dcm4py

dcm4py is a Python library for converting DICOM images to NumPy arrays and vice versa, allowing you to convert NumPy arrays back into DICOM images.

Features

• Convert DICOM images to NumPy arrays for easy manipulation.
• Optionally normalize the NumPy arrays with a custom range.
• Convert processed NumPy arrays back into DICOM format.

Installation

You can install requirements via pip:

pip install -r requirements.txt

Usage

Convert DICOM to NumPy

To generate a numpy array from DICOM images, use the dcm2npy function and input the path to the directory of DICOM images. The normalize parameter to True to normalize the array values, and the range to normalize to can be defined in the norm_range parameter.

from dcm4py import dcm2npy
slices, dataset = dcm2npy('/path/to/input_directory', normalize=True, norm_range=(0, 1))

Convert NumPy to DICOM

After processing the slices NumPy array, you can convert it back to DICOM images in a specified folder using the npy2dcm function. If the array was previously normalized, it can be denormalized by setting the denormalize parameter to True. The current range of the array must be specified.

from dcm4py import npy2dcm
npy2dcm(output_path='/path/to/output_directory', slices=slices, dataset=dataset, denormalize=True, current_range=(0, 1))