3D Breast Research Articles

Evaluation of breast skin and tissue stiffness using a non-invasive aspiration device and impact of clinical predictors.

A personalized 3D breast model could present a real benefit for preoperative discussion with patients, surgical planning, and guidance. Breast tissue biomechanical properties have been poorly studied in vivo, although they are important for breast deformation simulation. The main objective of our study was to determine breast skin thickness and breast skin and adipose/fibroglandular tissue stiffness. The secondary objective was to assess clinical predictors of elasticity and thickness: age, smoking status, body mass index, contraception, pregnancies, breastfeeding, menopausal status, history of radiotherapy or breast surgery. Participants were included at the Montpellier University Breast Surgery Department from March to May 2022. Breast skin thickness was measured by ultrasonography, breast skin and adipose/fibroglandular tissue stiffnesses were determined with a VLASTIC non-invasive aspiration device at three different sites (breast segments I-III). Multivariable linear models were used to assess clinical predictors of elasticity and thickness. In this cohort of 196 women, the mean breast skin and adipose/fibroglandular tissue stiffness values were 39 and 3 kPa, respectively. The mean breast skin thickness was 1.83 mm. Only menopausal status was significantly correlated with breast skin thickness and adipose/fibroglandular tissue stiffness. The next step will be to implement these stiffness and thickness values in a biomechanical breast model and to evaluate its capacity to predict breast tissue deformations.

Decellularized extracellular matrix-based bioengineered 3D breast cancer scaffolds for personalized therapy and drug screening.

Breast cancer (BC) is the second deadliest cancer after lung cancer. Similar to all cancers, it is also driven by a 3D microenvironment. The extracellular matrix (ECM) is an essential component of the 3D tumor micro-environment, wherein it functions as a scaffold for cells and provides metabolic support. BC is characterized by alterations in the ECM. Various studies have attempted to mimic BC-specific ECMs using artificial materials, such as Matrigel. Nevertheless, research has proven that naturally derived decellularized extracellular matrices (dECMs) are superior in providing the essential in vivo-like cues needed to mimic a cancer-like environment. Developing in vitro 3-D BC models is not straightforward and requires extensive analysis of the data established by researchers. For the benefit of researchers, in this review, we have tried to highlight all developmental studies that have been conducted by various scientists so far. The analysis of the conclusions drawn from these studies is also discussed. The advantages and drawbacks of the decellularization methods employed for generating BC scaffolds will be covered, and the review will shed light on how dECM scaffolds help develop a BC environment. The later stages of the article will also focus on immunogenicity issues arising from decellularization and the origin of the tissue. Finally, this review will also discuss the biofabrication of matrices, which is the core part of the bioengineering process.

3D Breast Cancer Model on Silk Fibroin-Integrated Microfluidic Chips.

To imitate in vivo environment of cells, microfluidics offer controllable fashions at micro-scale and enable regulate flow-related parameters precisely, leveraging the current state of 3D systems to 4D level through the inclusion of flow and shear stress. In particular, integrating silk fibroin as an adhering layer with microfluidic chips enables to form more comprehensive and biocompatible network between cells since silk fibroin holds outstanding mechanical and biological properties such as easy processability, biocompatibility, controllable biodegradation, and versatile functionalization. In this chapter, we describe design and fabrication of a microfluidic chip, with silk fibroin-covered microchannels for the formation of 3D structures, such as MCF-7 (human breast cancer) cell spheroids as a model system. All the steps performed here are characterized by surface-sensitive tools and standard tissue culture methods. Overall, this strategy can be easily integrated into various high-tech application areas such as drug delivery systems, regenerative medicine, and tissue engineering in near future.

