 2 protocols. A total of 155 lesions were pathomorphologically verified. We analysed morphologic patterns on low-energy (LE) images and recombined images (RI) by defining the shape, margin, and dynamic patterns based on delayed images.

Results

The comparative analysis revealed that the shape and margins on RI were more significant than those on LE images. The dynamic indicators of CESM were found to be highly significant in dense breasts. The correlation between kinetic curve and histological results demonstrated that a persistent type of curve was common for benign lesions, accounting for 15/22 (68.1%); plateau and washout – for malignant lesions, accounting for 24/89 (26.9%) and 61/89 (68.5%), respectively. Delayed image leads to an increase of specificity up to 12.4%, which is statistically significant. The area under the curve (AUC) in CESM with delayed image is larger than that in CESM without delayed image (p < 0.01).

Conclusions

CESM is sensitive for the differential diagnosis of breast lesions. CESM with delayed image has higher specificity than CESM without delayed image. Delayed images with plateau and washout are typical for malignancy.

Mammographic breast density, its changes, and breast cancer risk in premenopausal and postmenopausal women

BACKGROUND

The risk of breast cancer related to changes in breast density over time, including its regression and persistence, remains controversial. The authors investigated the relationship between breast density and its changes over time with the development of breast cancer in premenopausal and postmenopausal women.

METHODS

The current cohort study included 74,249 middle-aged Korean women (aged ≥35 years) who were free of breast cancer at baseline and who underwent repeated screening mammograms. Mammographic breast density was categorized according to the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS). A dense breast was defined as heterogeneously dense or extremely dense, and changes in dense breasts between baseline and subsequent follow-up were classified as none, developed, regressed, or persistent dense breast.

RESULTS

During a median follow-up of 6.1 years (interquartile range, 4.1-8.8 years), a total of 803 incident breast cancers were identified. Baseline breast density was found to be positively associated with incident breast cancer in a dose-response manner, and this association did not significantly differ by menopausal status. The multivariable-adjusted hazard ratios (HRs) for breast cancer comparing "heterogeneously dense" and "extremely dense" categories with the nondense category were 1.96 (95% confidence interval [95% CI], 1.40-2.75) and 2.86 (95% CI, 2.04-4.01), respectively. With respect to changes in dense breasts over time, multivariable-adjusted HRs for breast cancer comparing persistent dense breast with none were 2.37 (95% CI, 1.34-4.21) in premenopausal women and 3.61 (95% CI, 1.78-7.30) in postmenopausal women.

CONCLUSIONS

Both baseline dense breasts and their persistence over time were found to be strongly associated with an increased risk of incident breast cancer in premenopausal and postmenopausal women.

LAY SUMMARY

Both baseline breast density and its changes over time were found to be independently associated with the risk of breast cancer in both premenopausal and postmenopausal women. The risk of incident breast cancer increased in women with persistent dense breasts, whereas the breast cancer risk decreased as dense breasts regressed. The findings of the current study support that both dense breasts at baseline and their persistence over time are independent risk factors for developing breast cancer.

Automated mammographic density measurement using Quantra™: comparison with the Royal Australian and New Zealand College of Radiology synoptic scale

