A5080238943- Digital Radiography and Breast Imaging
- AI in cancer detection
- Advanced X-ray and CT Imaging
- Medical Imaging Techniques and Applications
- Radiation Dose and Imaging
- Radiomics and Machine Learning in Medical Imaging
- Global Cancer Incidence and Screening
- Advanced Image Fusion Techniques
- Infrared Thermography in Medicine
Hospital de Câncer de Barretos
2019-2024
Accurate risk assessment is essential for the success of population screening programs in breast cancer. Models with high sensitivity and specificity would enable to target more elaborate efforts high-risk populations, while minimizing overtreatment rest. Artificial intelligence (AI)-based models have demonstrated a significant advance over used today clinical practice. However, responsible deployment novel AI requires careful validation across diverse populations. To this end, we validate...
Purpose To investigate the use of an affine‐variance noise model, with correlated quantum and spatially dependent gain, for simulation in virtual clinical trials (VCT) digital breast tomosynthesis (DBT). Methods Two distinct technologies were considered: amorphous‐selenium (a‐Se) detector direct conversion a thallium‐doped cesium iodide (CsI(Tl)) indirect conversion. A VCT framework was used to generate noise‐free projections uniform three‐dimensional simulated phantom, whose geometry...
Noise negatively impacts the detection and characterization of lesions in mammography. While denoising filters may be used to suppress noise, they might also affect conspicuity small due signal blurring smearing. In previous works, we designed validated a pipeline, dedicated mammography, capable suppressing noise avoiding excessive blur smear. This is achieved by fine-tuned noisy-denoised image blending step, which leverages Poisson-Gaussian model. current work, investigate impact pipeline...
Several clinical image databases are currently available to support scientific research in the medical field. These images generally used validate studies based on measuring sensitivity and specificity of a particular task. In case digital mammography, radiation dose directly influences quality consequently performance radiologists. Therefore, it is important conduct find balance between dose. Image processing methods typically employed optimize this relationship. For evaluation these...
The validation of many dose optimization methods in x-ray imaging requires clinical images from a range signal-to-ratio regimes. This data is commonly generated through computer simulation. For this purpose, our group developed method to simulate reduction for digital breast tomosynthesis. In the previous work, tests were performed system that features an amorphous selenium detector with minimal pixel correlation. current we evaluate simulation performance silicon system, which yields...
In digital mammography, the physics of acquisition system and post-processing algorithms can cause image noise to be spatially correlated. Noise correlation is characterized by non-constant power spectral density negatively affect quality. Although literature explores ways quantify frequency dependence in there still a lack studies that explore effect this phenomenon on clinical tasks. Thus, aim work evaluate impact quality mammography detectability lesions using virtual trial (VCT) tool....
Digital Breast Tomosynthesis (DBT) is a medical imaging modality that has been increasingly used for breast cancer screening. To improve the accuracy in early detection of cancer, it common to use tools based on image processing quality DBT images and, consequently, visibility lesions clinical interest. Microcalcification (MC) clusters are class findings may indicate stages cancer. evaluate impact methods lesions, human perception studies usually performed, where evaluated using positive and...
In this work, we investigated and measured the noise in Digital Breast Tomosynthesis (DBT) slices considering back-projection (BP) algorithm for image reconstruction. First, presented our open-source DBT reconstruction toolbox validated with a freely available virtual clinical trials (VCT) software, comparing results at Food Drug Administration's (FDA) repository. A anthropomorphic breast phantom was generated VCT environment noise-free projections were simulated. Slices reconstructed by...
It is well-known that x-ray systems featuring indirect detectors are affected by noise spatial correlation. In the case of digital breast tomosynthesis (DBT), this phenomenon might affect perception small details in image, such as microcalcifications. work, we propose use a deep convolutional neural network (CNN) to restore DBT projections degraded with correlated using framework cycle generative adversarial (cycle-GAN). To generate pairs images for training procedure, used virtual clinical...
The majority of the denoising algorithms available in literature are designed to treat signal-independent Gaussian noise. However, digital breast tomosynthesis (DBT) systems, noise model seldom presents signal-independence. In this scenario, variance-stabilizing transforms (VSTs) may be used convert signaldependent into approximately noise, enabling use 'off-the-shelf' techniques. accurate stabilization variance requires a robust estimation system's coefficients, usually obtained using...