Multi-scale feature fusion for breast cancer detection using circular dilated convolutional transformer optimized by enhanced wombat algorithm

Breast cancer is one of the most prevalent and life-threatening diseases among women worldwide. Early detection of breast cancer being critical for increasing survival rates. Ultrasound image is commonly used for breast cancer screening due to its non-invasive, safe, and cost-effective. However, ultrasound images are often of low quality and have significant noise, which can hinder the effectiveness of classification models. This study proposes an enhanced breast cancer classification model that leverages transfer learning in combination with attention mechanisms to improve diagnostic performance. The main contribution of this research is the introduction of Dense-SASE, a novel architecture that combines DenseNet-121 with two powerful attention modules: Scaled-Dot Product Attention and Squeeze-and-Excitation (SE) Block. These mechanisms are integrated to improve feature representation and allow the model to focus on the most relevant regions of the ultrasound images. The proposed method was evaluated on a publicly available breast ultrasound image dataset, with classification performed across three categories: normal, benign, and malignant. Experimental results demonstrate that the Dense-SASE model achieves an accuracy of 98.29%, a precision of 97.97%, a recall of 98.98%, and an F1-score of 98.44%. Additionally, Grad-CAM visualizations demonstrated the model's capability to localize lesion areas effectively, avoiding non-informative regions, and confirming the model's interpretability. In conclusion, the Dense-SASE model significantly improves the accuracy and reliability of breast cancer classification in ultrasound images. By effectively learning and focusing on clinically relevant features, this approach offers a promising solution for computer-aided diagnosis (CAD) systems and has the potential to assist radiologists in early and accurate breast cancer detection.

Background: Our team has shown that younger versus older women, and both Blacks and Hispanics breast cancer (BC) survivors versus White women have a higher mortality after a second cancer. We have also shown that county-level neighborhood poverty is associated with increased cancer mortality after a BC diagnosis. In this study, we assess to what extent neighborhood-level socioeconomic status (SES) contributes to high BC mortality after second cancers and compare the association to BC mortality after a first cancer. We hypothesize that social determinants are critical to address in women who develop a second cancer despite now being part of the health system. Methods: Women diagnosed with a first BC and/or second cancer after first BC between 2007 and 2019 were identified from the Maryland (MD) Cancer Registry. Census tract-level neighborhood SES environments for poverty and median income were determined using local spatial autocorrelation analysis with Local Moran’s I of MD’s five-year census-tract estimates from the American Community Survey. MD census tract estimates above or below the state mean were identified as high (H)- or low (L)- tracts, respectively. We defined four neighborhood environments for each SES indicator as H-tracts among H-tracts (HH-tracts); L-tracts among L-tracts (LL-tracts); H-tracts among LL-tracts (HL-tracts); L-tracts among HH-tracts (LH-tracts). Women at least 20 years of age diagnosed with a local or regional primary BC, received surgery, and survived more than 365 days were assigned a tract-level neighborhood SES environment for each SES indicator associated with their first and second BC diagnosis. Time-to-event analyses were used to compare BC deaths across each tract-level SES indicator neighborhood environment separately. The data was analyzed using Multivariate Cox proportional hazard regression models adjusted for age at diagnosis, tract median income, cancer stage, treatment, and Estrogen Receptor status. Models were also stratified by race/ethnicity (Non-Hispanic White, Non-Hispanic Black), age at diagnosis (<50, 50-69, =>70), and treatment (hormone therapy and chemotherapy). Results: The analytical population included 15,149 MD women diagnosed with early-stage BC, of which 2,436 developed a second cancer, and BC deaths after first BC and second cancer diagnosis was 2039 and 315, respectively. We detected a 66% (HR=1.66, 95%CI=1.09, 2.53), 14% (HR=1.14, 95%CI=0.66,1.97), and 15% (HR=1.15, 95%CI=0.73,1.81) increase in BC death for women in LL, LH, and HL median income census tracts when compared to HH median income census tracts in the neighborhood income model after second cancer diagnosis. A 30% (HR=1.30, 95%CI=0.89, 1.90), 31% (HR=1.31, 95%CI=0.90, 1.93), and 20% (HR=1.20 95%CI=0.80, 1.79) increase in BC deaths for women in HH, LH, and HL poverty census tracts when compared to LL poverty census tracts was observed in the neighborhood poverty model after second cancer diagnosis. Notably, we observed increased BC death for women in LL, LH, and HL income neighborhood environments as well as in HH and HL poverty neighborhood environments at first BC diagnosis though not as robust as the second cancer neighborhood environment observations. Finally, neighborhood-level SES environments at first BC diagnosis were not statistically associated with BC death upon second diagnosis. Conclusion: Overall, our findings highlight the importance of neighborhood-level SES, such as poverty and median income, at the time of second cancer diagnosis as a strong indicator of BC mortality after second cancer, possibly even more than after a first BC. These findings provide initial evidence of the need for continued surveillance of neighborhood-level SES factors after a first BC diagnosis, and early interventions to improve survival among this group of high-risk women. Citation Format: Ifeoma Nwigwe, Michael Desjardins, Kala Visvanathan. The impact of neighborhood-level social determinants on breast cancer mortality after a second primary breast cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P5-04-03.