Smartphone-based augmented reality patient education in radiation oncology

We built an augmented reality (AR) patient education application for portable iOS and Android devices that allows patients to view a virtual simulation of themselves receiving radiation treatment. We created software that reads data from the clinical treatment planning system and renders the patient’s actual radiotherapy plan in AR on a tablet or smartphone. The patient's CT simulation data are converted into a 3D translucent virtual human shown being treated with visible radiation beams from a virtual linear accelerator. We conducted a patient study to determine if showing patients this AR simulation improves patient understanding of radiotherapy and/or reduces anxiety about treatment. A total of 75 patients completed this study. The most common plans were 3D breast tangents and intensity modulated radiotherapy lung plans. Patients were administered questionnaires both before and after their AR viewing experience. After their AR viewing, 95% of patients indicated that they had a better understanding of how radiotherapy will be used to treat their cancer. Of the 35 patients who expressed anxiety about radiotherapy beforehand, 21 (60%) indicated that they had decreased anxiety after the AR session. In our single-arm prospective patient study, we found that this simplified low-cost tablet-based personalized AR simulation can be a helpful educational tool for cancer patients undergoing radiotherapy.

Deep learning-based approach in surface thermography for inverse estimation of breast tumor size

Background and objectiveIn early breast cancer diagnosis, tumor size is key to improving the patient's survival chances. It helps doctors to determine the adequate treatment for each case. Surface thermography presents encouraging results to detect thermal patterns of breast tumor abnormality. However, the early and accurate estimation of tumor size based on temperature characteristics is quite challenging due to unavailability of labeled clinical data. This work proposes a Feed-Forward Deep Neural Network (FF-DNN) for an inverse estimation of breast tumor size using thermographic data. MethodsA 3D breast model was created by the COMSOL Multiphysics software incorporating tumors in the gland and covered by fat, muscle, and skin layers. Several tumor configurations are included to generate a large amount of training data. An initial thermal data analysis was performed to affirm the influence of breast tumor size on the skin surface temperature. Then, 1400 different cases were prepared, and the relevant features were extracted to train the deep learning model. The coefficient of determination (R2) and the Mean Square Error (MSE) were used as metrics for evaluating the predictive model. ResultsThe analysis of the normalized temperature variations demonstrated the influence of tumor size on the surface temperature of the breast. Thus, the comparison between the FF-DNN against the Convolutional Neural Network (CNN) showed the reliability of the proposed approach. As result, the prediction accuracy indicates the capacity of the proposed FF-DNN model to estimate tumor size from the provided relevant features with an MSE value of 0.194 and an R2 value of 0.998. ConclusionThe findings of this study indicate that surface thermography, when paired with deep learning, holds promise as a valuable diagnostic tool to improve the prognosis of breast cancer.

Transforming electronic plastics into bioadaptive 3D porous construct for advanced cell culture applications

Electronic plastics (e-plastics) are indispensable in modern society, but their low recycling rate and environmental persistence have raised significant concerns. Prevailing plastic recycling strategies are inadequate to fully capture the economic benefits inherent to e-plastics, providing limited incentives for recycling. Therefore, there is an urgent need to develop innovative approaches aimed at maximizing the capture of value from e-plastics. Herein, acrylonitrile butadiene styrene (ABS) from discarded keyboards was unconventionally “re-tooled” to produce highly porous bioadaptive 3D sponge-like constructs for advanced in vitro applications. The ABSponge was surface functionalized via layer-by-layer (LBL) electrostatic deposition method to generate 3D human breast, colorectal and bone cancer spheroids as a drug screening tool or adapted for co-culturing of cancer spheroids and cancer-associated-fibroblasts to emulate the complex tumor niche. Collectively, our findings reveal the promising potential of using discarded keyboards as a "waste-to-resource" feedstock for advanced in-vitro biotechnological applications, achieving waste reduction and maximizing value-capture.

