Mammography Examination Research Articles

The application of traditional machine learning and deep learning techniques in mammography: a review.

Breast cancer, the most prevalent malignant tumor among women, poses a significant threat to patients' physical and mental well-being. Recent advances in early screening technology have facilitated the early detection of an increasing number of breast cancers, resulting in a substantial improvement in patients' overall survival rates. The primary techniques used for early breast cancer diagnosis include mammography, breast ultrasound, breast MRI, and pathological examination. However, the clinical interpretation and analysis of the images produced by these technologies often involve significant labor costs and rely heavily on the expertise of clinicians, leading to inherent deviations. Consequently, artificial intelligence(AI) has emerged as a valuable technology in breast cancer diagnosis. Artificial intelligence includes Machine Learning(ML) and Deep Learning(DL). By simulating human behavior to learn from and process data, ML and DL aid in lesion localization reduce misdiagnosis rates, and improve accuracy. This narrative review provides a comprehensive review of the current research status of mammography using traditional ML and DL algorithms. It particularly highlights the latest advancements in DL methods for mammogram image analysis and offers insights into future development directions.

PACS-integrated machine learning breast density classifier: clinical validation

ObjectiveTo test the performance of a novel machine learning-based breast density tool. The tool utilizes a convolutional neural network to predict the BI-RADS based density assessment of a study. The clinical density assessments of 33,000 mammographic examinations (164,000 images) from one academic medical center (Site A) were used for training. Materials and methodsThis was an IRB approved HIPAA compliant study performed at two academic medical centers. The validation data set was composed of 500 studies from one site (Site A) and 700 from another (Site B). At Site A, each study was assessed by three breast radiologists and the majority (consensus) assessment was used as truth. At Site B, if the tool agreed with the clinical reading, then it was considered to have correctly predicted the clinical reading. In cases where the tool and the clinical reading disagreed, then the study was evaluated by three radiologists and the consensus reading was used as the clinical reading. ResultsFor the classification into the four categories of the Breast Imaging Reporting and Data System (BI-RADS®), the AI classifier had an accuracy of 84.6% at Site A and 89.7% at Site B. For binary classification (dense vs. non-dense), the AI classifier had an accuracy of 94.4% at Site A and 97.4% at Site B. In no case did the classifier disagree with the consensus reading by more than one category. ConclusionsThe automated breast density tool showed high agreement with radiologists' assessments of breast density.

Evaluating clinical efficacy of hospital-based surveillance with mammography and ultrasonography for breast cancer

Periodic mammography and/or sonography examinations are conducted across numerous hospitals nationalwidely, especially for antedees with a positive mammography screening. Despite the regular practice, clinical efficacy of hospital-based breast cancer surveillance remains unclear. Specifically, the impact of surveillance interval upon survival and prognostic surrogates stratified by menopausal status, as well as malignant transition rate should be deciphered. We retrieved cancer registry to ascertain 841 breast cancers with surveillance history through administration data. Healthy controls underwent breast surveillance and were concurrently free of cancer. More benign diseases rather than cancers were identified from premenopausal women (age ≤50 years) with sonography alone within one year, as well as older women (age >50) with both mammography and sonography one to two years before a cancer or benign diagnosis. Among breast cancers, mammography alone during the antecedent one to two years had a protective effect for diagnosing carcinoma in situ rather than invasive cancer (age-adjusted odds ratio: 0.048, P=0.016). Three-state time homogeneous Markov model showed that hospital-based breast surveillance within 2 years of disease onset reduced the malignant transition rate by 65.16% (59.79-76.74%). The clinical efficacy of breast cancer surveillance was evidenced.