Use of output from a computerized analysis of a breast image by radiologists may help them in the tasks of detection or diagnostic, and potentially improve the overall interpretation of breast images and the subsequent patient care. Many factors motivate the attempts to aid or automate radiological diagnosis. Inadequacies in interpretation performance may be due to the presence of image noise or normal anatomical structure as well as to known limitations in the human search and perception process. Developments in breast CAD have led to clinically‐used detection systems in screening mammography and preclinical classification systems for diagnostic breast imaging. Diagnostic breast CAD systems provide output for the characterization of lesions on special‐view mammography, breast sonography, and breast MRI in order to aid in patient management decisions, such as biopsy decisions. As imaging continues to expand in the digital era, computer‐aided diagnosis (CAD) may become an integrated tool in the online diagnostic workup of suspect breast lesions using multi‐modality images and advanced PACS. This presentation reviews research in computerized analysis of mammographic, sonographic, and magnetic resonance breast images for detection and diagnosis. It will include the characterization of lesions and the estimation of the probability of malignancy for use in the diagnostic workup of suspect lesions. CAD systems in diagnostic workup usually involve having the computer extract the margin of the lesion from the surrounding parenchyma, extract characteristics (features) of the lesions, merge these computer‐extracted features into an estimate of the probability of malignancy, and as an option, retrieve automatically similar lesions from an online reference library. The aim of CAD in diagnostic workup is to increase classification sensitivity and specificity as well as to reduce intra‐and inter‐observer variability. While the breadth and depth of CAD is increasing, continued and expanded efforts are needed for collecting and confirming databases, establishing methods for evaluation, integrating effectively and efficiently with PACS and RIS systems, and providing means for clinical evaluation.Learning objectives:1. To appreciate the development of CAD in multi‐modality imaging for detection and diagnosis of breast cancer.2. To understand the benefits and challenges as CAD moves into the digital era and is integrated with PACS and HIS.3. To recognize the necessary steps for advancing and integrating CAD clinically.COI: Maryellen Giger is a shareholder and consultant, and receives research funding from R2 Technology, Inc.

Breast cancer (BC) and genital organ cancers (GOC) are known to have a major impact on the quality of life of patients. The aim of this study was to analyze the risk of depression, anxiety, and adjustment disorders in women in Germany with a suspected but unconfirmed diagnosis of BC or GOC in their medical history. This study included women who received a suspected diagnosis of BC or GOC and were followed between 2007 and 2015 (index date). These women were matched (1:1:1) by age to women with a confirmed diagnosis of BC or GOC and women without a cancer diagnosis. The main outcome measure of the study was the rate of depression, anxiety, and adjustment disorder diagnoses within 3years of the index date. The present analysis included a total of 4,842 patients (mean age=49.3years). Within 3years of the index date, 23.5% of women with a confirmed diagnosis of BC or GOC, 14.1% of those with a suspected diagnosis of BC or GOC, and 10.5% of those without a cancer diagnosis developed depression, anxiety, or an adjustment disorder (log-rank p value <0.001). Women with a suspected diagnosis of cancer were at a higher risk for these psychiatric conditions than those without a cancer diagnosis (BC and GOC: HR 1.32; BC: HR 1.21; GOC: HR 1.50). A suspected diagnosis of BC or GOC in a woman's medical history is associated with an increased risk of developing depression, anxiety, and adjustment disorders.

