Lesion volume is an important predictor for prognosis in breast cancer. We make a step towards a more accurate lesion volume measurement on digital mammograms by developing a model that allows to estimate lesion volumes on processed mammograms, which are the images routinely used by radiologists in clinical practice as well as in breast cancer screening and are available in medical centers. Processed mammograms are obtained from raw mammograms, which are the X-ray data coming directly from the scanner, by applying certain vendor-specific non-linear transformations. At the core of our volume estimation method is a physics-based algorithm for measuring lesion volumes on raw mammograms. We subsequently extend this algorithm to processed mammograms via a deep learning image-to-image translation model that produces synthetic raw mammograms from processed mammograms in a multi-vendor setting. We assess the reliability and validity of our method using a dataset of 1778 mammograms with an annotated mass. Firstly, we investigate the correlations between lesion volumes computed from mediolateral oblique and craniocaudal views, with a resulting Pearson correlation of 0.93 [95% confidence interval (CI) 0.92 - 0.93]. Secondly, we compare the resulting lesion volumes from true and synthetic raw data, with a resulting Pearson correlation of 0.998 [95% CI 0.998 - 0.998] . Finally, for a subset of 100 mammograms with a malign mass and concurrent MRI examination available, we analyze the agreement between lesion volume on mammography and MRI, resulting in an intraclass correlation coefficient of 0.81 [95% CI 0.73 - 0.87] for consistency and 0.78 [95% CI 0.66 - 0.86] for absolute agreement. In conclusion, we developed an algorithm to measure mammographic lesion volume that reached excellent reliability and good validity, when using MRI as ground truth.
Data scarcity and class imbalance are two fundamental challenges in many machine learning applications to healthcare. Breast cancer classification in mammography exemplifies these challenges, with a malignancy rate of around 0.5% in a screening population, which is compounded by the relatively small size of lesions (~1% of the image) in malignant cases. Simultaneously, the prevalence of screening mammography creates a potential abundance of non-cancer exams to use for training. Altogether, these characteristics lead to overfitting on cancer cases, while under-utilizing non-cancer data. Here, w…
Deep learning approaches to breast cancer detection in mammograms have recently shown promising results. However, such models are constrained by the limited size of publicly available mammography datasets, in large part due to privacy concerns and the high cost of generating expert annotations. Limited dataset size is further exacerbated by substantial class imbalance since "normal" images dramatically outnumber those with findings. Given the rapid progress of generative models in synthesizing realistic images, and the known effectiveness of simple data augmentation techniques (e.g. horizontal…
Kheiron Medical Technologies Ltd.（Kheiron医疗技术有限公司）；Imperial College London（帝国理工学院）
The ability to generate synthetic medical images is useful for data augmentation, domain transfer, and out-of-distribution detection. However, generating realistic, high-resolution medical images is challenging, particularly for Full Field Digital Mammograms (FFDM), due to the textural heterogeneity, fine structural details and specific tissue properties. In this paper, we explore the use of progressively trained generative adversarial networks (GANs) to synthesize mammograms, overcoming the underlying instabilities when training such adversarial models. This work is the first to show that gen…
Evaluating the degree of malignancy of a massive lesion on the basis of the mere visual analysis of the mammogram is a non-trivial task. We developed a semi-automated system for massive-lesion characterization with the aim to support the radiological diagnosis. A dataset of 226 masses has been used in the present analysis. The system performances have been evaluated in terms of the area under the ROC curve, obtaining A_z=0.80+-0.04.
Mammography is widely recognized as the most reliable technique for early detection of breast cancers. Automated or semi-automated computerized classification schemes can be very useful in assisting radiologists with a second opinion about the visual diagnosis of breast lesions, thus leading to a reduction in the number of unnecessary biopsies. We present a computer-aided diagnosis (CADi) system for the characterization of massive lesions in mammograms, whose aim is to distinguish malignant from benign masses. The CADi system we realized is based on a three-stage algorithm: a) a segmentation t…
MAM-E: Mammographic synthetic image generation with diffusion models
Computer Vision and Robotics Institute (ViCOROB), University of Girona（赫罗纳大学计算机视觉与机器人研究所（ViCOROB））；Clínica Girona, Institute of Diagnostic Imaging (IDI) Girona, University of Girona（赫罗纳大学赫罗纳诊所诊断影像研究所（IDI））
A system for the automated detection of massive lesions in mammograms is presented. The approach we adopted is a pixel-based and multi-level one. Each pixel in a mammogram is flagged with the appropriate class membership, e.g. massive lesions or normal breast tissue.