The Relationship Between Body Mass Index and Breast Volume on Mammography Examination

Breast cancer is the most common type of cancer suffered by women worldwide. The development and incidence of breast cancer is influenced by various risk factors, one of which is breast volume. The amount of breast volume is influenced by several factors such as genetics, demographics, reproduction, lifestyle, and Body Mass Index (BMI). This study aims to determine the relationship between Body Mass Index and breast volume of research subjects at RSUD Dr. Saiful Anwar Malang, East Java. This type of research is observational analytic with a cross-sectional study approach and analysis using Kruskal Wallis test, the Mann Whitney test, and Spearman correlation test. Based on the Kruskal Wallis test BMI-left breast volume, the sig value obtained was 0.020 (sig < 0.05) while in the analysis of BMI-right breast volume the sig value obtained was 0.083 (sig > 0.05). Follow-up analysis using the Mann-Whitney BMI test-left breast volume showed significant results in the normal-overweight (p=0.050) and normal-obese (p=0.007) BMI groups. While the BMI-volume analysis of the right breast showed significant results only in the normal-obese BMI group (p = 0.044). Spearman correlation test of BMI-left breast volume obtained sig = 0.000 and r = 0.603. While the Spearman test of BMI-right breast volume obtained sig = 0.004 and r = 0.511. This study concludes that there is a significant relationship between BMI and breast volume which is positive with a fairly strong correlation strength.

Health Literacy and Breast and Cervical Cancer Screening Behaviors in Women

Background: Breast cancer and cervical cancer, the most common forms of cancer in women worldwide, are on a fast and steady rise. Cancer screening tests are an important tool to combat cancer-related morbidity and mortality. Considering the importance of health literacy in promoting community health, this study aimed to investigate the relationship between health literacy and the history of screening behaviors of common cancers in women referred to Bushehr healthcare centers in 2019. Methods: This cross-sectional study was carried out on 380 women referred to urban health centers in Bushehr in 2019. Data were collected by using a questionnaire consisting of three parts: demographic characteristics, breast and cervical cancer screening behavior, and health literacy (HELIA) questionnaire. Data were analyzed via SPSS 21 software using the Chi-square test, an independent t-test, and descriptive statistical methods. Furthermore, P<0.05 was considered statistically significant. Results: The mean age of the subjects was 34.88±9.15 years. Moreover, 3.2% of subjects had inadequate health literacy, 13.9% had health literacy at a border level, and 82.9% had sufficient health literacy. In this study, 46.8% and 88.2% of the women did not undergo Pap tests and mammography, respectively, and 73.9% were never referred to the health centers for clinical examination of breasts. In addition, health literacy had a significant relationship with monthly breast examination and undergoing Pap smear (P<0.001), but health literacy had no significant relationship with undergoing mammography and clinical breast examination (P>0.05). Conclusion: The results of this study indicated that women with higher levels of health literacy are more likely to get in the habit of doing monthly breast self-examination and undergoing the Pap test. Thus, it is suggested that this result be considered in prevention programs (primary and secondary) to increase women’s health.

Characteristics of breast cancers detected by screening mammography in Taiwan: a single institute’s experience

Background/aimBreast cancer is the most common female malignancy in the world. Nearly ninety percent of screening-detected breast cancers were diagnosed with earlier stages of 0 to II in Taiwan. It’s widely acknowledged that mammography screening of breast cancer can achieve the goal of early diagnosis and treatment in terms of preventive task while neglected interval cancers are associated with unfavorable tumor characteristics and worse outcomes. The purpose of this study was to explore the characteristics of screening-detected breast cancers in Taiwan.Materials and methodsBoth screening and diagnostic mammography examinations with accompanied BI-RADS (breast imaging-reporting and data system) classification were extracted from the health information system and linked to cancer registry in Taiwan. Enrolled population included those attending their first mammography between 2012 and 2016, excluding subjects with previous breast cancer, or with missing or incomplete data. We compared treatment outcomes between breast cancers with either initial screening or diagnostic mammography (control group), and investigated the compositions of breast cancers detected by screening mammography through direct chart reviews.ResultsA total of 84,246 screening and 61,230 diagnostic mammography sessions were performed from 2010 to 2020. More positive results (BI-RADS 0, 3, 4 and 5) were observed for those attending the first diagnostic than the first screening mammography (43.58% versus 16.12%, p < 0.001). Earlier stages (0 and I) distribution (92% versus 81%, p < 0.0001), better survivorship (overall survival: 96.91% versus 92.17%, p = 0.007) and a lower HER2 (human epidermal growth factor receptor II) positive status and lower tumor grade were noted in breast cancers with initial screening rather than diagnostic mammography. Among 26,103 mammography screening invitees between 2012 and 2016, 325 breast cancers were ascertained from cancer registry. Of these, 234 had one, 72 had two and 19 had three episodes of mammography before cancer diagnosis. Extensive chart reviews revealed that women with and without breast symptoms constituted 29.9 and 70.1% of the 325 screening-detected breast cancers, with the latter further divided into false negative results (interval cancer), diagnosed at the first mammography, diagnostic at the secondary or subsequent mammography and those with a delayed biopsy or confirmatory imaging constituted (5.2, 47, 10.5 and 7.4%).ConclusionScreening-detected breast cancers were a mixture of women with and without symptoms, those with a false negative result, true negative results with cancer detected at subsequent mammography and non-adherers. Despite this, efficacy of mammography screening was ascertained in Taiwan from this study. To further enhance earlier detection and decrease false negativity, the impact of repeated mammography, and additional sonography for symptomatic women, compliance following a positive screening mammography should not be overemphasized.

