CheXmask: a large-scale dataset of anatomical segmentation masks for multi-center chest x-ray images DOI Creative Commons
Nicolás Gaggion, Candelaria Mosquera,

Lucas Mansilla

et al.

arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

The development of successful artificial intelligence models for chest X-ray analysis relies on large, diverse datasets with high-quality annotations. While several databases images have been released, most include disease diagnosis labels but lack detailed pixel-level anatomical segmentation labels. To address this gap, we introduce an extensive multi-center dataset uniform and fine-grain annotations coming from six well-known publicly available databases: CANDID-PTX, ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest, VinDr-CXR, resulting in 676,803 masks. Our methodology utilizes the HybridGNet model to ensure consistent segmentations across all datasets. Rigorous validation, including expert physician evaluation automatic quality control, was conducted validate Additionally, provide individualized indices per mask overall estimation dataset. This serves as a valuable resource broader scientific community, streamlining assessment innovative methodologies analysis. CheXmask is at: https://physionet.org/content/chexmask-cxr-segmentation-data/

Language: Английский

Complex Organ Mask Guided Radiology Report Generation DOI

Tiancheng Gu,

Dongnan Liu, Zhiyuan Li

et al.

2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Journal Year: 2024, Volume and Issue: unknown, P. 7980 - 7989

Published: Jan. 3, 2024

The goal of automatic report generation is to generate a clinically accurate and coherent phrase from single given X-ray image, which could alleviate the workload traditional radiology reporting. However, in real-world scenario, radiologists frequently face challenge producing extensive reports derived numerous medical images, thereby multi-image perspective needed. In this paper, we propose Complex Organ Mask Guided (termed as COMG) model, incorporates masks multiple organs (e.g., bones, lungs, heart, mediastinum), pro-vide more detailed information guide model's attention these crucial body regions. Specifically, leverage prior knowledge disease corresponding each organ fusion process enhance identification phase during process. Additionally, cosine similarity loss introduced target function ensure convergence cross-modal consistency facilitate model optimization. Experimental results on two public datasets show that COMG achieves 11.4% 9.7% improvement terms BLEU@4 scores over SOTA KiUT IU-Xray MIMIC, respectively. code publicly available at https://github.com/GaryGuTC/COMG_model.

Language: Английский

Citations

11
CheXmask: a large-scale dataset of anatomical segmentation masks for multi-center chest x-ray images DOI Creative Commons
Nicolás Gaggion, Candelaria Mosquera,

Lucas Mansilla

et al.

Scientific Data, Journal Year: 2024, Volume and Issue: 11(1)

Published: May 17, 2024

Language: Английский

Citations

10
Towards Unifying Anatomy Segmentation: Automated Generation of a Full-Body CT Dataset DOI
Alexander Jaus, Constantin Seibold,

Kelsey Hermann

et al.

2022 IEEE International Conference on Image Processing (ICIP), Journal Year: 2024, Volume and Issue: unknown, P. 41 - 47

Published: Sept. 27, 2024

Language: Английский

Citations

5
LLM-Driven Chest X-Ray Report Generation With a Modular, Reduced-Size Architecture DOI

Talles Viana Vargas,

Hélio Pedrini, André Santanchè

et al.

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 199 - 211

Published: Jan. 1, 2025

Language: Английский

Citations

0
Explainable variable-weight multi-modal based deep learning framework for catheter malposition detection DOI Creative Commons
Yuhan Wang, Hak‐Keung Lam

Information Fusion, Journal Year: 2025, Volume and Issue: unknown, P. 103170 - 103170

Published: April 1, 2025

Language: Английский

Citations

0
Multi-label pathology editing of chest X-rays with a Controlled Diffusion Model DOI

Huan Chu,

Xiaolong Qi, Huiling Wang

et al.

Medical Image Analysis, Journal Year: 2025, Volume and Issue: 103, P. 103584 - 103584

Published: April 20, 2025

Language: Английский

Citations

0
From Posts to Knowledge: Annotating a Pandemic-Era Reddit Dataset to Navigate Mental Health Narratives DOI Creative Commons

Saima Rani,

Khandakar Ahmed, Sudha Subramani

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(4), P. 1547 - 1547

Published: Feb. 15, 2024

Mental illness is increasingly recognized as a substantial public health challenge worldwide. With the advent of social media, these platforms have become pivotal for individuals to express their emotions, thoughts, and experiences, thereby serving rich resource mental research. This paper devoted creation comprehensive dataset an innovative data annotation methodology explore underlying causes issues. Our approach included extraction over one million Reddit posts from five different subreddits, spanning pre-pandemic, during-pandemic, post-pandemic periods. These were methodically annotated using set specific criteria, aimed at identifying various root causes. rigorous process produced richly categorized dataset, invaluable detailed analysis. The complete unlabelled along with subset that has been expertly annotated, prepared release, outlined in availability section. critical training fine-tuning machine learning models identify foundational triggers individual issues, offering valuable insights practical interventions future research this domain.

Language: Английский

Citations

2
Anatomy Completor: A Multi-class Completion Framework for 3D Anatomy Reconstruction DOI
Jianning Li, Antonio Pepe, Gijs Luijten

et al.

Lecture notes in computer science, Journal Year: 2023, Volume and Issue: unknown, P. 1 - 14

Published: Jan. 1, 2023

Language: Английский

Citations

3
Anatomy-Guided Pathology Segmentation DOI
Alexander Jaus, Constantin Seibold, Simon Reiß

et al.

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 3 - 13

Published: Jan. 1, 2024

Language: Английский

Citations

0
CheXmask: a large-scale dataset of anatomical segmentation masks for multi-center chest x-ray images DOI Creative Commons
Nicolás Gaggion, Candelaria Mosquera,

Lucas Mansilla

et al.

arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

The development of successful artificial intelligence models for chest X-ray analysis relies on large, diverse datasets with high-quality annotations. While several databases images have been released, most include disease diagnosis labels but lack detailed pixel-level anatomical segmentation labels. To address this gap, we introduce an extensive multi-center dataset uniform and fine-grain annotations coming from six well-known publicly available databases: CANDID-PTX, ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest, VinDr-CXR, resulting in 676,803 masks. Our methodology utilizes the HybridGNet model to ensure consistent segmentations across all datasets. Rigorous validation, including expert physician evaluation automatic quality control, was conducted validate Additionally, provide individualized indices per mask overall estimation dataset. This serves as a valuable resource broader scientific community, streamlining assessment innovative methodologies analysis. CheXmask is at: https://physionet.org/content/chexmask-cxr-segmentation-data/

Language: Английский

Citations

0