A step-by-step review on patient-specific biomechanical finite element models for breast MRI to x-ray mammography registration

Estimation of the biomechanical mammographic deformation of the breast using machine learning models

BACKGROUND

A typical problem in the registration of MRI and X-ray mammography is the nonlinear deformation applied to the breast during mammography. We have developed a method for virtual deformation of the breast using a biomechanical model automatically constructed from MRI. The virtual deformation is applied in two steps: unloaded state estimation and compression simulation. The finite element method is used to solve the deformation process. However, the extensive computational cost prevents its usage in clinical routine.

METHODS

We propose three machine learning models to overcome this problem: an extremely randomized tree (first model), extreme gradient boosting (second model), and deep learning-based bidirectional long short-term memory with an attention layer (third model) to predict the deformation of a biomechanical model. We evaluated our methods with 516 breasts with realistic compression ratios up to 76%.

FINDINGS

We first applied one-fold validation, in which the second and third models performed better than the first model. We then applied ten-fold validation. For the unloaded state estimation, the median RMSE for the second and third models is 0.8 mm and 1.2 mm, respectively. For the compression, the median RMSE is 3.4 mm for both models. We evaluated correlations between model accuracy and characteristics of the clinical datasets such as compression ratio, breast volume, and tissue types.

INTERPRETATION

Using the proposed models, we achieved accurate results comparable to the finite element model, with a speedup of factor 240 using the extreme gradient boosting model. These proposed models can replace the finite element model simulation, enabling clinically relevant real-time application.

A review of bioengineering techniques applied to breast tissue: Mechanical properties, tissue engineering and finite element analysis

Female breast cancer was the most prevalent cancer worldwide in 2020, according to the Global Cancer Observatory. As a prophylactic measure or as a treatment, mastectomy and lumpectomy are often performed at women. Following these surgeries, women normally do a breast reconstruction to minimize the impact on their physical appearance and, hence, on their mental health, associated with self-image issues. Nowadays, breast reconstruction is based on autologous tissues or implants, which both have disadvantages, such as volume loss over time or capsular contracture, respectively. Tissue engineering and regenerative medicine can bring better solutions and overcome these current limitations. Even though more knowledge needs to be acquired, the combination of biomaterial scaffolds and autologous cells appears to be a promising approach for breast reconstruction. With the growth and improvement of additive manufacturing, three dimensional (3D) printing has been demonstrating a lot of potential to produce complex scaffolds with high resolution. Natural and synthetic materials have been studied in this context and seeded mainly with adipose derived stem cells (ADSCs) since they have a high capability of differentiation. The scaffold must mimic the environment of the extracellular matrix (ECM) of the native tissue, being a structural support for cells to adhere, proliferate and migrate. Hydrogels (e.g., gelatin, alginate, collagen, and fibrin) have been a biomaterial widely studied for this purpose since their matrix resembles the natural ECM of the native tissues. A powerful tool that can be used in parallel with experimental techniques is finite element (FE) modeling, which can aid the measurement of mechanical properties of either breast tissues or scaffolds. FE models may help in the simulation of the whole breast or scaffold under different conditions, predicting what might happen in real life. Therefore, this review gives an overall summary concerning the human breast, specifically its mechanical properties using experimental and FE analysis, and the tissue engineering approaches to regenerate this particular tissue, along with FE models.

Incorporating heterogeneity and anisotropy for surgical applications in breast deformation modeling

BACKGROUND

Simulating soft-tissue breast deformations is of interest for many applications including image fusion, longitudinal registration, and image-guided surgery. For the surgical use case, positional changes cause breast deformations that compromise the use of preoperative imaging to inform tumor excision. Even when acquiring imaging in the supine position, which better reflects surgical presentation, deformations still occur due to arm motion and orientation changes. A biomechanical modeling approach to simulate supine breast deformations for surgical applications must be both accurate and compatible with the clinical workflow.

METHODS

A supine MR breast imaging dataset from n = 11 healthy volunteers was used to simulate surgical deformations by acquiring images in arm-down and arm-up positions. Three linear-elastic modeling approaches with varying levels of complexity were used to predict deformations caused by this arm motion: a homogeneous isotropic model, a heterogeneous isotropic model, and a heterogeneous anisotropic model using a transverse-isotropic constitutive model.