Consistency of Artificial Intelligence (AI)-based Diagnostic Support Software in Short-term Digital Mammography Reimaging After Core Needle Biopsy.

To evaluate the consistency in the performance of Artificial Intelligence (AI)-based diagnostic support software in short-term digital mammography reimaging after core needle biopsy. Of 276 women who underwent short-term (<3 mo) serial digital mammograms followed by breast cancer surgery from Jan. to Dec. 2017, 550 breasts were included. All core needle biopsies for breast lesions were performed between serial exams. All mammography images were analyzed using a commercially available AI-based software providing an abnormality score (0-100). Demographic data for age, interval between serial exams, biopsy, and final diagnosis were compiled. Mammograms were reviewed for mammographic density and finding. Statistical analysis was performed to evaluate the distribution of variables according to biopsy and to test the interaction effects of variables with the difference in AI-based score according to biopsy. AI-based score of 550 exams (benign or normal in 263 and malignant in 287) showed significant difference between malignant and benign/normal exams (0.48 vs. 91.97 in first exam and 0.62 vs. 87.13 in second exam, P<0.0001). In comparison of serial exams, no significant difference was found in AI-based score. AI-based score difference between serial exams was significantly different according to biopsy performed or not (-0.25 vs. 0.07, P = 0.035). In linear regression analysis, there was no significant interaction effect of all clinical and mammographic characteristics with mammographic examinations performed after biopsy or not. The results from AI-based diagnostic support software for digital mammography was relatively consistent in short-term reimaging even after core needle biopsy.

Impact of BMI on Prevalence of Dense Breasts by Race and Ethnicity.

Density notification laws require notifying women of dense breasts with dense breast prevalence varying by race/ethnicity. We evaluated whether differences in body mass index (BMI) account for differences in dense breasts prevalence by race/ethnicity. Prevalence of dense breasts (heterogeneously or extremely dense) according to Breast Imaging Reporting and Data System and obesity (BMI > 30 kg/m2) were estimated from 2,667,207 mammography examinations among 866,033 women in the Breast Cancer Surveillance Consortium (BCSC) from January 2005 through April 2021. Prevalence ratios (PR) for dense breasts relative to overall prevalence by race/ethnicity were estimated by standardizing race/ethnicity prevalence in the BCSC to the 2020 U.S. population, and adjusting for age, menopausal status, and BMI using logistic regression. Dense breasts were most prevalent among Asian women (66.0%) followed by non-Hispanic/Latina (NH) White (45.5%), Hispanic/Latina (45.3%), and NH Black (37.0%) women. Obesity was most prevalent in Black women (58.4%) followed by Hispanic/Latina (39.3%), NH White (30.6%), and Asian (8.5%) women. The adjusted prevalence of dense breasts was 19% higher [PR = 1.19; 95% confidence interval (CI), 1.19-1.20] in Asian women, 8% higher (PR = 1.08; 95% CI, 1.07-1.08) in Black women, the same in Hispanic/Latina women (PR = 1.00; 95% CI, 0.99-1.01), and 4% lower (PR = 0.96; 95% CI, 0.96-0.97) in NH White women relative to the overall prevalence. Clinically important differences in breast density prevalence are present across racial/ethnic groups after accounting for age, menopausal status, and BMI. If breast density is the sole criterion used to notify women of dense breasts and discuss supplemental screening it may result in implementing inequitable screening strategies across racial/ethnic groups. See related In the Spotlight, p. 1479.