Purpose: This technology evaluation study assesses the limits of agreement between the mammographic density (MD) measurement of Quantra™ from different breasts and mammographic views and its agreement with the Royal Australian and New Zealand College of Radiologists (RANZCR) synoptic scale. Approach: MD of 800 women was assessed by Quantra™ and seven radiologists using the RANZCR synoptic scale. Bland-Altman analysis was used to assess the limits of agreement between Quantra™ MD measures from both breasts and mammographic views. The agreement between Quantra™ and the RANZCR synoptic scale was assessed using weighted kappa ( K w ). The receiver operating characteristics area under the curve (AUC) was used to assess the performance of Quantra™ in reproducing RANZCR MD ratings. Results: There was no significant bias in the mean MD of Quantra™ from both breasts: left versus right craniocaudal (CC) views ( B = - 0.14 ; p = 0.36 ) and right versus left mediolateral oblique (MLO) ( B = - 0.021 ; p = 0.18 ). However, MD measures from the same breast but different views showed significant bias: right CC versus right MLO ( B = 0.064 ; p < 0.0001 ) and left CC versus left MLO ( B = 0.56 ; p < 0.0001 ). Quantra™ demonstrated substantial agreement with the RANZCR synoptic scale on four- and two-category scales ( K w = 0.62 ; 0.59 to 0.66 and 0.76; 0.72 to 0.81, respectively). Quantra™ better reproduced the RANZCR synoptic scale on a two-category scale ( AUC = 0.88 ; 0.84 to 0.91) than a four-category scale ( AUC = 0.62 ; 0.58 to 0.67 to 0.78; 0.74 to 0.82). Conclusions: Quantra™ reproduces MD classification using the RANZCR synoptic scale on a two-category scale and should help in identification of women with dense breasts who may need adjunctive imaging for early detection of breast cancer.

Detecting mammographically occult cancer in women with dense breasts using deep convolutional neural network and Radon Cumulative Distribution Transform

We have applied the Radon Cumulative Distribution Transform (RCDT) as an image transformation to highlight the subtle difference between left and right mammograms to detect mammographically occult (MO) cancer in women with dense breasts and negative screening mammograms. We developed deep convolutional neural networks (CNNs) as classifiers for estimating the probability of having MO cancer. We acquired screening mammograms of 333 women (97 unilateral MO cancer) with dense breasts and at least two consecutive mammograms and used the immediate prior mammograms, which radiologists interpreted as negative. We used fivefold cross validation to divide our dataset into a training and independent test sets with ratios of 0.8:0.2. We set aside 10% of the training set as a validation set. We applied RCDT on the left and right mammograms of each view. We applied inverse Radon transform to represent the resulting RCDT images in the image domain. We then fine-tuned a VGG16 network pretrained on ImageNet using the resulting images per each view. The CNNs achieved mean areas under the receiver operating characteristic (AUC) curve of 0.73 (standard error, SE = 0.024) and 0.73 (SE = 0.04) for the craniocaudal and mediolateral oblique views, respectively. We combined the scores from two CNNs by training a logistic regression classifier and it achieved a mean AUC of 0.81 (SE = 0.032). In conclusion, we showed that inverse Radon-transformed RCDT images contain information useful for detecting MO cancers and that deep CNNs could learn such information.

Application of the downgrade criteria to supplemental screening ultrasound for women with negative mammography but dense breasts

We investigated whether the application of the downgrade criteria to supplemental screening ultrasound (US) for women with negative mammography but dense breasts can reduce the rate of Breast Imaging Reporting and Data System (BI-RADS) categories 3 to 4a without a loss of cancer detection.This retrospective study was approved by the Institutional Review Board, and the need to obtain informed consent was waived. A total of 3171 consecutive women (978 women, 1173 women, and 1020 women in the first, second, and third year, respectively) with negative mammography but dense breast who underwent radiologist-performed, hand-held supplemental screening US from March 2010 to February 2013 were included. Downgrade criteria for BI-RADS category 2 were complicated cysts ≤5 mm observed as circumscribed, homogeneous, and hypoechoic lesions and circumscribed oval-shaped solid masses ≤5 mm. Changes in the distribution of BI-RADS category, biopsy rate, and cancer detection yield over 3 years were analyzed. Performances of less-experienced (12 fellows with <2 years of experience) and experienced (3 staffs with >12 years of experience) radiologists were compared. Outcomes of initial examinations (prevalence screening) and noninitial examinations (incidence screening) were compared.Application of the downgrade criteria reduced BI-RADS categories 3 to 4a in both less-experienced (from 39.4% to 16.0%, P < 0.001) and experienced radiologists (from 22.6% to 11.1%, P < 0.001) over 3 years. Biopsy rates also significantly decreased from 6.5% to 2.4% (P < 0.001). Cancer detection yield of supplemental screening US was 2.8 per 1000 examinations (9 of 3171: 2 ductal carcinoma in situ and 7 invasive cancers). There were no differences in cancer detection yield per each year (P = 0.539). There was no interval cancer. In noninitial examinations, BI-RADS categories 3 to 4a rates, biopsy rates, and cancer detection rates were lower compared to initial examinations.Application of the downgrade criteria reduced BI-RADS categories 3 to 4a without a loss of cancer detection. We suggest that our downgrade criteria can be used to reduce the false positive rate in the supplemental screening US. Further large-scale, multicenter, prospective studies are needed to validate the effectiveness of the downgrade criteria.