MAMBO: High-Resolution Generative Approach for Mammography Images
The Institute for Artificial Intelligence Research and Development of Serbia（塞尔维亚人工智能研究与开发研究所）；University of Trento（特伦托大学）；Faculty of Technical Sciences, University of Novi Sad（诺维萨大学技术科学学院）；University of Pisa（比萨大学）；Fondazione Bruno Kessler（布鲁诺·凯斯勒基金会）；University of Verona（威尼斯大学）
Universitat Rovira i Virgili（罗维拉-威尔吉利大学）；Bioinformatics Institute, A*STAR（新加坡科技研究局生物信息学研究所）；Hospital Universitari Sant Joan de Reus（雷乌斯圣琼大学医院）
Mammogram inspection in search of breast tumors is a tough assignment that radiologists must carry out frequently. Therefore, image analysis methods are needed for the detection and delineation of breast masses, which portray crucial morphological information that will support reliable diagnosis. In this paper, we proposed a conditional Generative Adversarial Network (cGAN) devised to segment a breast mass within a region of interest (ROI) in a mammogram. The generative network learns to recognize the breast mass area and to create the binary mask that outlines the breast mass. In turn, the ad…
Computer Vision and Robotics Institute, University of Girona（赫罗纳大学计算机视觉与机器人研究所）；University of Barcelona（巴塞罗那大学）；Faculty of Medicine, University of Girona（赫罗纳大学医学院）；Parc Tauli Hospital Universitari（陶利公园大学医院）；Institut d’Investigacio Parc Tauli I3PT（陶利公园研究所）
Medical image synthesis has gained a great focus recently, especially after the introduction of Generative Adversarial Networks (GANs). GANs have been used widely to provide anatomically-plausible and diverse samples for augmentation and other applications, including segmentation and super resolution. In our previous work, Deep Convolutional GANs were used to generate synthetic mammogram lesions, masses mainly, that could enhance the classification performance in imbalanced datasets. In this new work, a deeper investigation was carried out to explore other aspects of the generated images evalu…
Medical University of Gdansk（格但斯克医科大学）；Karolinska University Hospital（卡罗林斯卡大学医院）；Karolinska Institutet（卡罗林斯卡学院）
Computer-aided detection systems based on deep learning have shown good performance in breast cancer detection. However, high-density breasts show poorer detection performance since dense tissues can mask or even simulate masses. Therefore, the sensitivity of mammography for breast cancer detection can be reduced by more than 20% in dense breasts. Additionally, extremely dense cases reported an increased risk of cancer compared to low-density breasts. This study aims to improve the mass detection performance in highdensity breasts using synthetic high-density full-field digital mammograms (FFD…
A computer-aided detection (CADe) system for the identification of microcalcification clusters in digital mammograms has been developed. It is mainly based on the application of wavelet transforms for image filtering and neural networks for both the feature extraction and the classification procedures. This CADe system is easily adaptable to different databases. We report and compare the FROC curves obtained on the private database we used for developing the CADe system and on the publicly available MIAS database. The results achieved on the two databases show the same trend, thus demonstratin…
Shanghai United Imaging Intelligence Co., Ltd.（上海联影智能医疗科技有限公司）；Institute for Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University（上海交通大学生物医学工程学院医学影像技术研究所）
Computer-aided diagnosis with deep learning techniques has been shown to be helpful for the diagnosis of the mammography in many clinical studies. However, the image styles of different vendors are very distinctive, and there may exist domain gap among different vendors that could potentially compromise the universal applicability of one deep learning model. In this study, we explicitly address style variety issue with the proposed multi-resolution and multi-reference neural style transfer (mr2NST) network. The mr2NST can normalize the styles from different vendors to the same style baseline w…
The CALMA (Computer Assisted Library for MAmmography) project is a five years plan developed in a physics research frame in collaboration between INFN (Istituto Nazionale di Fisica Nucleare) and many Italian hospitals. At present a large database of digitized mammographic images (more than 6000) was collected and a software based on neural network algorithms for the search of suspicious breast lesions was developed. Two tools are available: a microcalcification clusters hunter, based on supervised and unsupervised feedforward neural network, and a massive lesions searcher, based on a hibrid ap…