The impact of AI suggestions on radiologists’ decisions: a pilot study of explainability and attitudinal priming interventions in mammography examination

Various studies have shown that medical professionals are prone to follow the incorrect suggestions offered by algorithms, especially when they have limited inputs to interrogate and interpret such suggestions and when they have an attitude of relying on them. We examine the effect of correct and incorrect algorithmic suggestions on the diagnosis performance of radiologists when (1) they have no, partial, and extensive informational inputs for explaining the suggestions (study 1) and (2) they are primed to hold a positive, negative, ambivalent, or neutral attitude towards AI (study 2). Our analysis of 2760 decisions made by 92 radiologists conducting 15 mammography examinations shows that radiologists’ diagnoses follow both incorrect and correct suggestions, despite variations in the explainability inputs and attitudinal priming interventions. We identify and explain various pathways through which radiologists navigate through the decision process and arrive at correct or incorrect decisions. Overall, the findings of both studies show the limited effect of using explainability inputs and attitudinal priming for overcoming the influence of (incorrect) algorithmic suggestions.

Evaluation of a population-based breast cancer screening in North China

BackgroundDespite mammography-based screening for breast cancer has been conducted in many countries, there are still little data on participation and diagnostic yield in population-based breast cancer screening in China.MethodsWe enrolled 151,973 eligible women from four cities in Hebei Province within the period 2013–2021 and followed up until December 31, 2021. Participants aged 40–74 who assessed as high risk were invited to undergo breast ultrasound and mammography examination. Overall and group-specific participation rates were calculated. Multivariable analyses were used to estimate the factors associated with participation rates. The diagnostic yield of both screening and no screening groups was calculated. We further analyzed the stage distribution and molecular subtype of breast cancer cases by different modes of cancer detection.ResultsA total of 42,547 participants were evaluated to be high risk of breast cancer. Among them, 23,009 subjects undertook screening services, with participation rate of 54.08%. Multivariable logistic regression model showed that aged 45–64, high education level, postmenopausal, current smoking, alcohol consumption, family history of breast cancer, and benign breast disease were associated with increased participation of screening. After median follow-up of 3.79 years, there were 456 breast cancer diagnoses of which 65 were screen-detected breast cancers (SBCs), 27 were interval breast cancers (IBCs), 68 were no screening cancers, and 296 were cancers detected outside the screening program. Among them, 92 participants in the screening group (0.40%) and 364 in the non-screening group (0.28%) had breast cancer detected, which resulted in an odds ratio of 1.42 (95% CI 1.13–1.78; P = 0.003). We observed a higher detection rate of breast cancer in the screening group, with ORs of 2.42 (95% CI 1.72–3.41) for early stage (stages 0–I) and 2.12 (95% CI 1.26–3.54) for luminal A subtype. SBCs had higher proportion of early stage (71.93%) and luminal A subtype (47.22%) than other groups.ConclusionsThe significant differences in breast cancer diagnosis between the screening and non-screening group imply an urgent need for increased breast cancer awareness and early detection in China.

Diagnostic value of mammography density of breast masses by using deep learning.