Single-port three-dimensional (3D) endoscopic-assisted breast surgery—preliminary results and patient-reported satisfaction in 145 breast cancer and gynecomastia cases

BackgroundMinimal-accessed (robotic and endoscopic) breast cancer surgery is increasingly performed due to better cosmetic results and acceptable oncological outcomes. This study aims to demonstrate the clinical safety and patient-reported cosmetic satisfaction of single-port three-dimensional endoscopic-assisted breast surgery (S-P 3D EABS), which is our new endoscopic surgical innovation, in both malignant and benign breast conditions.MethodsPatients who underwent S-P 3D EABS from 1 August 2018 to 31 July 2022 in a single institution were enrolled. Clinical outcomes of this procedure were retrospectively reviewed, and the patient-reported cosmetic satisfaction was evaluated by a questionnaire and reported herein.ResultsDuring the study period, 145 patients underwent 164 procedures of S-P 3D EABS. One hundred fifty (91.5%) procedures were endoscopic-assisted nipple-sparing mastectomy (S-P 3D E-NSM; 117 therapeutic procedures for breast cancer, 13 prophylactic mastectomies, 20 procedures for gynecomastia). Fourteen (8.5%) procedures of endoscopic-assisted breast-conserving surgery (S-P 3D E-BCS) were performed (12 S-P 3D E-BCS, 2 S-P 3D E-BCS with 3D videoscope-assisted partial breast reconstruction, which was 1 case of latissimus dorsi flap and 1 case of omental flap). The mean operative time was 245 ± 110 min in S-P 3D E-NSM and 260 ± 142 min in S-P 3D E-BCS. The mean intraoperative blood loss was 49.7 ± 46.9 ml in S-P 3D E-NSM and 32.8 ± 17.5 ml in S-P 3D E-BCS. Subnipple biopsy showed positive malignancy in 3 (2.6%) S-P 3D E-NSM patients. None of the S-P 3D E-BCS patients found margin involvement; however, 3 (2.6%) reported margin involvement in S-P 3D E-NSM patients. Thirty-two complications were found (24.6%): 7 (5.3%) transient nipple-areolar complex (NAC) ischemia, 7 (5.3%) partial NAC necrosis, 1 (0.7%) total NAC necrosis, and 1 (0.7%) implant loss. During the mean follow-up time of 34 months, there were 2 (1.5%) patients with locoregional recurrence, 9 (6.9%) distant metastasis, and 2 (1.5%) mortality. 78.6% (77/98) of patients answering the cosmetic-evaluated questionnaire reported good and excellent overall satisfaction.ConclusionsS-P 3D EABS is a novel surgical innovation, which is able to perform safely in either malignant or benign breast conditions and offer promising cosmetic results.

Microwave Breast Sensing via Deep Learning for Tumor Spatial Localization by Probability Maps.

microwave imaging (MWI) has emerged as a promising modality for breast cancer screening, offering cost-effective, rapid, safe and comfortable exams. However, the practical application of MWI for tumor detection and localization is hampered by its inherent low resolution and low detection capability. this study aims to generate an accurate tumor probability map directly from the scattering matrix. This direct conversion makes the probability map independent of specific image formation techniques and thus potentially complementary to any image formation technique. An approach based on a convolutional neural network (CNN) is used to convert the scattering matrix into a tumor probability map. The proposed deep learning model is trained using a large realistic numerical dataset of two-dimensional (2D) breast slices. The performance of the model is assessed through visual inspection and quantitative measures to assess the predictive quality at various levels of detail. the results demonstrate a remarkably high accuracy (0.9995) in classifying profiles as healthy or diseased, and exhibit the model's ability to accurately locate the core of a single tumor (within 0.9 cm for most cases). overall, this research demonstrates that an approach based on neural networks (NN) for direct conversion from scattering matrices to tumor probability maps holds promise in advancing state-of-the-art tumor detection algorithms in the MWI domain.

3D bioprinted breast tumor-stroma models for pre-clinical drug testing

The use of three-dimensional (3D) bioprinting has been proposed for the reproducible production of 3D disease models that can be used for high-throughput drug testing and personalized medicine. However, most such models insufficiently reproduce the features and environment of real tumors. We report the development of bioprinted in vitro 3D tumor models for breast cancer, which physically and biochemically mimic important aspects of the native tumor microenvironment, designed to study therapeutic efficacy. By combining a mix of breast decellularized extracellular matrix and methacrylated hyaluronic acid with tumor-derived cells and non-cancerous stromal cells of biological relevance to breast cancer, we show that biological signaling pathways involved in tumor progression can be replicated in a carefully designed tumor-stroma environment. Finally, we demonstrate proof-of-concept application of these models as a reproducible platform for investigating therapeutic responses to commonly used chemotherapeutic agents.

