Overview

Schema Version

https://atlas.rsna.org/schemas/2025-11/dataset.json

Name

FASTR-SCANN cardiac MRI T1 mapping dataset (internal data with external test reference)

Link

https://pubs.rsna.org/doi/10.1148/ryai.210294

Indexing

Keywords: Cardiac MRI, T1 mapping, Extracellular volume, MOLLI, STONE, Segmentation, Deep learning, U-Net, Relaxometry, Myocardium

Content: CA, MR

RadLex: RID35976, RID10312, RID10794, RID1398, RID10728, RID49531

SNOMED: 399020009, 233873004, 414545008

Author(s)

Nitish Bhatt

Venkat Ramanan

Ady Orbach

Labonny Biswas

Matthew Ng

Fumin Guo

Xiuling Qi

Lancia Guo

Laura Jimenez-Juan

Idan Roifman

Graham A. Wright

Nilesh R. Ghugre

Organization(s)

University of Toronto

Sunnybrook Research Institute

Sunnybrook Health Sciences Centre

St Michael’s Hospital

Contact

Nitish Bhatt (corresponding author)

Funding

Ontario Research Fund (ORF-RE7–21); Natural Sciences and Engineering Research Council (NSERC) Discovery Program (RGPIN-2019–06367); N.R.G. supported by Heart and Stroke Foundation of Canada National New Investigator award.

Ethical review

Research Ethics Board of Sunnybrook Health Sciences Centre approved the study; all patients provided written informed consent.

Comments

Retrospective internal dataset (Sunnybrook Health Sciences Centre) of cardiac MRI T1 mapping used to train and test FASTR-SCANN; evaluated also on an independent, publicly available external dataset for generalization.

Date

Published: 2022-11-02

References

[1] Bhatt N, Ramanan V, Orbach A, Biswas L, Ng M, Guo F, Qi X, Guo L, Jimenez-Juan L, Roifman I, Wright GA, Ghugre NR. "A Deep Learning Segmentation Pipeline for Cardiac T1 Mapping Using MRI Relaxation–based Synthetic Contrast Augmentation". Radiology: Artificial Intelligence. 2022-11-01. doi:10.1148/ryai.210294. PMID: 36523641. PMCID: PMC9745444.

[2] Fahmy A. "Replication data for: automated analysis of cardiovascular magnetic resonance myocardial native T1 mapping images using fully convolutional neural networks". Harvard Dataverse. 2021-01-01. Available from: https://dataverse.harvard.edu/

Dataset

Motivation

To enable accurate automated myocardial segmentation and T1/ECV analysis across native and postcontrast T1 maps using synthetic contrast augmentation.

Sampling

100 consecutive patient examinations between 2016 and 2019.

Partitioning scheme

Internal data split into training (60 patients; 358 images) and internal test (40 patients; 240 images). External test included 147 patients; 735 images (native T1 only).

Missing information

Image file formats, spatial resolution, and full demographic breakdown not reported; data-sharing availability for internal dataset not specified.

Relationships between instances

Per patient: native and postcontrast T1 series; per slice T1 maps reconstructed from multiple inversion times; per-segment analysis based on AHA 17-segment model.

Noise

Potential motion across inversion times addressed via deformable registration; failure cases included thin myocardium, partial volume blurring, and dilated myocardium.

External data

An independent, publicly available external dataset (147 patients) was used for external testing (native T1 only).

Confidentiality

Retrospective patient MRI data collected under institutional ethics approval with informed consent.

Sensitive data

Clinical cardiac MRI with associated hematocrit levels for ECV computation.