FINDINGS

The average target registration errors for subsurface anatomical features were 5.4 ± 1.5 mm for the homogeneous isotropic model, 5.3 ± 1.5 mm for the heterogeneous isotropic model, and 4.7 ± 1.4 mm for the heterogeneous anisotropic model. A statistically significant improvement in target registration error was observed between the heterogeneous anisotropic model and both the homogeneous and the heterogeneous isotropic models (P < 0.01).

INTERPRETATION

While a model that fully incorporates all constitutive complexities of anatomical structure likely achieves the best accuracy, a computationally tractable heterogeneous anisotropic model provided significant improvement and may be applicable for image-guided breast surgeries.

Development and validation of a four-dimensional registration technique for DCE breast MRI

BACKGROUND

Patient motion can degrade image quality of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) due to subtraction artifacts. By objectively and subjectively assessing the impact of principal component analysis (PCA)-based registration on pretreatment DCE-MRIs of breast cancer patients, we aim to validate four-dimensional registration for DCE breast MRI.

RESULTS

After applying a four-dimensional, PCA-based registration algorithm to 154 pretreatment DCE-MRIs of histopathologically well-described breast cancer patients, we quantitatively determined image quality in unregistered and registered images. For subjective assessment, we ranked motion severity in a clinical reading setting according to four motion categories (0: no motion, 1: mild motion, 2: moderate motion, 3: severe motion with nondiagnostic image quality). The median of images with either moderate or severe motion (median category 2, IQR 0) was reassigned to motion category 1 (IQR 0) after registration. Motion category and motion reduction by registration were correlated (Spearman's rho: 0.83, p < 0.001). For objective assessment, we performed perfusion model fitting using the extended Tofts model and calculated its volume transfer coefficient K^trans as surrogate parameter for motion artifacts. Mean K^trans decreased from 0.103 (± 0.077) before registration to 0.097 (± 0.070) after registration (p < 0.001). Uncertainty in perfusion quantification was reduced by 7.4% after registration (± 15.5, p < 0.001).

CONCLUSIONS

Four-dimensional, PCA-based image registration improves image quality of breast DCE-MRI by correcting for motion artifacts in subtraction images and reduces uncertainty in quantitative perfusion modeling. The improvement is most pronounced when moderate-to-severe motion artifacts are present.

Using Breast Tissue Information and Subject-Specific Finite-Element Models to Optimize Breast Compression Parameters for Digital Mammography

Digital mammography has become a first-line diagnostic tool for clinical breast cancer screening due to its high sensitivity and specificity. Mammographic compression force is closely associated with image quality and patient comfort. Therefore, optimizing breast compression parameters is essential. Subjects were recruited for digital mammography and breast magnetic resonance imaging (MRI) within a month. Breast MRI images were used to calculate breast volume and volumetric breast density (VBD) and construct finite element models. Finite element analysis was performed to simulate breast compression. Simulated compressed breast thickness (CBT) was compared with clinical CBT and the relationships between compression force, CBT, breast volume, and VBD were established. Simulated CBT had a good linear correlation with the clinical CBT (R2 = 0.9433) at the clinical compression force. At 10, 12, 14, and 16 daN, the mean simulated CBT of the breast models was 5.67, 5.13, 4.66, and 4.26 cm, respectively. Simulated CBT was positively correlated with breast volume (r > 0.868) and negatively correlated with VBD (r < –0.338). The results of this study provides a subject-specific and evidence-based suggestion of mammographic compression force for radiographers considering image quality and patient comfort.

A particle filter approach to dynamic kidney pose estimation in robotic surgical exposure

PURPOSE

Traditional soft tissue registration methods require direct intraoperative visualization of a significant portion of the target anatomy in order to produce acceptable surface alignment. Image guidance is therefore generally not available during the robotic exposure of structures like the kidneys which are not immediately visualized upon entry into the abdomen. This paper proposes guiding surgical exposure with an iterative state estimator that assimilates small visual cues into an a priori anatomical model as exposure progresses, thereby evolving pose estimates for the occluded structures of interest.