A Multi-Layer Breast Cancer Model to Study the Synergistic Effect of Photochemotherapy.

Breast cancer is one of the most common cancers among women. The development of new and effective therapeutic approaches in the treatment of breast cancer is an important challenge in modern oncology. Two-dimensional (2D) cell cultures are most often used in the study of compounds with potential anti-tumor nature. However, it is necessary to develop advanced three-dimensional (3D) cell models that can, to some extent, reflect the physiological conditions. The use of miniature cancer-on-a-chip microfluidic systems can help to mimic the complex cancer microenvironment. In this report, we developed a 3D breast cancer model in the form of a cell multilayer, composed of stromal cells (HMF) and breast cancer parenchyma (MCF-7). The developed cell model was successfully used to analyze the effectiveness of combined sequential photochemotherapy, based on doxorubicin and meso-tetraphenylporphyrin. We proved that the key factor that allows achieving the synergistic effect of combination therapy are the order of drug administration to the cells and the sequence of therapeutic procedures. To the best of our knowledge, studies on the effectiveness of combination photochemotherapy depending on the sequence of the component drugs were performed for the first time under microfluidic conditions on a 3D multilayered model of breast cancer tissue.

Platinum(IV)-Gold(I) Agents with Promising Anticancer Activity: Selected Studies in 2D and 3D Triple-Negative Breast Cancer Models.

New heterometallic binuclear and trinuclear platinum(IV)-gold(I) compounds of the type [Pt(L)n Cl2 (OH){(OOC-4-C6 H4 -PPh2 )AuCl}x ] (L=NH3 , n=2; x=1, 2; L=diaminocyclohexane, DACH, n=1; x=2) are described. These compounds are cytotoxic and selective against a small panel of renal, bladder, ovarian, and breast cancer cell lines. We selected a trinuclear PtAu2 compound containing the PtIV core based on oxaliplatin, to further investigate its cell-death pathway, cell and organelle uptake and anticancer effects against the triple-negative breast cancer (TNBC) MDA-MB-231 cell line. This compound induces apoptosis and accumulates mainly in the nucleus and mitochondria. It also exerts remarkable antimigratory and antiangiogenic properties, and has a potent cytotoxic effect against TNBC 3D spheroids. Trinuclear compounds do not seem to display relevant interactions with calf thymus (CT) DNA and plasmid (pBR322) even in the presence of reducing agents, but inhibit pro-angiogenic enzyme thioredoxin reductase (TrxR) in TNBC cells.

Cost-effective, portable, patient-dedicated three-dimensional automated breast ultrasound for point-of-care breast cancer screening

Breast cancer screening has substantially reduced mortality across screening populations. However, a clinical need persists for more accessible, cost-effective, and robust approaches for increased-risk and diverse patient populations, especially those with dense breasts where screening mammography is suboptimal. We developed and validated a cost-effective, portable, patient-dedicated three-dimensional (3D) automated breast ultrasound (ABUS) system for point-of-care breast cancer screening. The 3D ABUS system contains a wearable, rapid-prototype 3D-printed dam assembly, a compression assembly, and a computer-driven 3DUS scanner, adaptable to any commercially available US machine and transducer. Acquisition is operator-agnostic, involves a 40-second scan time, and provides multiplanar 3D visualization for whole-breast assessment. Geometric reconstruction accuracy was evaluated with a 3D grid phantom and tissue-mimicking breast phantoms, demonstrating linear measurement and volumetric reconstruction errors < 0.2 mm and < 3%, respectively. The system’s capability was demonstrated in a healthy male volunteer and two healthy female volunteers, representing diverse patient geometries and breast sizes. The system enables comfortable ultrasonic coupling and tissue stabilization, with adjustable compression to improve image quality while alleviating discomfort. Moreover, the system effectively mitigates breathing and motion, since its assembly affixes directly onto the patient. While future studies are still required to evaluate the impact on current clinical practices and workflow, the 3D ABUS system shows potential for adoption as an alternative, cost-effective, dedicated point-of-care breast cancer screening approach for increased-risk populations and limited-resource settings.

