\"In 3D computer graphics, a voxel represents a value on a regular grid in three-dimensional space. As with pixels in a 2D bitmap, voxels themselves do not typically have their position (i.e. coordinates) explicitly encoded with their values. Instead, rendering systems infer the position of a voxel based upon its position relative to other voxels (i.e., its position in the data structure that makes up a single volumetric image).\"
\nhttps://en.wikipedia.org/wiki/Voxel
\nLast revised by Dr Daniel J Bell on 04 Sep 2018
\n\"Voxel size is an important component of image quality. Voxel is the 3-D analog of a pixel. Voxel size is related to both the pixel size and slice thickness. Pixel size is dependent on both the field of view and the image matrix. The pixel size is equal to the field of view divided by the matrix size. The matrix size is typically 128x, 256x or 512x. Pixel size is typically between 0.5 and 1.5 mm. The smaller the pixel size, the greater the image spatial resolution.\"
\n\"Increased voxel size results in an increased signal-to-noise ratio. The trade-off for increased voxel size is decreased spatial resolution. Voxel size can be influenced by receiver coil characteristics. For examples, surface coils indirectly improve resolution by enabling a smaller voxel size for the same signal-to-noise ratio.\"
\n\"Voxel size can contribute to artifacts in MRI. Many MR artifacts are attributable to errors in the underlying spatial encoding of the radiofrequency signals arising from image voxels. Motion artefacts can occur in the phase-encoding direction because a specific tissue voxel may change location between acquisition cycles, leading to phase encoding errors. This manifests as a streak or ghost in the final image, and can be reduced with image gating and regional presaturation techniques.\"
\nhttps://radiopaedia.org/articles/voxel-size-1
\nAuthors: Andrew L Goertzen, Freek J Beekman, Simon R Cherry - DOI:10.1109/NSSMIC.2000.949326
\n\"In computer simulations of X-ray CT systems one can either use continuous geometrical descriptions for phantoms or a voxelized representation. The voxelized approach allows arbitrary phantoms to be defined without being confined to geometrical shapes. \"
\n\"The disadvantage of the voxelized approach is that inherent errors are introduced due to the phantom voxelization. To study effects of phantom discretization, analytical CT simulations were run for a fan-beam geometry with phantom voxel sizes ranging from 0.0625 to 2 times the reconstructed pixel size and noise levels corresponding to 10 to 10 photons per detector pixel prior to attenuation. Differences in the filtered back-projection (FBP) images caused by different phantom matrix sizes were assessed by calculating the difference between reconstructions based on the finest matrix and coarser matrix simulations.\"
\n\"In noise free simulations, all phantom matrix sizes produced a measurable difference from the almost continuous case. When even a small amount of noise was added to the projection data, the differences due to the phantom discretization were masked by the noise, and in all cases there was almost no improvement by using a phantom matrix that was more than twice as fine as the reconstruction matrix.\"
\nhttps://www.researchgate.net/publication/3914486_Effect_of_voxel_size_in_CT_simulations
\nAutomatic detection and voxel-wise mapping of lumbar spine Modic changes with deep learning
\nAuthors: Kenneth T. Gao, Radhika Tibrewala, Madeline Hess, Upasana U. Bharadwaj, Gaurav Inamdar, Thomas M. Link, Cynthia T. Chin, Valentina Pedoia, Sharmila Majumdar
\n\"Modic changes (MCs) are the most prevalent classification system for describing magnetic resonance imaging (MRI) signal intensity changes in the vertebrae. However, there is a growing need for novel quantitative and standardized methods of characterizing these anomalies, particularly for lesions of transitional or mixed nature, due to the lack of conclusive evidence of their associations with low back pain. This retrospective imaging study aims to develop an interpretable deep learning-based detection tool for voxel-wise mapping of MCs.\"
\n\"The model successfully identified the presence of changes in 85.7% of samples in the unseen test set with a sensitivity of 0.71 (±0.072), specificity of 0.95 (±0.022), and Cohen's kappa score of 0.63. In the AI-assisted experiment, the agreement between the junior radiologist and the senior neuroradiologist significantly improved from Cohen's kappa score of 0.52 to 0.58 (p < 0.05).\"
\n\"That deep learning-based approach demonstrates substantial agreement with radiologists and may serve as a tool to improve inter-rater reliability in the assessment of MCs.\"
\nhttps://onlinelibrary.wiley.com/doi/10.1002/jsp2.1204
