-
MAMA-MIA: A Large-Scale Multi-Center Breast Cancer DCE-MRI Benchmark Dataset with Expert Segmentations
Authors:
Lidia Garrucho,
Claire-Anne Reidel,
Kaisar Kushibar,
Smriti Joshi,
Richard Osuala,
Apostolia Tsirikoglou,
Maciej Bobowicz,
Javier del Riego,
Alessandro Catanese,
Katarzyna Gwoździewicz,
Maria-Laura Cosaka,
Pasant M. Abo-Elhoda,
Sara W. Tantawy,
Shorouq S. Sakrana,
Norhan O. Shawky-Abdelfatah,
Amr Muhammad Abdo-Salem,
Androniki Kozana,
Eugen Divjak,
Gordana Ivanac,
Katerina Nikiforaki,
Michail E. Klontzas,
Rosa García-Dosdá,
Meltem Gulsun-Akpinar,
Oğuz Lafcı,
Ritse Mann
, et al. (8 additional authors not shown)
Abstract:
Current research in breast cancer Magnetic Resonance Imaging (MRI), especially with Artificial Intelligence (AI), faces challenges due to the lack of expert segmentations. To address this, we introduce the MAMA-MIA dataset, comprising 1506 multi-center dynamic contrast-enhanced MRI cases with expert segmentations of primary tumors and non-mass enhancement areas. These cases were sourced from four…
▽ More
Current research in breast cancer Magnetic Resonance Imaging (MRI), especially with Artificial Intelligence (AI), faces challenges due to the lack of expert segmentations. To address this, we introduce the MAMA-MIA dataset, comprising 1506 multi-center dynamic contrast-enhanced MRI cases with expert segmentations of primary tumors and non-mass enhancement areas. These cases were sourced from four publicly available collections in The Cancer Imaging Archive (TCIA). Initially, we trained a deep learning model to automatically segment the cases, generating preliminary segmentations that significantly reduced expert segmentation time. Sixteen experts, averaging 9 years of experience in breast cancer, then corrected these segmentations, resulting in the final expert segmentations. Additionally, two radiologists conducted a visual inspection of the automatic segmentations to support future quality control studies. Alongside the expert segmentations, we provide 49 harmonized demographic and clinical variables and the pretrained weights of the well-known nnUNet architecture trained using the DCE-MRI full-images and expert segmentations. This dataset aims to accelerate the development and benchmarking of deep learning models and foster innovation in breast cancer diagnostics and treatment planning.
△ Less
Submitted 19 June, 2024;
originally announced June 2024.
-
CMOS pixel sensors optimized for large ionizing dynamic
Authors:
W. Ren,
J. Baudot,
L. Federici,
C. Finck,
C. Hu-Guo,
M. Kachel,
C. -A. Reidel,
C. Schui,
R. Sefri,
E. Spiriti,
U. Weber,
Y. Zhao
Abstract:
Monolithic active pixel sensors (MAPS) are now well established as a technology for tracking charged particles, especially when low material budget is desirable. For such applications, sensors focus on spatial resolution and pixels with digital output or modest charge measurement ability are well suited. Within the European Union STRONG-2020 project, which focuses on experiments using hadrons, the…
▽ More
Monolithic active pixel sensors (MAPS) are now well established as a technology for tracking charged particles, especially when low material budget is desirable. For such applications, sensors focus on spatial resolution and pixels with digital output or modest charge measurement ability are well suited. Within the European Union STRONG-2020 project, which focuses on experiments using hadrons, the TIIMM (Tracking and Ions Identifications with Minimal Material budget) joint research activity intends to expand granular MAPS capacity to energy-loss (ΔE) measurement for ion species identification. The TIIMM prototypes are developed in the Tower Jazz 180 nm CMOS image sensor (CIS) process. The Time-Over-Threshold (ToT) method is applied to the sensor for the energy-loss measurement. The main design details and the preliminary test results from laboratory measurements of the initial TIIMM prototype are presented in this work.
△ Less
Submitted 24 January, 2022;
originally announced January 2022.
-
Total nuclear reaction cross-section database for radiation protection in space and heavy-ion therapy applications
Authors:
F Luoni,
F Horst,
CA Reidel,
A Quarz,
L Bagnale,
L Sihver,
U Weber,
RB Norman,
W de Wet,
M Giraudo,
G Santin,
J Norbury,
M Durante
Abstract:
Realistic nuclear reaction cross-section models are an essential ingredient of reliable heavy-ion transport codes. Such codes are used for risk evaluation of manned space exploration missions as well as for ion-beam therapy dose calculations and treatment planning. Therefore, in this study, a collection of total nuclear reaction cross-section data has been generated within a GSI-ESA-NASA collabora…
▽ More
Realistic nuclear reaction cross-section models are an essential ingredient of reliable heavy-ion transport codes. Such codes are used for risk evaluation of manned space exploration missions as well as for ion-beam therapy dose calculations and treatment planning. Therefore, in this study, a collection of total nuclear reaction cross-section data has been generated within a GSI-ESA-NASA collaboration. The database includes the experimentally measured total nucleus-nucleus reaction cross-sections. The Tripathi, Kox, Shen, Kox-Shen, and Hybrid-Kurotama models are systematically compared with the collected data. Details about the implementation of the models are given. Literature gaps are pointed out and considerations are made about which models fit best the existing data for the most relevant systems to radiation protection in space and heavy-ion therapy.
△ Less
Submitted 6 September, 2021; v1 submitted 25 May, 2021;
originally announced May 2021.