The Use of Ultrasound in Breast Cancer Screening of Asymptomatic Women with Dense Breast Tissue: A Narrative Review

INTRODUCTION

Mammography is the standard screening modality for breast cancer; however, sensitivity reduces with increasing breast density, resulting in the potential for masking of cancer. Ultrasound is a potential supplemental screening tool, but its routine use is controversial.

METHODS

A database search was performed with keywords "ultrasound" and "breast density and screening", including variations. Articles were included if they assessed the use of hand-held ultrasound as a supplemental screening modality in women with dense breasts.

DISCUSSION

Twelve articles were identified. No high-level evidence articles were identified. Cancer detection rates increased with the addition of ultrasound-to-mammography screening protocols. However, this was associated with increased costs per cancer detected, an increased biopsy rate, and a low positive predictive value. The survival benefit, cost versus benefit, and psychological impact of the addition of ultrasound is unknown.

CONCLUSIONS

The addition of ultrasound to a screening program in an asymptomatic population of women with dense breast tissue detects additional cancers compared with mammography alone. Knowledge regarding a survival or cost benefit associated with increased cancer detection, and the psychological impact of the addition of ultrasound is unknown. Further research is needed to assess whether the addition of ultrasound is cost-effective with respect to clinical outcome and survival.

Radiologist-performed hand-held ultrasound screening at average risk of breast cancer: results from a single health screening center

BACKGROUND

Ultrasound (US) screening is not currently recommended as a routine screening modality in the general population of average risk. The cancer detection yield and positive predictive value in an average risk general population who undergo breast screening by experienced radiologists is unclear.

PURPOSE

To determine the performance of screening breast US in women at an average risk for breast cancer undergoing breast screening by experienced radiologists.

MATERIAL AND METHODS

This study received institutional review board approval, and informed consent was waived. A retrospective review of our database revealed 1526 women who underwent prevalence screening US at a single health screening center and had negative findings on digital mammography (MG). The Breast Imaging and Reporting Data System (BI-RADS) final assessments of the breast US were analyzed retrospectively, with the reference standard defined as a combination of pathology and a 12-month follow-up. The cancer detection rate and positive predictive value (PPV2) of the biopsies were calculated according to breast density.

RESULTS

The average time to perform a screening US examination was 15-20 min. Of 1526 women, 1095 (71.8%) were classified as BI-RADS category 1 or 2; 340 (22.3%) were classified as category 3; and 91 (6.0%) were classified as category 4. Five malignant lesions were found in women with dense breasts. The overall cancer detection rate was 3.3, and the cancer detection rate for dense breasts was 5.1 per 1000 screens (4.1 per 1000 screens [heterogeneously dense breast], 7.7 per 1000 screens [extremely dense breast]). The PPV2 for biopsies was 5.3%.

CONCLUSION

The radiologist-performed screening US offered to women with an average risk and dense breasts can detect additional mammographically occult breast cancers. In screening US, a relatively high rate of BI-RADS category 3 and 4 lesions was observed.

