Dr. Oliver Diaz Profile

Dr. Oliver Diaz

Assistant Professor at Univ. de Barcelona

SPIE Involvement:

Author

Area of Expertise:

Mammography , Digital Breast Tomosynthesis , Monte Carlo simulations , Breast CT , Breast MRI , Breast Ultrasound

Publications (27)

Proceedings Article | 2 April 2024 Presentation + Paper

Leveraging epistemic uncertainty to improve tumour segmentation in breast MRI: an exploratory analysis

Smriti Joshi, Richard Osuala, Lidia Garrucho, Apostolia Tsirikoglou, Javier del Riego, Katarzyna Gwoździewicz, Kaisar Kushibar, Oliver Diaz, Karim Lekadir

Proceedings Volume 12926, 1292616 (2024) https://doi.org/10.1117/12.3006783

KEYWORDS: Tumors, Image segmentation, Magnetic resonance imaging, Breast, Cancer, Uncertainty analysis, Machine learning

Read Abstract +

Proceedings Article | 2 April 2024 Paper

Pre- to post-contrast breast MRI synthesis for enhanced tumour segmentation

Richard Osuala, Smriti Joshi, Apostolia Tsirikoglou, Lidia Garrucho Moras, Walter L. Pinaya, Oliver Diaz, Karim Lekadir

Proceedings Volume 12926, 129260Y (2024) https://doi.org/10.1117/12.3006961

KEYWORDS: Deep learning, Adversarial training, Realistic image synthesis, Image contrast enhancement, Artificial intelligence, Image segmentation, Magnetic resonance imaging, Breast, Contrast agents, 3D image processing

Read Abstract +

Proceedings Article | 7 April 2023 Presentation + Paper

End-to-end mammographic breast density quantification with deep learning: preliminary study on simulated mammograms

Sjoerd A. Tunissen, Andrea Motta, Franziska Mauter, Eloy García, Oliver Diaz, John Boone, Ioannis Sechopoulos, Marco Caballo

Proceedings Volume 12465, 124650K (2023) https://doi.org/10.1117/12.2654131

KEYWORDS: Mammography, Breast density, Breast, Deep learning, Convolution, Image segmentation, Adipose tissue, 3D modeling

Read Abstract +

SPIE Journal Paper | 20 February 2023 Open Access

medigan: a Python library of pretrained generative models for medical image synthesis

Richard Osuala, Grzegorz Skorupko, Noussair Lazrak, Lidia Garrucho, Eloy García, Smriti Joshi, Socayna Jouide, Michael Rutherford, Fred Prior, Kaisar Kushibar, Oliver Díaz, Karim Lekadir

JMI, Vol. 10, Issue 06, 061403, (February 2023) https://doi.org/10.1117/12.10.1117/1.JMI.10.6.061403

KEYWORDS: Data modeling, Education and training, Medical imaging, Process modeling, Statistical modeling, Mammography, Visual process modeling, Performance modeling, Systems modeling, Gallium nitride

Read Abstract +

PurposeDeep learning has shown great promise as the backbone of clinical decision support systems. Synthetic data generated by generative models can enhance the performance and capabilities of data-hungry deep learning models. However, there is (1) limited availability of (synthetic) datasets and (2) generative models are complex to train, which hinders their adoption in research and clinical applications. To reduce this entry barrier, we explore generative model sharing to allow more researchers to access, generate, and benefit from synthetic data.ApproachWe propose medigan, a one-stop shop for pretrained generative models implemented as an open-source framework-agnostic Python library. After gathering end-user requirements, design decisions based on usability, technical feasibility, and scalability are formulated. Subsequently, we implement medigan based on modular components for generative model (i) execution, (ii) visualization, (iii) search & ranking, and (iv) contribution. We integrate pretrained models with applications across modalities such as mammography, endoscopy, x-ray, and MRI.ResultsThe scalability and design of the library are demonstrated by its growing number of integrated and readily-usable pretrained generative models, which include 21 models utilizing nine different generative adversarial network architectures trained on 11 different datasets. We further analyze three medigan applications, which include (a) enabling community-wide sharing of restricted data, (b) investigating generative model evaluation metrics, and (c) improving clinical downstream tasks. In (b), we extract Fréchet inception distances (FID) demonstrating FID variability based on image normalization and radiology-specific feature extractors.Conclusionmedigan allows researchers and developers to create, increase, and domain-adapt their training data in just a few lines of code. Capable of enriching and accelerating the development of clinical machine learning models, we show medigan’s viability as platform for generative model sharing. Our multimodel synthetic data experiments uncover standards for assessing and reporting metrics, such as FID, in image synthesis studies.