Paired Image Generation with Diffusion-Guided Diffusion Models
University of Science and Technology of China（科学技术大学）；Suzhou Institute of Biomedical Engineering and Technology（生物医学工程与技术研究所）；Macao Polytechnic University（澳门 polytechnic 大学）；Suzhou Municipal Hospital（苏州 municipal 医院）；Guizhou Provincial People's Hospital（贵州省人民医院）
A breast neoplasia is often marked by the presence of microcalcifications and massive lesions in the mammogram: hence the need for tools able to recognize such lesions at an early stage. Our collaboration, among italian physicists and radiologists, has built a large distributed database of digitized mammographic images and has developed a Computer Aided Detection (CADe) system for the automatic analysis of mammographic images and installed it in some Italian hospitals by a GRID connection. Regarding microcalcifications, in our CADe digital mammogram is divided into wide windows which are proce…
A computer-aided detection (CADe) system for microcalcification cluster identification in mammograms has been developed in the framework of the EU-founded MammoGrid project. The CADe software is mainly based on wavelet transforms and artificial neural networks. It is able to identify microcalcifications in different kinds of mammograms (i.e. acquired with different machines and settings, digitized with different pitch and bit depth or direct digital ones). The CADe can be remotely run from GRID-connected acquisition and annotation stations, supporting clinicians from geographically distant loc…
Fundus2Globe: Generative AI-Driven 3D Digital Twins for Personalized Myopia Management
School of Optometry, The Hong Kong Polytechnic University（香港理工大学验光学院）；Swiss Federal Institute of Technology Lausanne (EPFL)（瑞士洛桑联邦理工学院）；Centre for Eye and Vision Research (CEVR)（眼与视觉研究中心）
The use of an automatic system for the analysis of mammographic images has proven to be very useful to radiologists in the investigation of breast cancer, especially in the framework of mammographic-screening programs. A breast neoplasia is often marked by the presence of microcalcification clusters and massive lesions in the mammogram: hence the need for tools able to recognize such lesions at an early stage. In the framework of the GPCALMA (GRID Platform for Computer Assisted Library for MAmmography) project, the co-working of italian physicists and radiologists built a large distributed dat…
Changes in Visual Attention Patterns for Detection Tasks due to Dependencies on Signal and Background Spatial Frequencies
Department of Biomedical Engineering, University of Houston（生物医学工程系，休斯顿大学）；Department of Physics, University of Houston（物理系，休斯顿大学）；Department of Electrical and Computer Engineering, University of Houston（电气与计算机工程系，休斯顿大学）
Cervical Cancer is one of the most common forms of cancer in women worldwide. Most cases of cervical cancer can be prevented through screening programs aimed at detecting precancerous lesions. During Digital Colposcopy, Specular Reflections (SR) appear as bright spots heavily saturated with white light. These occur due to the presence of moisture on the uneven cervix surface, which act like mirrors reflecting light from the illumination source. Apart from camouflaging the actual features, the SR also affects subsequent segmentation routines and hence must be removed. Our novel technique elimin…
Digital Breast Tomosynthesis (DBT) provides an insight into the fine details of normal fibroglandular tissues and abnormal lesions by reconstructing a pseudo-3D image of the breast. In this respect, DBT overcomes a major limitation of conventional X-ray mammography by reducing the confounding effects caused by the superposition of breast tissue. In a breast cancer screening or diagnostic context, a radiologist is interested in detecting change, which might be indicative of malignant disease. To help automate this task image registration is required to establish spatial correspondence between t…
Simulation-based image quality metrics are adapted and investigated for characterizing the parameter dependences of linear iterative image reconstruction for DBT. Three metrics based on 2D DBT simulation are investigated: (1) a root-mean-square-error (RMSE) between the test phantom and reconstructed image, (2) a gradient RMSE where the comparison is made after taking a spatial gradient of both image and phantom, and (3) a region-of-interest (ROI) Hotelling observer (HO) for signal-known-exactly/background-known-exactly (SKE/BKE) and signal-known-exactly/background-known-statistically (SKE/BKS)…