Breast cancer is the second leading cause of death after lung cancer in women all over the world. The survival rate of breast cancer patients depends on the stage of diagnosis; patients with stage 0 are more likely to reach cancer free state. Therefore, early detection of breast cancer is the key to patient survival. In order to enhance diagnostic accuracy of breast cancer, computer aided diagnosis (CAD) systems have been built. Ultrasound is one of the most frequently used methods for early detection of breast cancer. Currently, the accuracy of CAD systems based on ultrasound images is about 90% and needs further enhancement in order to save lives of the undetected. A meaningful approach to do this is to explore new and meaningful features with discriminating ability and incorporate them into CAD systems. Recently, from a thorough investigation of the images, we extracted a new geometric feature related to the mass shape in ultrasound images called Central Regularity Degree (CRD). The CRD reflects the degree of regularity of the middle part of the mass. To demonstrate the effect of CRD on differentiating malignant from benign masses and the potential improvement to the diagnostic accuracy of breast cancer using ultrasound, this study evaluated the diagnostic accuracy of different classifiers when the CRD was added to five powerful mass features obtained from previous studies including one geometric feature: Depth-Width ratio (DW); two morphological features: shape and margin; blood flow and age. Artificial Neural Networks (ANN), K Nearest Neighbour (KNN), Nearest Centroid, Linear Discriminant Analysis (LDA), and Receiver Operating Characteristic (ROC) analysis were employed for classification and evaluation. Ninety nine breast sonograms- 46 malignant and 53 benign- were evaluated. The results reveal that CRD is an effective feature discriminating between malignant and benign cases leading to improved accuracy of diagnosis of breast cancer. The best results were obtained by ANN where the area under ROC curve (Az) for training and testing using all features except CRD was 100% and 81.8%, respectively, and 100% and 95.45% using all features. Therefore, the overall improvement by adding CRD was about 14%, a significant improvement.

Ki67, PCNA, and MCM proteins: Markers of proliferation in the diagnosis of breast cancer

BackgroundThe diagnosis of thyroid cancer and distinguishing benign from malignant thyroid nodules by junior radiologists can be challenging. This study aimed to develop a computer-aided diagnosis (CAD) system based on the Thyroid Imaging Reporting and Data System (TI-RADS) to distinguish benign from malignant thyroid nodules by analyzing ultrasound images to improve the diagnostic performance of junior radiologists.Material/MethodsA modified TI-RADS based on a convolutional neural network (CNN) was used to develop the CAD system. This retrospective study reviewed 789 thyroid nodules from 695 patients and included radiologists with different diagnostic experience. Five study groups included the CAD group, the junior radiologist group, the intermediate-level radiologist group, the senior radiologist group, and the group in which the junior radiologist used the CAD system. The ultrasound findings were reviewed and compared with the histopathology diagnosis.ResultsThe CAD system for the diagnosis of thyroid cancer showed an accuracy of 80.35%, a sensitivity of 80.64%, a specificity of 80.13%, a positive predictive value (PPV) of 76.02%, a negative predictive value (NPV) of 84.12%, and an area under the receiver operating characteristic (ROC) curve (AUC) of 0.87. The accuracy of the junior radiologists in diagnosing thyroid cancer using CAD was similar to that of intermediate-level radiologists (79.21% vs. 77.57%; P=0.427).ConclusionsThe use of ultrasound CAD based on the TI-RADS showed potential for distinguishing between benign and malignant thyroid nodules and improved the diagnostic performance of junior radiologists.