\n\"A voxel is the smallest 3D element of the volume,2 and is typically represented as a cube or a box, with height, width and depth. Just as 2D images are made of several pixels (represented as squares, with height and width) and the smaller the pixel the better the quality of the picture, the same concept applies to a 3D data volume. Each three-dimensional voxel represents specific x-ray absorption.\"
\n\"The voxel size on CBCT images is isotropic, which means that all the sides are the same dimension with uniform resolution in all directions. In contrast, an MDCT voxel is in general nonisotropic meaning that one side of the voxel is different in dimension. This is considered an advantage of the CBCT because if a certain structure needs to be measured, the measurement will be exact in all the three orthogonal planes. There are different voxel sizes depending on the capabilities of each unit. The small field of view units may use a small voxel size of 0.076 mm, which enables visualization of very small changes to structures. Other voxel sizes available for CBCT units are variable, such as 0.2 mm, 0.3 mm, and 0.4 mm. It is important to note that the larger the voxel size, the less resolution the image will have and less capability to differentiate between small structures. The voxel size is dependent of the imaging objective and the size of the unit detector.\"
\nhttps://www.dentalcare.com/en-us/ce-courses/ce531/voxel
\nNiBabel: Access a cacophony of neuro-imaging file formats
\n\"A nibabel (and nipy) image is the association of three things:
\n\"The image data array: a 3D or 4D array of image data. An affine array that tells you the position of the image array data in a reference space. image metadata (data about the data) describing the image, usually in the form of an image header.\"
\nThis document describes how the affine array describes the position of the image data in a reference space. On the way we will define what we mean by reference space, and the reference spaces that Nibabel uses.
\nSNIPPET:
\nimport numpy as np
\nimport nibabel as nib
\naffine = np.eye(4) # identity affine
\nvoxel_data = np.random.normal(size=(10, 11, 12))
\nimg = nib.Nifti1Image(voxel_data, affine)
https://nipy.org/nibabel/image_orientation.html
\nhttps://nipy.org/nibabel/tutorials.html
\nIntroduction to Dicoms
\nhttps://nipy.org/nibabel/dicom/dicom_intro.html
\nConnecting voxel spaces By Michael Beregov
\nhttps://www.kaggle.com/code/boojum/connecting-voxel-spaces
Normalized Voxels: Align Planes and Crop By YU4U
\nhttps://www.kaggle.com/code/ren4yu/normalized-voxels-align-planes-and-crop
\"In 3D computer graphics, a voxel represents a value on a regular grid in three-dimensional space. As with pixels in a 2D bitmap, voxels themselves do not typically have their position (i.e. coordinates) explicitly encoded with their values. Instead, rendering systems infer the position of a voxel based upon its position relative to other voxels (i.e., its position in the data structure that makes up a single volumetric image).\"
\nhttps://en.wikipedia.org/wiki/Voxel
\nLast revised by Dr Daniel J Bell on 04 Sep 2018
\n\"Voxel size is an important component of image quality. Voxel is the 3-D analog of a pixel. Voxel size is related to both the pixel size and slice thickness. Pixel size is dependent on both the field of view and the image matrix. The pixel size is equal to the field of view divided by the matrix size. The matrix size is typically 128x, 256x or 512x. Pixel size is typically between 0.5 and 1.5 mm. The smaller the pixel size, the greater the image spatial resolution.\"
\n\"Increased voxel size results in an increased signal-to-noise ratio. The trade-off for increased voxel size is decreased spatial resolution. Voxel size can be influenced by receiver coil characteristics. For examples, surface coils indirectly improve resolution by enabling a smaller voxel size for the same signal-to-noise ratio.\"
\n\"Voxel size can contribute to artifacts in MRI. Many MR artifacts are attributable to errors in the underlying spatial encoding of the radiofrequency signals arising from image voxels. Motion artefacts can occur in the phase-encoding direction because a specific tissue voxel may change location between acquisition cycles, leading to phase encoding errors. This manifests as a streak or ghost in the final image, and can be reduced with image gating and regional presaturation techniques.\"
\nhttps://radiopaedia.org/articles/voxel-size-1
\nAuthors: Andrew L Goertzen, Freek J Beekman, Simon R Cherry - DOI:10.1109/NSSMIC.2000.949326
\n\"In computer simulations of X-ray CT systems one can either use continuous geometrical descriptions for phantoms or a voxelized representation. The voxelized approach allows arbitrary phantoms to be defined without being confined to geometrical shapes. \"