In order to explore the relationship between mammographic density of breast mass and its surrounding area and benign or malignant breast, this paper proposes a deep learning model based on C2FTrans to diagnose the breast mass using mammographic density. This retrospective study included patients who underwent mammographic and pathological examination. Two physicians manually depicted the lesion edges and used a computer to automatically extend and segment the peripheral areas of the lesion (0, 1, 3, and 5mm, including the lesion). We then obtained the mammary glands' density and the different regions of interest (ROI). A diagnostic model for breast mass lesions based on C2FTrans was constructed based on a 7: 3 ratio between the training and testing sets. Finally, receiver operating characteristic (ROC) curves were plotted. Model performance was assessed using the area under the ROC curve (AUC) with 95% confidence intervals (CI), sensitivity, and specificity. In total, 401 lesions (158 benign and 243 malignant) were included in this study. The probability of breast cancer in women was positively correlated with age and mass density and negatively correlated with breast gland classification. The largest correlation was observed for age (r = 0.47). Among all models, the single mass ROI model had the highest specificity (91.8%) with an AUC = 0.823 and the perifocal 5mm ROI model had the highest sensitivity (86.9%) with an AUC= 0.855. In addition, by combining the cephalocaudal and mediolateral oblique views of the perifocal 5mm ROI model, we obtained the highest AUC (AUC = 0.877P < 0.001). Deep learning model of mammographic density can better distinguish benign and malignant mass-type lesions in digital mammography images and may become an auxiliary diagnostic tool for radiologists in the future.

ULTRASOUND VS. MAMMOGRAPHY IN EVALUATING SUSPICIOUS BREAST LESIONS

This study aimed to compare the effectiveness of ultrasound and mammography in evaluating suspicious breast lesions and determine their respective diagnostic capabilities. A retrospective analysis was conducted at Khyber Teaching Hospital Peshawar over 6 months, from March 2021 to September 2021. The study included 100 patients with suspicious breast lesions who underwent ultrasound and mammography examinations. The demographic data, including age and breast density, were recorded. Both imaging modalities' diagnostic accuracy, sensitivity, specificity, and positive predictive value were calculated. Statistical analysis was performed using appropriate tests. The analysis included 100 patients with an average age of 51.3 years. The breast density distribution was as follows: 20% dense, 35% scattered fibroglandular, 30% heterogeneously dense, and 15% extremely dense. Ultrasound demonstrated higher sensitivity (75%) than mammography (65%) in the study population. When stratified by age, ultrasound had higher sensitivity in all age groups, with the highest sensitivity observed in patients aged &gt;60 years (85%). Regarding breast density, ultrasound showed higher sensitivity across all density categories, particularly in patients with dense breasts (80%) and heterogeneously dense breasts (85%). Our study findings suggest that ultrasound is a valuable imaging modality in evaluating suspicious breast lesions, offering higher sensitivity than mammography, especially in patients with dense breast tissue. Incorporating ultrasound as an adjunct to mammography can improve diagnostic accuracy and aid in the early detection of breast cancer. These results support the growing recognition of ultrasound as an important tool in breast cancer diagnosis and highlight the need for further research and collaboration to optimize its utilization in clinical practice.

The significance of morphological verification of calcifications in differential diagnsis of the mammary gland pathology

Relevance of the research. An important advantage of radiological imaging methods is their ability to detect areas of calcification in tissues, even in the form of microcalcifications, measuring up to 1 mm. These microcalcifications are formed under the influence of various factors and serve as important diagnostic criteria for a number of pathologies. Microcalcifications, detected during mammographic examination, are considered reliable markers of breast cancer. The classification of breast microcalcifications is based on their radiological and histological characteristics, which are evaluated during the diagnostic process to identify breast pathology. The aim of the study is to determine the frequency of detecting calcifications in the breast through mammographic examination and verify their significance as diagnostic criteria for breast cancer.Materials and methods. The study utilized a processed randomized sample of anonymized archival mammograms from 100 women of mature and elderly age who underwent screening or preventive examinations at the mammography room of the CNE "6th city polyclinic of Lviv". The examinations were conducted using the Fujifilm Amulet Innоvality machine (Japan) in the craniocaudal (CC) and mediolateral oblique (MLO) projections. Histological slides from the archives of the Histology, Cytology, and Embryology Department of Danylo Halytsky Lviv National Medical University and anonymized archival preparations of DOBRA DIAGNOSTYKA LLC were used for the study.The slides were examined under a UlabXSP-137TLED microscope at various magnifications and photographed using an XCAM 1080P camera.Results. It has been established that the frequency of detecting calcifications during screening/preventive examinations of mature and elderly women was 69% of the total number of subjects examined, more than half of them bilaterally. The detected calcifications had different localization, shape, size, and quantity. The analysis of scientific literature sources indicates the existence of different views on the presence of calcifications in breast tissue - some authors consider them to be a characteristic radiological sign of the breast, while others assert that the frequency of malignant neoplasms increases, including in cases of calcification of the glandular tissue. The results of mammary glands histological studying in mature and elderly women indicate that calcifications can be visualized in the context of age-related changes in the stroma or parenchyma of the organ, which don`t exhibit morphological signs of neoplastic development. These calcifications can be manifestations of age-related involution and are possibly formed as a result of calcification of blocked secretions. Conclusion. The presence of calcifications can`t be considered a definitive diagnostic criterion for breast cancer. The final verification of the diagnosis can only be made after conducting a histological examination.

