Notebook

BrainWeb-based multimodal models of 20 normal brains¶

Download and Preprocessing for PET-MR Simulations

This notebook will not re-download/re-process files if they already exist.

Output data
- ~/.brainweb/subject_*.npz: dtype(shape): float32(127, 344, 344)
Raw data source
- ~/.brainweb/subject_*.bin.gz: dtype(shape): uint16(362, 434, 362)
Prerequisites
- Python: requirements.txt (e.g. pip install -r requirements.txt)

Author: Casper da Costa-Luis <casper.dcl@physics.org>
Date: 2017-19
Licence: MPLv2.0

In [ ]:

from __future__ import print_function, division
%matplotlib notebook
import brainweb
from brainweb import volshow
import numpy as np
from os import path
from tqdm.auto import tqdm
import logging
logging.basicConfig(level=logging.INFO)

Raw Data¶

In [ ]:

# download
files = brainweb.get_files()

# read last file
data = brainweb.load_file(files[-1])

# show last subject
print(files[-1])
volshow(data, cmaps=['gist_ncar']);

Transform¶

$\ifcsname bm\endcsname\else\newcommand{\bm}[1]{\mathbf{#1}}\fi$

Convert raw image data:

Siemens Biograph mMR resolution (~2mm) & dimensions (127, 344, 344)
PET/T1/T2/uMap intensities
randomised structure for PET/T1/T2
- $\bm{\theta} \circ (\bm{1} + \gamma[2G_\sigma(\bm{\rho}) - \bm{1}])$
  - $\bm{\rho} = rand(127, 344, 344) \in [0, 1)$
  - Gaussian smoothing $\sigma = 1$
  - $\gamma = \left\{\matrix{1 & \text{for PET}\\ 0.75 & \text{for MR}}\right.$
  - $\bm{\theta}$ is the PET or MR piecewise constant phantom

In [ ]:

brainweb.seed(1337)

for f in tqdm(files, desc="mMR ground truths", unit="subject"):
    vol = brainweb.get_mmr_fromfile(
        f,
        petNoise=1, t1Noise=0.75, t2Noise=0.75,
        petSigma=1, t1Sigma=1, t2Sigma=1)

In [ ]:

# show last subject
print(f)
volshow([vol['PET' ][:, 100:-100, 100:-100],
         vol['uMap'][:, 100:-100, 100:-100],
         vol['T1'  ][:, 100:-100, 100:-100],
         vol['T2'  ][:, 100:-100, 100:-100]],
        cmaps=['hot', 'bone', 'Greys_r', 'Greys_r'],
        titles=["PET", "uMap", "T1", "T2"]);

In [ ]:

# add some lesions
brainweb.seed(1337)
im3d = brainweb.add_lesions(vol['PET'])
volshow(im3d[:, 100:-100, 100:-100], cmaps=['hot']);