Several studies have demonstrated an increased frequency of cancer in patients with systemic sclerosis (SSc), specially lung and breast cancers. The pathogenesis of the association between SSc and cancer is not fully established. The aim of this study was to describe new cases of the association between SSc and breast cancer and to perform a review of the literature. We retrospectively studied the medical files of eight patients followed in our institution for SSc and breast cancer. We analyzed them with data available in the literature for a total of 46 patients. Cutaneous extension of SSc was clearly mentioned in 17 cases: the SSc was limited in 10 cases and diffuse in 7 cases The median age at the diagnosis of cancer was 54 years (range: 40-71). The median duration between SSc onset and breast cancer diagnosis was 11.5 months (range: 0-288). The duration between SSc onset and breast cancer diagnosis was < or = 12 months in 27 of 44 patients (61.4%), and in 11 (25%) of them the diagnosis of both diseases was made simultaneously. It was clearly mentioned for 35 patients whether the diagnosis of breast cancer was made before or after the onset of SSc. The diagnosis of breast cancer was made before SSc onset in 17 of 35 patients (48.6%) and after SSc onset in 18 of 35 patients (51.4%). For 33 patients, the follow-up was available: 18 (54.5%) died, 11 (33.3%) of them within the 1st year after the diagnosis of the cancer. For none of the patients did the anticancer treatment improve the SSc. The close temporal relationship between SSc onset and breast cancer diagnosis is highly suggestive of a pathophysiological link. SSc is probably not a paraneoplastic disease since the anticancer treatment has no influence on the evolution of SSc. However, it can be suggested that SSc could be a disease facilitating breast cancer and/or metastases development.

This study was to analyze the clinical application value of magnetic resonance imaging (MRI) image features based on intelligent algorithms in the diagnosis and treatment of breast cancer and to provide an effective reference assessment for breast cancer diagnosis. The MRI diagnosis model (ACO-MRI) based on the ant colony algorithm (ACO) was proposed, which was compared with the diagnosis methods based on support vector machine (SVM) and proximity (KNN) algorithm, and the proposed algorithm was applied to MRI images to diagnose breast cancer. The results showed that the accuracy, sensitivity, and specificity of the ACO-MRI model were greater than those of the KNN and SVM algorithm. Moreover, the specificity was statistically considerable compared with the two algorithms of KNN and SVM ( P &lt; 0.05 ). By comparing 1/5 number of ants and the average gray path of the ACO-MRI model under 1/8 number of ants, it was found that the average gray path value of 1/8 number of ants was greatly higher than the average gray path value of 1/5 number of ants ( P &lt; 0.05 ). The differences in the overall distribution of breast MRI imaging features among Luminal A, Luminal B, HER-2 overexpression, and TN were compared. There were considerable differences in the overall distribution of the three breast MRI imaging features of the boundaries, morphology, and enhancement methods among the four groups ( P &lt; 0.05 ). In short, MRI image based on the intelligent algorithm ACO-MRI diagnosis model can effectively improve the diagnosis effect of breast cancer. Its image feature boundaries, morphology, and enhancement methods had good imaging features in the diagnosis of breast cancer.

Among patients with breast cancer, tumors that contain estrogen receptors (ER) are associated with improved survival and better response to hormone therapy than those not expressing these receptors. The purpose of these case comparison studies was to examine the relationship between carotenoids, vitamin A, and the tumor ER status in women at diagnosis of primary breast cancer. The focus of the first study was the relationship between dietary intake and ER status, and the focus of the second study was the relationship between ER status and the plasma carotenoid, retinol, and tocopherol concentrations. We evaluated tumor ER status and self-reported dietary intake in 142 women and plasma concentrations of carotenoids, retinol, and tocopherols in 149 women, at diagnosis of breast cancer, before any medical or surgical treatment. In the first study the overall odds of ER-positive status were increased in relation to number of mammograms in the past five years, number of breast-fed babies, dietary carotenoid intake, and more frequent intake of yellow and green vegetables. Overall odds of ER-positive status were decreased in relation to years of oral contraceptive use and preformed vitamin A intake. In the second study older women, women with higher plasma lutein concentration, and women not using beta-carotene supplements were more likely to be ER positive, when data were adjusted for body mass index and factors that may influence breast cancer risk or hormonal status. Significant independent relationships between plasma retinol or tocopherol concentrations and ER status were not observed. The strong and independent relationships between carotenoid intake, plasma lutein concentration, and ER status may relate to observations linking a carotenoid-rich diet with improved prognosis after diagnosis of breast cancer.