Pacemaker in patients undergoing mammography: A limitation for breast cancer diagnosis?

A pacemaker may affect the utility of a mammogram in several ways. The aim of this study is to summarize our institution's experience with mammograms among patients with a cardiac pacemaker, focusing on the diagnostic workup among patients with a newly diagnosed ipsilateral breast cancer. A retrospective search of all mammography reports between January 2011 and April 2021 was conducted for identifying cases of patients with a pacemaker. Demographic and clinical characteristics as well as mammography-derived quality parameters and findings were categorized and statistically compared. The incidence of pacemaker concurrence in mammographic examination, although apparently slightly under-documented, accounted for 0.33% of cases. Population mean age was 71.7 years, and most patients (79%) had a left-sided pacemaker. The pacemaker was much more likely to be projected on the medio-lateral-oblique (96%) than on the cranio-caudal view (10%), on the axilla rather than the breast, and on the retro-pectoral rather than the pre-pectoral region (P < 0.001 for all). Compression force decreased by up to 23.0% (P < 0.001) and breast thickness increased by up to 9.5% (P < 0.001) for the ipsilateral vs. the contralateral side. Among 11 patients with newly diagnosed ipsilateral breast cancer, the pacemaker partially projected on the tumour region in two cases, and significantly obscured the tumour in another two. Although rare, the coexistence of a pacemaker in patients undergoing mammography is associated with reduced image quality due to suboptimal breast visualization and reduced compression, and as a result, this may eventually lead to decreased diagnostic efficacy.

A GATE Monte Carlo framework for dosimetric evaluation in mammography in an Algerian hospital

A framework has been developed for dosimetric evaluation in mammography, using the GATE Monte Carlo (MC) platform, to simulate a MAMMOMAT 3000 Nova mammograph (Siemens) available at the University Hospital Center "1st November 1954" of Oran (EHU Oran 1er Novembre, 1954), Algeria. Calculated quantities such half-value layer (HVL), Entrance Surface Dose (ESD) and Mean Glandular Dose (MGD) have been compared to experimental data in order to validate the modeling of mammography examinations. Results are consistent with previous studies and show a good agreement between measurements and Monte Carlo calculations. By varying the tube voltage from 25 to 35 kV, we have estimated an increasing of a factor of 2.4 in ESD, and a factor of 2.75 for the MGD in a breast phantom. Furthermore, the current intensity of 100 mAs used for a beam quality combination (Mo/Mo) Anode/filter was found suitable for the tube voltages of 25–29 keV since the MGD does not exceed the limits set by the different quality insurance protocols. This GATE dose calculation framework thus provides a very useful tool for the optimization of mammography examinations at Oran hospital by allowing a better estimation of the dose delivered to patients according to the parameters of the examination.

Toward the strengthening of radioprotection during mammography examinations through transparent glass screens: A benchmarking between experimental and Monte Carlo simulation studies.

A lead-acrylic protective screen is suggested to reduce radiation exposure to the unexposed breast during mammography. The presence of toxic lead in its structure may harm the tissues with which it comes in contact. This study aimed to design a CdO-rich quaternary tellurite glass screen (C40) and evaluate its efficiency compared to the Lead-Acrylic protective screen. A three-layer advanced heterogeneous breast phantom designed in MCNPX (version 2.7.0) general-purpose Monte Carlo code. Lead acrylic and C40 shielding screens were modeled in the MCNPX and installed between the right and left breast. The reliability of the absorption differences between the lead acrylic and C40 glass were assessed. The results showed that C40 protective glass screen has much superior protection properties compared to the lead acrylic protective screen. The amount of total dose absorbed in the unexposed breast for C40 was found to be much less than that for lead-based acrylic. The protection provided by the C40 glass screen is 35-38% superior to that of the Lead-Acrylic screen. The C40 offer the opportunity to avoid the toxic Pb in the structure of Lead-Acrylic material and may be utilized for mammography to offer superior radioprotection to Lead-Acrylic and significantly lower the dose amount in the unexposed breast. It can be concluded that transparent glass screens may be utilized for radiation protection purposes in critical diagnostic radiology applications through mammography.

