U.S. flag An official website of the United States government

U.S. Food and Drug Administration

A Normative Dataset of Healthy Pediatric Cranial Computed Tomography (CT) Images

Catalog of Regulatory Science Tools to Help Assess New Medical Devices 

This regulatory science tool is a dataset of 100 healthy pediatric cranial computed tomography (CT) images with associated demographic data. It is intended to serve as a normative reference for developing, validating, and assessing the performance of medical devices, including software-based analysis tools such as artificial intelligence and machine learning (AI/ML) applications., designed for pediatric neuroimaging applications.

Technical Description

This tool consists of a curated dataset of medical images and associated demographic data. The dataset contains 100 cranial CT image series all from pediatric subjects who met inclusion criteria (no known neurological pathology or significant comorbidities) and spans a developmental age range from one month to ten years old. Each image is provided in the MetaImage (.mha) format and a corresponding spreadsheet (.xlsx) contains demographic data for each subject, including age in days, biological sex, and race/ethnicity.

Image files are named with a random numerical identifier that serves as a key to link to the demographic data in the spreadsheet. The tool is accompanied by an optional Python script to promote repeatable data loading and handling.

Intended Purpose 

The intended purpose of this dataset is to provide a normative reference for the development, verification, and evaluation of medical devices used in pediatric cranial imaging. The intended users of this tool include researchers and developers involved in pediatric cranial imaging.. This tool could be applicable to medical devices that acquire or analyze pediatric cranial CT images and could provide a baseline for evaluating devices that measure anatomical structures in pediatric subjects. Potential applications include AI/ML-based diagnostic software, image analysis tools, and image quality algorithms for pediatric neuroimaging. It is particularly relevant for devices intended to identify anatomical abnormalities, assess developmental stages, or segment structures in the pediatric brain.

Testing

The integrity and quality of the tool were verified through a multi-step process to ensure it is fit for its intended purpose. All images included in the dataset underwent institutional quality assurance procedures and were confirmed to be of diagnostic quality when originally obtained. Individuals with known neurological pathologies or significant comorbidities were excluded. A de-identification protocol was applied to all images and demographic data to protect subject confidentiality. The effectiveness of this process was verified before data release by individually evaluating each image. The Python script for data loading was tested to ensure it functions correctly, to support consistent data loading and facilitate reproducibility of analyses.

Limitations

The tool consists of 100 subjects from a single source, which may limit generalizability to other institutions or geographic regions. which may introduce institutional or regional biases. The dataset is cross-sectional and does not contain longitudinal data for tracking individual development over time. This is a normative dataset and does not include pathological cases.

Supporting Documentation

User Manual with links to images, excel sheet with demographic information, and Python script

Users should cite the following acknowledgement:

Data were collected with the support of NIH NICHD awards R42HD081712 and made available by Children's National Hospital to FDA, and released by FDA as a Regulatory Science Tool [U.S. Food and Drug Administration. (2026). A Normative Dataset of Healthy Pediatric Cranial Computed Tomography (CT) Images (RST26PD01.01). https://cdrh-rst.fda.gov/normative-dataset-healthy-pediatric-cranial-computed-tomography-ct-images].

Users should cite the following papers:

Porras AR, Paniagua B, Ensel S, Keating R, Rogers GF, Enquobahrie A, Linguraru MG. Locally Affine Diffeomorphic Surface Registration and Its Application to Surgical Planning of Fronto-Orbital Advancement. IEEE Trans Med Imaging. 2018 Jul;37(7):1690-1700. doi: 10.1109/TMI.2018.2816402. PMID: 29969419; PMCID: PMC6085886.

Porras AR, Tu L, Tsering D, Mantilla E, Oh A, Enquobahrie A, Keating R, Rogers GF, Linguraru MG. Quantification of Head Shape from Three-Dimensional Photography for Presurgical and Postsurgical Evaluation of Craniosynostosis. Plast Reconstr Surg. 2019 Dec;144(6):1051e-1060e. doi: 10.1097/PRS.0000000000006260. PMID: 31764657; PMCID: PMC6905129.

Contact

Tool Reference