\n\"The disadvantage of the voxelized approach is that inherent errors are introduced due to the phantom voxelization. To study effects of phantom discretization, analytical CT simulations were run for a fan-beam geometry with phantom voxel sizes ranging from 0.0625 to 2 times the reconstructed pixel size and noise levels corresponding to 10 to 10 photons per detector pixel prior to attenuation. Differences in the filtered back-projection (FBP) images caused by different phantom matrix sizes were assessed by calculating the difference between reconstructions based on the finest matrix and coarser matrix simulations.\"
\n\"In noise free simulations, all phantom matrix sizes produced a measurable difference from the almost continuous case. When even a small amount of noise was added to the projection data, the differences due to the phantom discretization were masked by the noise, and in all cases there was almost no improvement by using a phantom matrix that was more than twice as fine as the reconstruction matrix.\"
\nhttps://www.researchgate.net/publication/3914486_Effect_of_voxel_size_in_CT_simulations
\nAutomatic detection and voxel-wise mapping of lumbar spine Modic changes with deep learning
\nAuthors: Kenneth T. Gao, Radhika Tibrewala, Madeline Hess, Upasana U. Bharadwaj, Gaurav Inamdar, Thomas M. Link, Cynthia T. Chin, Valentina Pedoia, Sharmila Majumdar
\n\"Modic changes (MCs) are the most prevalent classification system for describing magnetic resonance imaging (MRI) signal intensity changes in the vertebrae. However, there is a growing need for novel quantitative and standardized methods of characterizing these anomalies, particularly for lesions of transitional or mixed nature, due to the lack of conclusive evidence of their associations with low back pain. This retrospective imaging study aims to develop an interpretable deep learning-based detection tool for voxel-wise mapping of MCs.\"
\n\"The model successfully identified the presence of changes in 85.7% of samples in the unseen test set with a sensitivity of 0.71 (±0.072), specificity of 0.95 (±0.022), and Cohen's kappa score of 0.63. In the AI-assisted experiment, the agreement between the junior radiologist and the senior neuroradiologist significantly improved from Cohen's kappa score of 0.52 to 0.58 (p < 0.05).\"
\n\"That deep learning-based approach demonstrates substantial agreement with radiologists and may serve as a tool to improve inter-rater reliability in the assessment of MCs.\"
\nhttps://onlinelibrary.wiley.com/doi/10.1002/jsp2.1204
\n\"A voxel is the smallest 3D element of the volume,2 and is typically represented as a cube or a box, with height, width and depth. Just as 2D images are made of several pixels (represented as squares, with height and width) and the smaller the pixel the better the quality of the picture, the same concept applies to a 3D data volume. Each three-dimensional voxel represents specific x-ray absorption.\"
\n\"The voxel size on CBCT images is isotropic, which means that all the sides are the same dimension with uniform resolution in all directions. In contrast, an MDCT voxel is in general nonisotropic meaning that one side of the voxel is different in dimension. This is considered an advantage of the CBCT because if a certain structure needs to be measured, the measurement will be exact in all the three orthogonal planes. There are different voxel sizes depending on the capabilities of each unit. The small field of view units may use a small voxel size of 0.076 mm, which enables visualization of very small changes to structures. Other voxel sizes available for CBCT units are variable, such as 0.2 mm, 0.3 mm, and 0.4 mm. It is important to note that the larger the voxel size, the less resolution the image will have and less capability to differentiate between small structures. The voxel size is dependent of the imaging objective and the size of the unit detector.\"
\nhttps://www.dentalcare.com/en-us/ce-courses/ce531/voxel
\nNiBabel: Access a cacophony of neuro-imaging file formats
\n\"A nibabel (and nipy) image is the association of three things:
\n\"The image data array: a 3D or 4D array of image data. An affine array that tells you the position of the image array data in a reference space. image metadata (data about the data) describing the image, usually in the form of an image header.\"
\nThis document describes how the affine array describes the position of the image data in a reference space. On the way we will define what we mean by reference space, and the reference spaces that Nibabel uses.
\nSNIPPET:
\nimport numpy as np
\nimport nibabel as nib
\naffine = np.eye(4) # identity affine
\nvoxel_data = np.random.normal(size=(10, 11, 12))
\nimg = nib.Nifti1Image(voxel_data, affine)
https://nipy.org/nibabel/image_orientation.html
\nhttps://nipy.org/nibabel/tutorials.html
\nIntroduction to Dicoms
\nhttps://nipy.org/nibabel/dicom/dicom_intro.html
\nConnecting voxel spaces By Michael Beregov
\nhttps://www.kaggle.com/code/boojum/connecting-voxel-spaces
Normalized Voxels: Align Planes and Crop By YU4U
\nhttps://www.kaggle.com/code/ren4yu/normalized-voxels-align-planes-and-crop