P22-28: Cisplatin loaded single-walled carbon nanotubes-induced apoptosis: a comparative study on 2D and 3D breast cancer cultures

P22-28: Cisplatin loaded single-walled carbon nanotubes-induced apoptosis: a comparative study on 2D and 3D breast cancer cultures

Biological evaluation of [4-(4-aminophenyl)-1-(4-fluorophenyl)-1H-pyrrol-3-yl](3,4,5-trimethoxyphenyl)methanone as potential antineoplastic agent in 2D and 3D breast cancer models.

Targeting tubulin polymerization and depolymerization represents a promising approach to treat solid tumors. In this study, we investigated the molecular mechanisms underlying the anticancer effects of a structurally novel tubulin inhibitor, [4-(4-aminophenyl)-1-(4-fluorophenyl)-1H-pyrrol-3-yl](3,4,5-trimethoxyphenyl)methanone (ARDAP), in two- and three-dimensional MCF-7 breast cancer models. At sub-cytotoxic concentrations, ARDAP showed a marked decrease in cell proliferation, colony formation, and ATP intracellular content in MCF-7 cells, by acting through a cytostatic mechanism. Additionally, drug exposure caused blockage of the epithelial-to-mesenchymal transition (EMT). In 3D cell culture, ARDAP negatively affected tumor spheroid growth, with inhibition of spheroid formation and reduction of ATP concentration levels. Notably, ARDAP exposure promoted the differentiation of MCF-7 cells by inducing: (i) expression decrease of Oct4 and Sox2 stemness markers, both in 2D and 3D models, and (ii) downregulation of the stem cell surface marker CD133 in 2D cell cultures. Interestingly, treated MCF7 cells displayed a major sensitivity to cytotoxic effects of the conventional chemotherapeutic drug doxorubicin. In addition, although exhibiting growth inhibitory effects against breast cancer cells, ARDAP showed insignificant harm to MCF10A healthy cells. Collectively, our results highlight the potential of ARDAP to emerge as a new chemotherapeutic agent or adjuvant compound in chemotherapeutic treatments.

3D Breast Cancer Segmentation in DCE-MRI Using Deep Learning With Weak Annotation.

Deep learning models require large-scale training to perform confidently, but obtaining annotated datasets in medical imaging is challenging. Weak annotation has emerged as a way to save time and effort. To develop a deep learning model for 3D breast cancer segmentation in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using weak annotation with reliable performance. Retrospective. Seven hundred and thirty-six women with breast cancer from a single institution, divided into the development (N = 544) and test dataset (N = 192). 3.0-T, 3D fat-saturated gradient-echo axial T1-weighted flash 3D volumetric interpolated brain examination (VIBE) sequences. Two radiologists performed a weak annotation of the ground truth using bounding boxes. Based on this, the ground truth annotation was completed through autonomic and manual correction. The deep learning model using 3D U-Net transformer (UNETR) was trained with this annotated dataset. The segmentation results of the test set were analyzed by quantitative and qualitative methods, and the regions were divided into whole breast and region of interest (ROI) within the bounding box. As a quantitative method, we used the Dice similarity coefficient to evaluate the segmentation result. The volume correlation with the ground truth was evaluated with the Spearman correlation coefficient. Qualitatively, three readers independently evaluated the visual score in four scales. A P-value <0.05 was considered statistically significant. The deep learning model we developed achieved a median Dice similarity score of 0.75 and 0.89 for the whole breast and ROI, respectively. The volume correlation coefficient with respect to the ground truth volume was 0.82 and 0.86 for the whole breast and ROI, respectively. The mean visual score, as evaluated by three readers, was 3.4. The proposed deep learning model with weak annotation may show good performance for 3D segmentations of breast cancer using DCE-MRI. 3 TECHNICAL EFFICACY: Stage 2.

Evaluation of Intraoperative Volumetric Assessment of Breast Volume Using 3D Handheld Stereo Photogrammetric Device.

