Computed tomography image dataset of neurocysticercosis
Description
The images correspond to neurocysticercosis found in CT images of positive patients. The procedure begins by identifying the presence of neurocysticercosis in the CT images. The images were cropped, yielding 1,059 images contained in the "RGB_originals" folder. These images are in PNG format using the RGB color model. The "PNG_grayscale" folder contains 1,059 images in single-channel grayscale configuration. The "DICOM_grayscale" folder contains 1,059 images in DICOM format with the original image configuration. Of the 1059 images, 150 images were randomly taken to which various processes of marking the neurocysticercosis were performed, being in the "pro_market" folder, obtaining a mask of the neurocysticercosis saving in the "pro_mask" folder, the neurocysticercosis was segmented saving them in the "pro_segmented" folder, these images were saved in two formats PNG located in the "pro_originals_PNG" folder and TIFF in the "pro_originals_TIFF" folder, from these marked images, values of characteristics such as area, perimeter, contrast, correlation, energy, homogeneity among others were extracted, which are found in the "pro_feature" folder that contains the values of the characteristics in csv and txt format.
Files
Institutions
- Universidad Privada del Norte