METHODS

Intraoperative surface observations of a right kidney are simulated using endoscope tracking and preoperative tomography from a representative robotic partial nephrectomy case. Clinically relevant random perturbations of the true kidney pose are corrected using this sequence of observations in a particle filter framework to estimate an optimal similarity transform for fitting a patient-specific kidney model at each step. The temporal response of registration error is compared against that of serial rigid coherent point drift (CPD) in both static and simulated dynamic surgical fields, and for varying levels of observation persistence.

RESULTS

In the static case, both particle filtering and persistent CPD achieved sub-5 mm accuracy, with CPD processing observations 75% faster. Particle filtering outperformed CPD in the dynamic case under equivalent computation times due to the former requiring only minimal persistence.

CONCLUSION

This proof-of-concept simulation study suggests that Bayesian state estimation may provide a viable pathway to image guidance for surgical exposure in the abdomen, especially in the presence of dynamic intraoperative tissue displacement and deformation.

Multi-Classification of Breast Cancer Lesions in Histopathological Images Using DEEP_Pachi: Multiple Self-Attention Head

Introduction and Background: Despite fast developments in the medical field, histological diagnosis is still regarded as the benchmark in cancer diagnosis. However, the input image feature extraction that is used to determine the severity of cancer at various magnifications is harrowing since manual procedures are biased, time consuming, labor intensive, and error-prone. Current state-of-the-art deep learning approaches for breast histopathology image classification take features from entire images (generic features). Thus, they are likely to overlook the essential image features for the unnecessary features, resulting in an incorrect diagnosis of breast histopathology imaging and leading to mortality. Methods: This discrepancy prompted us to develop DEEP_Pachi for classifying breast histopathology images at various magnifications. The suggested DEEP_Pachi collects global and regional features that are essential for effective breast histopathology image classification. The proposed model backbone is an ensemble of DenseNet201 and VGG16 architecture. The ensemble model extracts global features (generic image information), whereas DEEP_Pachi extracts spatial information (regions of interest). Statistically, the evaluation of the proposed model was performed on publicly available dataset: BreakHis and ICIAR 2018 Challenge datasets. Result: A detailed evaluation of the proposed model’s accuracy, sensitivity, precision, specificity, and f1-score metrics revealed the usefulness of the backbone model and the DEEP_Pachi model for image classifying. The suggested technique outperformed state-of-the-art classifiers, achieving an accuracy of 1.0 for the benign class and 0.99 for the malignant class in all magnifications of BreakHis datasets and an accuracy of 1.0 on the ICIAR 2018 Challenge dataset. Conclusion: The acquired findings were significantly resilient and proved helpful for the suggested system to assist experts at big medical institutions, resulting in early breast cancer diagnosis and a reduction in the death rate.

Ultrasound waves in tumors via needle irradiation for precise medicine

Grounded in the interdisciplinary crosstalk among physics and biological sciences, precision medicine-based diagnosis and treatment strategies have recently gained great attention for the actual applicability of new engineered approaches in many medical fields, particularly in oncology. Within this framework, the use of ultrasounds employed to attack cancer cells in tumors to induce possible mechanical damage at different scales has received growing attention from scholars and scientists worldwide. With these considerations in mind, on the basis of ad hoc elastodynamic solutions and numerical simulations, we propose a pilot study for in silico modeling of the propagation of ultrasound waves inside tissues, with the aim of selecting proper frequencies and powers to be irradiated locally through a new teragnostic platform based on Lab-on-Fiber technology, baptized as a hospital in the needle and already the object of a patent. It is felt that the outcomes and the related biophysical insights gained from the analyses could pave the way for envisaging new integrated diagnostic and therapeutic approaches that might play a central role in future applications of precise medicine, starting from the growing synergy among physics, engineering and biology.

