Arithmetic operation Tab

The Image Arithmetics options dialog, Arithmetic operation tab

Image arithmetic operations applied to a single dataset

Arithmetic Operation tab: Set the arithmetic operation to be applied on a single dataset. 

Most single series arithmetic operations produce an output image (S3) based on an arithmetic operation between current pixel in the input (source) image (S2) and some reference data (S1) which may e.g. be the average of multiple time-points or a ROI value.

Single series image operation

        Subtract (S3=S2-S1):  Subtract each image in the active image set from a range of baseline images. The baseline images to use are set in the Source Image(s) option (see below). Can also be performed directly by clicking the toolbar icon shown in  Image Subtract . (using the current setting for Source Images).

        Ratio (S3=S2/S1): Output image is the ratio of the source image and reference data

        Percent difference (S3=100 x (S2-S1)/S1): Creates relative (in percent) difference images. The baseline images (S2) are set in Source Image(s) option (see below).

        Time series variance: Calculates the variance (square of the std deviation) for each pixel in a time series. This is a useful function for assessing abnormal signal variations in e.g. fMRI time series.

        delSI->delR2: Conversion from deltaSI (change in MR signal intensity) to deltaR2 (change in T2 relaxation rate due to contrast enhancement) using the expression: delR2=ln(S1/S2)/TE where TE is the echo time (set in <TE> entry).

        delSI->delR1: Conversion from deltaSI (change in MR signal intensity) to deltaR1 (change in T1 relaxation rate due to contrast enhancement) using either the expression for a spoiled gradient echo (SPGR) sequence of a saturation recovery (SR) sequence. For both sequences, additional parameters (TR/TD, flip angle and baseline T1) must be specified. Pixel-wise baseline T1-values can also be extracted from an overlay image, if present.

        Average (Sum(Sn)/n): Calculates an average image from all the active images by default. It is also possible to average over sub-sets of the total dataset, as defined in the <Parameters> field. This is mainly useful for dynamic series where this option enables multiple time-points to be averaged, thereby �truncating� the temporal resolution.

        Multiply: Calculates the product of the reference and source image

        Temporal SNR: The temporal SNR is calculated as the pixel-wise mean across all time-points divided by the standard deviation

        Linear transformation (S3=f1*S2+f1): The output images are linear transformations of the input images where the transformation is specified by the two constants f1(slope) and f2 (offset). 

        pCASL quantification (GE): A special operation to extract quantitative CBF values from pCASL data generated on certain GE scanners where the input data must contain a 4D dataset where the first dynamics contains a subtracted label/control pCASL series and the second dynamic the corresponding Mo series.

 

Reference data (S1)

Defines the reference data to use in the arithmetic operations (where relevant) as:

        Current Image: The image visible in the active image window when the function was activated.

        Average of images:  Set the range of images to be used as source images. The mean of the source images will be used.

        Current ROI value: The active ROI value for each image is used a reference. If no ROI is active (visible) in one of the source images, the reference value will be set to zero for this image. This option is usually used in combination with a ROI buffer. See Region of interest analysis and the ROI buffer for more details on ROI functionality.

        Mean of all ROI values: Similar to above option but now the mean of all ROI values is used as reference for all images rather that each ROI separately. This option is usually used in combination with a ROI buffer. See Region of interest analysis and the ROI buffer for more details on ROI functionality.

        Image overlay: Reference data is obtained from the active overlay

3D vs 4D analysis:

nordicICE can perform arithmetic operations on both 2D (single slice) 3D (multi-slice) and 4D (dynamic- or multiple 3D datasets). The way nordicICE treats 4D datasets depends on the setting of the  Image Scroll Box. If the scroll box is set to <All> then the data will be treated as a 4D dataset and all calculations will be performed for all dynamic images for each slice. If the scoll box is not set to <All> bu e.g. to <Slice> or <Dyn. Scan> then the arithmetic operation will be performed for the selected slice / dynamic series only. Note that arithmetic analysis on 4D data will usually produce a 4D output dataset (e.g. where each 3D series is subtracted from the first 3D series).

 

Process.

Performs the selected operation

Related topics:

Drag & Drop tab
Pre process tab
Image scaling tab
Image Arithmetics functions
ROI buffer
Region of interest analysis

Logo

Description automatically generated