Overview

Schema Version

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

Name

MRI-based Prostate Radiation Therapy

Link

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8980936/

Indexing

Keywords: MRI, Neural Networks, Radiation Therapy, Prostate, Segmentation, Dosimetry, Low-dose-rate prostate brachytherapy, External urinary sphincter, Hydrogel rectal spacer

Content: GU, MR, RO

RadLex: RID39262, RID49531, RID22338, RID49590, RID12781, RID38870, RID16573, RID22286

Author(s)

Jeremiah W. Sanders

Rajat J. Kudchadker

Chad Tang

Henry Mok

Aradhana M. Venkatesan

Howard D. Thames

Steven J. Frank

Organization(s)

The University of Texas MD Anderson Cancer Center

License

Text: © RSNA, 2022 (article).

Contact

Corresponding author: Jeremiah W. Sanders (Health Insurance Portability and Accountability Act–compliant and institutional review board–approved study at The University of Texas MD Anderson Cancer Center)

Ethical review

Health Insurance Portability and Accountability Act–compliant and institutional review board–approved protocol at the University of Texas MD Anderson Cancer Center.

Date

Published: 2022-01-26

References

[1] Sanders JW, Kudchadker RJ, Tang C, Mok H, Venkatesan AM, Thames HD, Frank SJ. "Prospective Evaluation of Prostate and Organs at Risk Segmentation Software for MRI-based Prostate Radiation Therapy". Radiology: Artificial Intelligence. 2022-03-01. doi:10.1148/ryai.210151. PMID: 35391775. PMCID: PMC8980936.

Dataset

Motivation

Assess clinical performance of fully convolutional network–based automatic segmentation for prostate and organs at risk on postimplant MRI for LDRPBT using clinically relevant dosimetry metrics.

Sampling

Thirty consecutive patients with confirmed prostate cancer undergoing MRI-based LDRPBT between February and June 2021 at a single institution.

Partitioning scheme

No machine learning train/validation/test partitioning within this study; comparisons made between automatic and physician-refined contours and between patient-specific and fixed operating points.

Missing information

No public data repository or file formats specified; imaging resolution and file formats not reported.

Relationships between instances

For each patient, multiple organ segmentations (prostate, external urinary sphincter, seminal vesicles, rectum, bladder) were derived from a postimplant MRI examination; dosimetry computed from the same seed plan for automatic and physician-refined contours.

Noise

Presence of implanted radioactive seeds and, in 23% of patients, biodegradable hydrogel rectal spacers; images acquired with different MRI sequences (turbo spin-echo or fully balanced SSFP).

External data

Development of the DL model referenced 295 MRI scans; the present prospective evaluation used 30 consecutive patients imaged postimplant at the same institution.

Confidentiality

Patient MRI examinations from a single institution; HIPAA-compliant, IRB-approved.

Sensitive data

Clinical imaging of cancer patients; PHI handling governed by HIPAA and IRB approval.