Methods for assessing three-dimensional (3D) breast volume are becoming increasingly popular in breast surgery. However, the precision of intraoperative volumetric assessment is still unclear. Until now, only non-validated scanning systems have been used for intraoperative volumetric analyses. This study aimed to assess the feasibility, handling, and accuracy of a commercially available, validated, and portable device for intraoperative 3D volumetric evaluation. All patients who underwent breast surgery from 2020 to 2022 were identified from our institutional database. Intraoperative 3D volumetric assessments of 103 patients were included in this study. Standardized 3D volumetric measurements were obtained 3 months postoperatively to compare the intraoperatively generated volumetric assessment. All of the study participants were women with a mean age of 48.3 ± 14.7 years (range: 20-89). The mean time for intraoperative volumetric assessment was 8.7 ± 2.6 min. The postoperative 3D volumetric assessment, with a mean volume of 507.11 ± 206.29 cc, showed no significant difference from the intraoperative volumetric measurements of 504.24 ± 276.61 cc (p = 0.68). The mean absolute volume difference between the intraoperative simulations and postoperative results was 27.1 cc. Intraoperative 3D volumetric assessment using the VECTRA H2 imaging system seems to be a feasible, reliable, and accurate method for measuring breast volume. Based on this finding, we plan to investigate whether volumetric objective evaluations will help to improve breast symmetry in the future.

Mimicking 3D breast tumor-stromal interactions to screen novel cancer therapeutics

Most of the 3D breast tumor models used in drug screening studies only comprise tumor cells, keeping out other essential cell players of the tumor microenvironment. Tumor-associated macrophages and fibroblasts are frequently correlated with tumor progression and therapy resistance, and targeting these cells at the tumor site has been appointed as a promising therapeutic strategy. However, the translation of new therapies to the clinic has been hampered by the absence of cellular models that more closely mimic the features of in vivo breast tumor microenvironment. Therefore, the development of innovative 3D models able to provide consistent and predictive responses about the in vivo efficacy of novel therapeutics is still an unmet preclinical need.Herein, we have established an in vitro 3D heterotypic spheroid model including MCF-7 breast tumor cells, human mammary fibroblasts and human macrophages. To establish this model, different cell densities have been combined and characterized through the evaluation of the spheroid size and metabolic activity, as well as histological and immunohistochemistry analysis of the 3D multicellular structures. The final optimized 3D model consisted in a multicellular spheroid seeded at the initial density of 5000 cells and cell ratio of 1:2:1 (MCF-7:monocytes:fibroblasts). Our model recapitulates several features of the breast tumor microenvironment, including the formation of a necrotic core, spatial organization, and extracellular matrix production. Further, it was validated as a platform for drug screening studies, using paclitaxel, a currently approved drug for breast cancer treatment, and Gefitinib, a chemotherapeutic approved for lung cancer and in preclinical evaluation for breast cancer. Generally, the impact on the cell viability of the 3D model was less evident than in 2D model, reinforcing the relevance of such complex 3D models in addressing novel treatment approaches. Overall, the use of a 3D heterotypic spheroid of breast cancer could be a valuable tool to predict the therapeutic effect of new treatments for breast cancer patients, by recapitulating key features of the breast cancer microenvironment.

Mitigating Under-Sampling Artifacts in 3D Photoacoustic Imaging Using Res-UNet Based on Digital Breast Phantom.