VAI-B: a multicenter platform for the external validation of artificial intelligence algorithms in breast imaging.

Multiple vendors are currently offering artificial intelligence (AI) computer-aided systems for triage detection, diagnosis, and risk prediction of breast cancer based on screening mammography. There is an imminent need to establish validation platforms that enable fair and transparent testing of these systems against external data. We developed validation of artificial intelligence for breast imaging (VAI-B), a platform for independent validation of AI algorithms in breast imaging. The platform is a hybrid solution, with one part implemented in the cloud and another in an on-premises environment at Karolinska Institute. Cloud services provide the flexibility of scaling the computing power during inference time, while secure on-premises clinical data storage preserves their privacy. A MongoDB database and a python package were developed to store and manage the data on-premises. VAI-B requires four data components: radiological images, AI inferences, radiologist assessments, and cancer outcomes. To pilot test VAI-B, we defined a case-control population based on 8080 patients diagnosed with breast cancer and 36,339 healthy women based on the Swedish national quality registry for breast cancer. Images and radiological assessments from more than 100,000 mammography examinations were extracted from hospitals in three regions of Sweden. The images were processed by AI systems from three vendors in a virtual private cloud to produce abnormality scores related to signs of cancer in the images. A total of 105,706 examinations have been processed and stored in the database. We have created a platform that will allow downstream evaluation of AI systems for breast cancer detection, which enables faster development cycles for participating vendors and safer AI adoption for participating hospitals. The platform was designed to be scalable and ready to be expanded should a new vendor want to evaluate their system or should a new hospital wish to obtain an evaluation of different AI systems on their images.

Utilization of three-layers heterogeneous mammographic phantom through MCNPX code for breast and chest radiation dose levels at different diagnostic X-ray energies: A Monte Carlo simulation study.

We report the breast and chest radiation dose assessment for mammographic examinations using a three-layer heterogeneous breast phantom through the MCNPX Monte Carlo code. A three-layer heterogeneous phantom along with compression plates and X-ray source are modeled. The validation of the simulation code is obtained using the data of AAPM TG-195 report. Deposited energy amount as a function of increasing source energy is calculated over a wide energy range. The behavioral changes in X-ray absorption as well as transmission are examined using the F6 Tally Mesh extension of MCNPX code. Moreover, deposited energy amount is calculated for modeled body phantom in the same energy range. The diverse distribution of glands has a significant impact on the quantity of energy received by the various breast layers. In layers with a low glandular ratio, low-energy primary X-ray penetrability is highest. In response to an increase in energy, the absorption in layers with a low glandular ratio decreased. This results in the X-rays releasing their energy in the bottom layers. Additionally, the increase in energy increases the quantity of energy absorbed by the tissues around the breast.

Abstract OT3-18-03: The PRAIM study: A prospective multicenter observational study of an integrated Artificial Intelligence system with live monitoring

