HandCT: hands-on computational dataset for X-Ray Computed Tomography
HandCT: hands-on computational dataset for X-Ray Computed Tomography
HandCT is a computational dataset to train machine-learning models for X-Ray Computed Tomography (CT). It consists of a meshed hand model, of which pose and anatomical properties are computed at run-time from a script. As such, it is an accurate modeling of anatomical phantoms of only 1.35 mB, and reproducibility is ensured using random seeds. It allows the user to have full control over the imaging chain, from projection to reconstruction, and over the X-Ray interaction with the different parts of the model by a simple variable editing. This open-source solution relies on the freeware Blender for the modelling and Python for the computations. The first deals with modelling, rigging and deformations, whilst the later ensures transformations such as scaling, translation, or else forward projection. This dataset can be used to train and evaluate regularisation procedures for low-energy, dual-energy and scarce-view CT.
Valat, Emilien
8c6f8b31-e1b7-449d-a5a0-ce7ce7e472b4
Valat, Loth
a7c80819-2c14-4bc2-a697-aa7fe5a2eae4
Valat, Emilien
8c6f8b31-e1b7-449d-a5a0-ce7ce7e472b4
Valat, Loth
a7c80819-2c14-4bc2-a697-aa7fe5a2eae4
Abstract
HandCT is a computational dataset to train machine-learning models for X-Ray Computed Tomography (CT). It consists of a meshed hand model, of which pose and anatomical properties are computed at run-time from a script. As such, it is an accurate modeling of anatomical phantoms of only 1.35 mB, and reproducibility is ensured using random seeds. It allows the user to have full control over the imaging chain, from projection to reconstruction, and over the X-Ray interaction with the different parts of the model by a simple variable editing. This open-source solution relies on the freeware Blender for the modelling and Python for the computations. The first deals with modelling, rigging and deformations, whilst the later ensures transformations such as scaling, translation, or else forward projection. This dataset can be used to train and evaluate regularisation procedures for low-energy, dual-energy and scarce-view CT.
This record has no associated files available for download.
More information
Published date: 20 April 2022
Identifiers
Local EPrints ID: 473148
URI: http://eprints.soton.ac.uk/id/eprint/473148
PURE UUID: fed5890f-bd84-412e-9df9-7ec1f2a6bf2c
Catalogue record
Date deposited: 10 Jan 2023 18:48
Last modified: 19 Jul 2023 01:54
Export record
Altmetrics
Contributors
Contributor:
Loth Valat
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