Combined anxiety and depressive symptoms before diagnosis of breast cancer

1628P Impact of COVID-19 pandemia on the diagnosis of breast cancer in one region of north of Portugal: One year experience

BackgroundWomen’s return to work after diagnosis of breast cancer (BC) is becoming more prevalent. However, register-based national investigation on sickness absence (SA) and disability pension (DP) in BC women is lacking. The aim of the study was to explore SA and DP before and after a first BC diagnosis and the possibility to predict new cancer-related SA by using disease-related and sociodemographic factors.MethodsA longitudinal register study of the 3536 women in Sweden aged 19–64 with a first BC diagnosis in 2010 was conducted by linkage of five nationwide registers. Particularly, detailed information on SA and DP was obtained from the National Social Insurance Agency. Descriptive statistics on SA and DP 2 years before through 3 years after the BC diagnosis were performed. The risk of having a new SA spell due to BC or BC-related diagnoses was modeled using logistic regression.ResultsThe proportion of women with SA increased during the year following the BC diagnosis date and declined over the next 2 years to proportions before diagnosis. At the time of BC diagnosis, half of the women began a new SA spell > 14 days with cancer, cancer-related, or mental diagnosis. Disease-related and sociodemographic factors including occupational sector, living area, age, cancer stage, educational level, and number of previous SA days showed statistical significance (p < 0.05) in predicting a new SA around BC diagnosis. By using these factors, it was possible to correctly predict 67% of the new SA spell.ConclusionsSA among women with BC was elevated mainly in the first year after diagnosis. New SA following BC diagnosis can accurately be predicted.

Histopathological analysis is important for detection of the breast cancer (BC). Computer-aided diagnosis and detection systems are developed to assist the radiologist in the diagnosis process and to relieve the patient from unnecessary pain. In this study, a computer-aided diagnosis system for the early detection of benign and malignant breast cancer is proposed. The proposed system consists of feature extraction, feature ensemble, and classification stages. Various preprocessing steps such as grayscale conversion, noise filtering, and image resizing are employed on the input histopathological images. The local texture descriptors namely Local Binary Pattern (LBP), Frequency Decoded LBP (FDLBP), Binary Gabor Pattern (BGP), Local Phase Quantization (LPQ), Binarized Statistical Image Features (BSIF), CENsus TRansform hISTogram (CENTRIST), and Pyramid Histogram of Oriented Gradients (PHOG) are employed for feature extraction from the histopathologic images. The obtained features are then concatenated for the construction of the ensemble of the features. Three classifiers namely Support Vector Machines (SVM), K-nearest neighbor (KNN), and Neural Networks (NN) are used in the detection of the BC and the classification accuracy score is used for performance evaluation. A dataset called BreaKHis is used in the studies. There are 9109 microscopic pictures in BreaKHis, with 2480 benign samples and 5429 malignant samples. During the collection of the data, 82 patients' breast tumor tissues were envisioned using various magnification factors such as 40X, 100X, 200X, and 400X. The accuracy score is used to assess the acquired findings. The results show that the proposed method has the potential to use accurate BC detection.

&lt;p&gt;A Computer-Aided Diagnosis (CAD) system can perform an accurate diagnosis and help radiologists by presenting a second opinion about breast density. However, the development of a robust CAD system for breast density classification is still an open problem. In this study, we proposed a CAD system based on hybrid intelligent machine learning technique for automatic classification of breast density on mammogram images. The proposed technique employs gradient orientation pattern HOG and texture descriptor CLBP-HF as features and K Nearest Neighbor (KNN) as classifier. The experiments were carried out on benchmarks public domain MIAS and DDSM datasets. The classification accuracy is 96.4% whereas recall and precision are 96.59 and 96.75% on MIAS dataset. Moreover, the comparison with the state-of-the-art breast density classification methods shows that the proposed method outperforms the existing methods on both MIAS and DDSM datasets, the improvement is significant on both datasets. The proposed method will help radiologists in assessing the breast density, which is important for breast cancer diagnosis.&lt;/p&gt;