Molecular breast imaging: an emerging modality for breast cancer screening

Screening mammography is recognized as an imperfect imaging tool that performs poorly in women with dense breast tissue - a limitation which has driven demand for supplemental screening techniques. One potential supplemental technique is molecular breast imaging (MBI). Significant improvements in gamma camera technology allow MBI to be performed at low radiation doses, comparable with those of tomosynthesis and mammography. A recent screening trial in women with dense breast tissue yielded a cancer detection rate of 3.2 per 1000 for mammography alone and 12.0 per 1000 for the combination of mammography and MBI. MBI also demonstrated a lower recall rate than that of mammography. MBI is a promising supplemental screening technique in women with dense breast tissue.

Role of dual-energy contrast-enhanced digital mammography as a problem-solving tool in dense breasts: A case report

Breast density is strongly and independently related to the risk of breast cancer. Women with very dense breasts may have up to five times higher chances of developing malignancy as compared to those with less breast density. Detecting lesions in extremely dense or heterogeneous breasts on screening mammogram can be difficult. We report a case of incidental bilateral breast malignancy in an asymptomatic patient in whom mammogram and screening ultrasound were equivocal.

Evaluation of screening whole-breast sonography as a supplemental tool in conjunction with mammography in women with dense breasts

OBJECTIVES

The purpose of this study was to evaluate the use and performance of supplemental screening whole-breast sonography in conjunction with mammography in asymptomatic women with dense breast tissue.

METHODS

A total of 28,796 asymptomatic women underwent screening mammography. Among 20,864 women with dense breasts (72%), 8359 underwent additional sonography as part of their screening examinations. We classified women with mammographically dense breasts into mammography-only and mammography-plus-sonography groups. The reference standard was a combination of pathologic results and clinical follow-up at 2 years. We compared the recall rate, cancer detection yield, sensitivity, specificity, and positive predictive value in each group.

RESULTS

Among the 20,864 women with dense breasts, 35 cancers were diagnosed, with a mean size of 13 mm. The cancer detection yield was 0.480 per 1000 women in the mammography-only group and increased to 2.871 in the mammography-plus-sonography group. Of 24 cancers detected in the mammography-plus-sonography group, the mean size was 11 mm, and the axillary lymph nodes were negative in 19 of 20. The sensitivity was significantly higher in the mammography-plus-sonography group than the mammography-only group (100% versus 54.55%; P = .002). The positive predictive values of sonographically prompted biopsy were 11.1% for the mammography-plus-sonography group and 50% for the mammography-only group.

CONCLUSIONS

Supplemental screening whole-breast sonography increases the cancer detection yield by 2.391 cancers per 1000 women with dense breast tissue over that of mammography alone. It is beneficial for increased detection of breast cancers that are predominantly small and node negative; however, it also raises the number of false-positive results.

Role of Breast Ultrasound for the Detection and Differentiation of Breast Lesions

Diagnosis of breast cancer has been widely improved since the development of high-resolution ultrasound equipment. In the past, ultrasound was only considered useful for the diagnosis of cysts. Meanwhile, it improves the differential diagnosis of benign and malignant lesions, local preoperative staging and guided interventional diagnosis. In dense breasts, mammography has limited sensitivity. Furthermore, women with dense parenchyma have a highly increased risk of breast cancer development. Ultrasound is useful to examine dense breast tissue. Recent studies have shown that the detection of small cancers with high-resolution ultrasound is increased by 3-4 cancers per 1,000 women without clinical or mammographic abnormalities. Furthermore, stage distribution is similar between mammographically and sonographically detected carcinomas. Ultrasound is routinely used for curative diagnosis, to overcome the limitations of mammography. However, within the mammographic screening in Germany, breast density is not considered as important. Ultrasound is only used if a suspicious lesion is detected by mammography. Interestingly, 2 years ago, a screening project started in Austria in which ultrasound is always added in cases of dense breasts. Preliminary data show that the detection of additional carcinomas is increased in the same order as shown in previous studies. Therefore, an improved cancer detection and differentiation can be expected with high-resolution ultrasound.