Abstract Background. Several retrospective studies have illustrated the potential clinical benefit of artificial intelligence (AI) systems for breast cancer screening. Some systems optimize normal mammography examination triaging, while others aim to improve cancer detection. However, no AI system has shown specificity high enough to replace human radiologists, suggesting that AI should play a different role in the breast screening pathway. The decision-referral approach is a promising alternative that has demonstrated the most potential to improve radiologist screening sensitivity and specificity while reducing workload. This collaborative human-AI approach combines AI pre-screening to triage normal examinations and post-screening to prevent missed cancers. The actual performance of decision-referral, including the interaction with human radiologists, can ideally be evaluated in a prospective real-world setting. Trial design. The PRAIM (PRospective, multicenter observational study of an integrated AI system with live monitoring to support breast cancer screening) study (German Trial Register: DRKS00027322) is a prospective controlled observational non-inferiority study to compare the use of CE-marked screening software including AI support (Vara) via the decision-referral approach, with standard screening for women participating in the German breast cancer screening program. Ethics approval was obtained from the University of Lübeck Research Ethics Committee (22-043). Examinations assessed by readers using Vara are compared to examinations without Vara (control). Eligibility criteria and target accrual. Women ages 50 to 69 years old undergoing biennial breast cancer screening within the national screening program are eligible for inclusion. We expect the inclusion of approximately 400,000 women within the inclusion period of 1.5 years. Statistical methods. The primary outcome is the screen-detected cancer rate, defined as biopsy-confirmed cancer diagnoses per 1000 screening examinations. For each screening site, rates over the prospective observation period are calculated for examinations read with AI and without. To control for systematically different screen-detected cancer rates across screening sites, a historical 5-year rate is computed for each site and subtracted from the corresponding prospective rates. Non-inferiority of the screen-detected cancer rate for the AI group compared to the control group is evaluated with a weighted, mixed-effects linear regression model. AI is considered as non-inferior if the lower bound of the two-sided 95 % confidence interval for the estimated difference in screen-detected cancer rates of AI and non-AI group is not below -10 %, which corresponds to a deviation of -0.6 screen-detected cancers per 1000 examinations. Citation Format: Danalyn Byng, Nora Eisemann, Dominik Schüler, Stefan Bunk, Christian Leibig, Moritz Brehmer, Susanne Elsner, Alexander Katalinic. The PRAIM study: A prospective multicenter observational study of an integrated Artificial Intelligence system with live monitoring [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr OT3-18-03.

Cost-effectiveness evaluation of risk-based breast cancer screening in Urban Hebei Province

To evaluate the implementations of Cancer Screening Program in Urban Hebei and to model the cost-effectiveness of a risk-based breast Cancer Screening Program. Women aged 40–74 years were invited to participate the Cancer Screening Program in Urban Hebei form 2016 to 2020 by completing questionnaires to collect information about breast cancer exposure. Clinical screening including ultrasound and mammography examination were performed. We developed a Markov model to estimate the lifetime costs and benefits, in terms of quality-adjusted life years (QALY), of a high-risk breast Cancer Screening Program. Nine screening strategies and no screening were included in the study. The age-specific incidence, transition probability data and lifetime treatment costs were derived and adopted from other researches. Average cost-effectiveness ratios (ACERs) were estimated as the ratios of the additional costs of the screening strategies to the QLYG compared to no screening. Incremental cost-effectiveness ratios (ICERs) were calculated based on the comparison of a lower cost strategies to the next more expensive and effective strategies after excluding dominated strategies and extendedly dominated strategies. ICERs were used to compare with a willingness-to-pay (WTP) threshold. Sensitivity analysis was explored the influence factors. A total of 84,029 women completed a risk assessment questionnaire, from which 20,655 high-risk breast cancer females were evaluated, with a high-risk rate of 24.58%. There were 13,392 high-risk females completed the screening program, with participation rate was 64.84%. Undergoing ultrasound, mammography and combined screening, the suspicious positive detection rates were 15.00%, 9.20% and 19.30%, and the positive detection rates were 2.11%, 2.76% and 3.83%, respectively. According to the results by Markov model, at the end of 45 cycle, the early diagnosis rates were 55.53%, 60.68% and 62.47% underwent the annual screening by ultrasound, mammography and combined, the proportion of advanced cancer were 17.20%, 15.85% and 15.36%, respectively. Different screening method and interval yield varied. In the exploration of various scenarios, annual ultrasound screening is the most cost-effective strategy with the ICER of ¥116,176.15/QALY. Sensitivity analyses demonstrated that the results are robust. Although it was not cost effective, combined ultrasound and mammography screening was an effective strategy for higher positive detection rate of breast cancer. High-risk population-based breast cancer screening by ultrasound annually was the most cost-effective strategy in Urban Hebei Province.