DOWNLOAD PAPER

Proceedings Article | 13 July 2022 Paper

Simultaneous pectoral muscle and nipple location in MLO mammograms, considering image quality assumptions

E. García, R. Martí, J. Martí, J. del Riego, C. Aynes, A. Oliver, O. Diaz

Proceedings Volume 12286, 122860E (2022) https://doi.org/10.1117/12.2625778

KEYWORDS: Nipple, Breast, Image segmentation, Mammography, Skin, Image quality, Image processing algorithms and systems, Algorithm development, Tissues, Lung cancer

Read Abstract +

Proceedings Article | 13 July 2022 Paper

Sharing generative models instead of private data: a simulation study on mammography patch classification

Zuzanna Szafranowska, Richard Osuala, Bennet Breier, Kaisar Kushibar, Karim Lekadir, Oliver Diaz

Proceedings Volume 12286, 122860Q (2022) https://doi.org/10.1117/12.2625781

KEYWORDS: Data modeling, Mammography, Transformers, Breast cancer, Breast, Classification systems, Convolutional neural networks, Realistic image synthesis, Data acquisition, Digital mammography, Image classification

Read Abstract +

Proceedings Article | 22 May 2020 Paper

First approach to estimate breast radiation dose in a DBT prone biopsy table

Lucas Paixão, Oliver Diaz, Margarita Chevalier

Proceedings Volume 11513, 115130H (2020) https://doi.org/10.1117/12.2564320

KEYWORDS: Breast, Biopsy, Digital breast tomosynthesis, Monte Carlo methods, Biological research, X-rays, Sensors, Target acquisition, X-ray detectors, Dielectrophoresis

Read Abstract +

Proceedings Article | 22 May 2020 Paper

Realistic compressed breast phantoms for medical physics applications

E. García, C. Fedon, M. Caballo, R. Martí, I. Sechopoulos, O. Diaz

Proceedings Volume 11513, 1151304 (2020) https://doi.org/10.1117/12.2564273

KEYWORDS: Breast, Tissues, Image compression, Digital breast tomosynthesis, 3D modeling, Medical physics, Image segmentation, X-rays, Mammography, Skin

Read Abstract +

Proceedings Article | 22 May 2020 Paper

Monte Carlo dose evaluation of different fibroglandular tissue distribution in breast imaging

L. M. Arana Peña, C. Fedon, E. García, O. Diaz, R. Longo, D. Dance, I. Sechopoulos

Proceedings Volume 11513, 115130G (2020) https://doi.org/10.1117/12.2564278

KEYWORDS: Breast, Tissues, Monte Carlo methods, Image compression, Mammography, X-rays, Breast imaging, 3D modeling, Image segmentation

Read Abstract +

Proceedings Article | 22 May 2020 Paper

Quality analysis of DCGAN-generated mammography lesions

Basel Alyafi, Oliver Diaz, Premkumar Elangovan, Joan Vilanova, Javier del Riego, Robert Marti

Proceedings Volume 11513, 115130B (2020) https://doi.org/10.1117/12.2560473

KEYWORDS: Mammography, Image processing, Tissues, Breast, Data modeling, Visualization, Databases, Medical imaging

Read Abstract +

Proceedings Article | 22 May 2020 Paper

Evaluation of elastic parameters for breast compression using a MRI-mammography registration approach

E. García, Y. Diez, A. Oliver, N. Karssemeijer, J. Martí, R. Martí, O. Diaz

Proceedings Volume 11513, 1151306 (2020) https://doi.org/10.1117/12.2564155

KEYWORDS: Tissues, Breast, Magnetic resonance imaging, Magnesium, Mammography, Image processing, Image registration, Image compression, Image segmentation, X-ray imaging

Read Abstract +

Proceedings Article | 16 March 2020 Paper

DCGANs for realistic breast mass augmentation in x-ray mammography

Basel Alyafi, Oliver Diaz, Robert Marti

Proceedings Volume 11314, 1131420 (2020) https://doi.org/10.1117/12.2543506

KEYWORDS: Breast, Mammography, Tissues, Breast cancer, X-rays, Computer aided diagnosis and therapy, Cancer, Image processing, Databases

Read Abstract +

Proceedings Article | 1 March 2019 Paper

Validation of modelling tools for simulating wide-angle DBT systems

Premkumar Elangovan, Alistair Mackenzie, David Dance, Kevin Wells, Oliver Diaz, Lucy Warren, Kenneth Young

Proceedings Volume 10948, 109482E (2019) https://doi.org/10.1117/12.2512225

KEYWORDS: Digital breast tomosynthesis, Breast, Polymethylmethacrylate, Monte Carlo methods, Modulation transfer functions, Aluminum, Modeling, Clinical trials, Digital mammography, Computer simulations

Read Abstract +

Proceedings Article | 1 March 2019 Paper

Can breast models be simplified to estimate scattered radiation in breast tomosythesis?

Oliver Diaz, Prekumar Elangovan, David Dance, Kevin Wells, Kenneth Young

Proceedings Volume 10948, 109485J (2019) https://doi.org/10.1117/12.2512287

KEYWORDS: Breast, Digital breast tomosynthesis, Monte Carlo methods, Receptors, Photons, Image compression, Signal detection, Sensors, Tissues, Digital imaging

Read Abstract +

SPIE Journal Paper | 20 February 2019 Open Access

Automatic mass detection in mammograms using deep convolutional neural networks

Richa Agarwal, Oliver Diaz, Xavier Lladó, Moi Hoon Yap, Robert Martí

JMI, Vol. 6, Issue 03, 031409, (February 2019) https://doi.org/10.1117/12.10.1117/1.JMI.6.3.031409

KEYWORDS: Mammography, Databases, Breast, Data modeling, Convolutional neural networks, Image classification, Image segmentation, Digital imaging, Image analysis, CAD systems

Read Abstract +

DOWNLOAD PAPER