In recent years, photoacoustic (PA) imaging has rapidly grown as a non-invasive screening technique for breast cancer detection using three-dimensional (3D) hemispherical arrays due to their large field of view. However, the development of breast imaging systems is hindered by a lack of patients and ground truth samples, as well as under-sampling problems caused by high costs. Most research related to solving these problems in the PA field were based on 2D transducer arrays or simple regular shape phantoms for 3D transducer arrays or images from other modalities. Therefore, we demonstrate an effective method for removing under-sampling artifacts based on deep neural network (DNN) to reconstruct high-quality PA images using numerical digital breast simulations. We constructed 3D digital breast phantoms based on human anatomical structures and physical properties, which were then subjected to 3D Monte-Carlo and K-wave acoustic simulations to mimic acoustic propagation for hemispherical transducer arrays. Finally, we applied a 3D delay-and-sum reconstruction algorithm and a Res-UNet network to achieve higher resolution on sparsely-sampled data. Our results indicate that when using a 757 nm laser with uniform intensity distribution illuminated on a numerical digital breast, the imaging depth can reach 3 cm with 0.25 mm spatial resolution. In addition, the proposed DNN can significantly enhance image quality by up to 78.4%, as measured by MS-SSIM, and reduce background artifacts by up to 19.0%, as measured by PSNR, even at an under-sampling ratio of 10%. The post-processing time for these improvements is only 0.6 s. This paper suggests a new 3D real time DNN method addressing the sparse sampling problem based on numerical digital breast simulations, this approach can also be applied to clinical data and accelerate the development of 3D photoacoustic hemispherical transducer arrays for early breast cancer diagnosis.

Frequency constraint-based adversarial attack on deep neural networks for medical image classification

The security of AI systems has gained significant attention in recent years, particularly in the medical diagnosis field. To develop a secure medical image classification system based on deep neural networks, it is crucial to design effective adversarial attacks that can embed hidden, malicious behaviors into the system. However, designing a unified attack method that can generate imperceptible attack samples with high content similarity and be applied to diverse medical image classification systems is challenging due to the diversity of medical imaging modalities and dimensionalities. Most existing attack methods are designed to attack natural image classification models, which inevitably corrupt the semantics of pixels by applying spatial perturbations. To address this issue, we propose a novel frequency constraint-based adversarial attack method capable of delivering attacks in various medical image classification tasks. Specially, our method introduces a frequency constraint to inject perturbation into high-frequency information while preserving low-frequency information to ensure content similarity. Our experiments include four public medical image datasets, including a 3D CT dataset, a 2D chest X-Ray image dataset, a 2D breast ultrasound dataset, and a 2D thyroid ultrasound dataset, which contain different imaging modalities and dimensionalities. The results demonstrate the superior performance of our model over other state-of-the-art adversarial attack methods for attacking medical image classification tasks on different imaging modalities and dimensionalities.

How Camera-to-Subject Distance and Height Affect Breast Measurement and Follow-Up Assessment in Plastic Surgery.

Due to several factors that affect photograph quality, bias is inevitably present in two-dimensional (2D) breast photography. The principal variables affecting image performance at a fixed focus length are the distance between the camera and the subjects and the photography angles. This study aimed to investigate the effects of camera-to-subject distances and camera height on breast measurement parameters to understand the trend of breast deformation and provide guidance for the accurate evaluation of planar follow-up. We enlisted 16 volunteers with various breast cup sizes (A-D). Frontal and lateral photos were obtained with a steady focus of 50 mm at distances between 1.10 m and 2.20 m and at heights between 30 cm above the nipple and 30 cm below the nipple at intervals of 10 cm. Two researchers independently evaluated each volunteer's breast aesthetic parameters, including 11 linear parameters, 3 area parameters, and 3 ratio parameters, using Vernier calipers and Photoshop. The correlation coefficient of the two investigators ranged from 0.922 to 0.999. The results measured by Photoshop were 29.67 ± 5.23% greater than those of the Vernier caliper (p < 0.01). In contrast to ratio parameters, which showed no significant changes in each distance group (p = 1.00), linear parameters and area parameters significantly increased as object distance decreased (p < 0.05). The lower pole of the breast grew wider and flatter and occupied a larger proportion of the breast as height declined. Camera-to-subject distances of 1.5-1.7 m are recommended for stabilized and uniform breast photography. Varying shooting height affects breast distortion. Quantifying the relationship between photographic conditions and breast morphology enables plastic surgeons to conduct more comprehensive and accurate assessments. This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266. Bullet point list: 1. The breast morphology will get more distortion with a smaller camera-to-subject distance. 2. Camera-to-subject distances of 1.5~1.7m are recommended for stabilized and uniform breast photography. 3. Height rather than distance affects the breast proportion.