Artificial intelligence for breast cancer analysis: Trends & directions

Breast cancer is one of the leading causes of death among women. Early detection of breast cancer can significantly improve the lives of millions of women across the globe. Given importance of finding solution/framework for early detection and diagnosis, recently many AI researchers are focusing to automate this task. The other reasons for surge in research activities in this direction are advent of robust AI algorithms (deep learning), availability of hardware that can run/train those robust and complex AI algorithms and accessibility of large enough dataset required for training AI algorithms. Different imaging modalities that have been exploited by researchers to automate the task of breast cancer detection are mammograms, ultrasound, magnetic resonance imaging, histopathological images or any combination of them. This article analyzes these imaging modalities and presents their strengths and limitations. It also enlists resources from where their datasets can be accessed for research purpose. This article then summarizes AI and computer vision based state-of-the-art methods proposed in the last decade to detect breast cancer using various imaging modalities. Primarily, in this article we have focused on reviewing frameworks that have reported results using mammograms as it is the most widely used breast imaging modality that serves as the first test that medical practitioners usually prescribe for the detection of breast cancer. Another reason for focusing on mammogram imaging modalities is the availability of its labelled datasets. Datasets availability is one of the most important aspects for the development of AI based frameworks as such algorithms are data hungry and generally quality of dataset affects performance of AI based algorithms. In a nutshell, this research article will act as a primary resource for the research community working in the field of automated breast imaging analysis.

Measurement of the hyperelastic properties of 72 normal homogeneous and heterogeneous ex vivo breast tissue samples

The mechanical properties of normal soft tissues, including breast tissue, have been of interest to the biomedical research community as there are many clinical and industrial applications that can benefit from quantitative information characterizing such properties. For instance, computer assisted surgery planning, elastography for breast cancer diagnosis, and bra design can all involve biomechanical modeling of the breast to predict its deformation or stress distribution. It is known that most biological soft tissues, including breast tissue, exhibit nonlinear mechanical response over large strains. As such, it is necessary to model such tissues as hyperelastic. In this work, we used indentation testing to estimate the hyperelastic parameters of 4 models (3rd order Ogden, 5-term polynomial, Veronda-Westman and Yeoh) estimated from 72 healthy ex vivo breast tissue samples covering adipose, fibroglandular, and mixed tissue. All estimated parameter sets were confirmed to represent stable material using Drucker's stability criterion. We observed that all three tissue types were statistically similar solidifying the use of homogenous breast modelling over large strain simulation.

The identification of the unloaded configuration of breast tissue with unknown non-homogenous stiffness parameters using surface measured data in deformed configuration

Large deformation analysis of the breast is known as a useful approach for locating the tumor and treatment strategies of breast cancer, for which knowing the breast stiffness parameters and unloaded configuration is crucial to obtain reliable results. In this study, an iterative inverse finite element algorithm is developed to identify the unloaded configuration of the breast while its stiffness constants are unknown and its internal structure is assumed to be non-homogeneous. The position vector of surface points in the deformed configuration of the breast is employed to obtain the unknowns of the inverse problem. An objective function based on the difference between the position vector of the calculated and measured deformed configurations is defined. Thereafter, the objective function is minimized using a gradient-based method. The sensitivity analysis for material parameters is performed using an analytic direct differentiation approach. Through several numerical examples, the effectiveness of the proposed inverse method for identifying the unloaded configuration of a uniform, a computational breast phantom with a single inclusion as well as a computational breast phantom with randomly distributed stiffness, is demonstrated. The effects of the number of load cases, measurement error, and initial guesses on the results of the inverse problem are investigated, as well. It is observed that the unloaded configuration of the computational breast phantom with a single inclusion or heterogeneous breast tissues can be accurately found by considering an equivalent homogenous material for the tissue.

Mechanical and hemodynamic responses of breast tissue under mammographic-like compression during functional dynamic optical imaging

Studying tissue hemodynamics following breast compression has the potential to reveal new contrast mechanisms for evaluating breast cancer. However, how compression will be distributed and, consequently, how hemodynamics will be altered inside the compressed breast remain unclear. To explore the effect of compression, 12 healthy volunteers were studied by applying a step compression increase (4.5-53.4 N) using an optical imaging system capable of concurrently measuring pressure distribution and hemodynamic responses. Finite element analysis was used to predict the distribution of internal fluid pressure (IFP) in breast models. Comparisons between the measured pressure distribution and the reconstructed hemodynamic images for the healthy volunteers indicated significant (p < 0.05) negative correlations. The findings from a breast cancer patient showed that IFP distribution during compression strongly correlates with the observed differential hemodynamic images. We concluded that dynamic breast compression results in non-uniform internal pressure distribution throughout the breast that could potentially drive directed blood flow. The encouraging results obtained highlight the promise of developing dynamic optical imaging biomarkers for breast cancer by interpreting differential hemodynamic images of breast tissue during compression in the context of measured pressure distribution and predicted IFP.

Deformable Mapping Method to Relate Lesions in Dedicated Breast CT Images to Those in Automated Breast Ultrasound and Digital Breast Tomosynthesis Images

This work demonstrates the potential for using a deformable mapping method to register lesions between dedicated breast computed tomography (bCT) and both automated breast ultrasound (ABUS) and digital breast tomosynthesis (DBT) images (craniocaudal [CC] and mediolateral oblique [MLO] views). Two multi-modality breast phantoms with external fiducial markers attached were imaged by the three modalities. The DBT MLO view was excluded for the second phantom. The automated deformable mapping algorithm uses biomechanical modeling to determine corresponding lesions based on distances between their centers of mass (d_COM) in the deformed bCT model and the reference model (DBT or ABUS). For bCT to ABUS, the mean d_COM was 5.2 ± 2.6 mm. For bCT to DBT (CC), the mean d_COM was 5.1 ± 2.4 mm. For bCT to DBT (MLO), the mean d_COM was 4.7 ± 2.5 mm. This application could help improve a radiologist's efficiency and accuracy in breast lesion characterization, using multiple imaging modalities.

Three-layer heterogeneous mammographic phantoms for Monte Carlo simulation of normalized glandular dose coefficients in mammography

Normalized glandular dose (DgN) coefficients obtained using homogeneous breast phantoms are commonly used in breast dosimetry for mammography. However, glandular tissue is heterogeneously distributed in the breast. This study aimed to construct three-layer heterogeneous mammographic phantoms (THEPs) to examine the effect of glandular distribution on DgN coefficient. Each layer of THEPs was set to 25%, 50%, or 75% glandular fraction to emulate heterogeneous glandular distribution. Monte Carlo simulation was performed to attain mean glandular dose (MGD) and air kerma at 22-36 kVp and W/Al, W/Rh, and W/Ag target-filter combinations. The heterogeneous DgN coefficient was calculated as functions of the mean glandular fraction (MGF), breast thickness, tube voltage, and half-value layer. At 50% MGF, the heterogeneous DgN coefficients for W/Al, W/Rh, and W/Ag differed by 40.3%, 36.7%, and 31.2%. At 9-cm breast thickness, the DgN values of superior and inferior glandular distributions were 25.4% higher and 29.2% lower than those of uniform distribution. The proposed THEPs can be integrated with conventional breast dosimetry to consider the heterogeneous glandular distribution in clinical practice.

Anthropomorphic breast model repository for research and development of microwave breast imaging technologies

A repository of anthropomorphic numerical breast models is made available for the scientific community to support research and development of microwave imaging technologies for diagnostic and therapeutic applications. These models are constructed from magnetic resonance imaging (MRI) scans acquired at our university hospital. Our 3D breast modelling method is used to translate the MRI scans into 3D models representing the geometry and microwave-frequency properties of tissues in the breast. The reconstructed models demonstrate anatomical realism, reconfigurable complexity, and flexibility to adapt to simulations of various microwave imaging techniques and prototype systems. With these models, realistic and rigorous test scenarios can be defined in simulations to support feasibility analysis, performance verification and design improvements of developing microwave imaging techniques, prior to testing on experimental systems. A repository of breast models is created which includes breasts of varying classification - fatty, scattered, heterogeneous, and dense. In addition, the models include brief documentation to facilitate researchers in selecting a model by matching its features with their requirements.