Image Data Stephen Bailey Instructor DataCamp Biomedical Image - - PowerPoint PPT Presentation

▶
image data
SMART_READER_LITE
LIVE PREVIEW

Image Data Stephen Bailey Instructor DataCamp Biomedical Image - - PowerPoint PPT Presentation

Jul 31, 2023 196 likes •474 views

DataCamp Biomedical Image Analysis in Python BIOMEDICAL IMAGE ANALYSIS IN PYTHON Image Data Stephen Bailey Instructor DataCamp Biomedical Image Analysis in Python Biomedical imaging: more than a century of discovery 1895 2017 DataCamp

slide-1
SLIDE 1

DataCamp Biomedical Image Analysis in Python

Image Data

BIOMEDICAL IMAGE ANALYSIS IN PYTHON

Stephen Bailey

Instructor

slide-2
SLIDE 2

DataCamp Biomedical Image Analysis in Python

Biomedical imaging: more than a century of discovery

1895 2017

slide-3
SLIDE 3

DataCamp Biomedical Image Analysis in Python

Course objectives

Toolbox ImageIO NumPy SciPy matplotlib

slide-4
SLIDE 4

DataCamp Biomedical Image Analysis in Python

Loading images

imageio: read and save images Image objects are NumPy arrays.

Slice the array by specifying values along each available dimension.

import imageio im = imageio.imread('body-001.dcm') type(im) imageio.core.Image im Image([[125, 135, ..., 110], [100, 130, ..., 100], ..., [100, 150, ..., 100]], dtype=uint8) im[0, 0] 125 im[0:2, 0:2] Image([[125, 135], [100, 130]], dtype=uint8)

slide-5
SLIDE 5

DataCamp Biomedical Image Analysis in Python

Metadata

Metadata: the who, what, when, where and how of image acquisition Accessible in Image objects through the meta dictionary attribute

im.meta Dict([('StudyDate', '2017-01-01'), ('Modality', 'MR'), ('PatientSex', F), ... ('shape', (256, 256)]) im.meta['Modality'] 'CT' im.meta.keys()

  • dict_keys(['StudyDate',

'SeriesDate', 'PatientSex', ... 'shape'])

slide-6
SLIDE 6

DataCamp Biomedical Image Analysis in Python

Plotting images

Matplotlib's imshow() function displays 2D image data Many colormaps available but often shown in grayscale (cmap='gray') Axis ticks and labels are often not useful for images

import matplotlib.pyplot as plt plt.imshow(im, cmap='gray') plt.axis('off') plt.show()

slide-7
SLIDE 7

DataCamp Biomedical Image Analysis in Python

Let's practice!

BIOMEDICAL IMAGE ANALYSIS IN PYTHON

slide-8
SLIDE 8

DataCamp Biomedical Image Analysis in Python

N-dimensional images

BIOMEDICAL IMAGE ANALYSIS IN PYTHON

Stephen Bailey

Instructor

slide-9
SLIDE 9

DataCamp Biomedical Image Analysis in Python

Images of all shapes and sizes

im[row, col]

slide-10
SLIDE 10

DataCamp Biomedical Image Analysis in Python

Images of all shapes and sizes

vol[pln, row, col]

slide-11
SLIDE 11

DataCamp Biomedical Image Analysis in Python

Images of all shapes and sizes

im[row, col, ch]

slide-12
SLIDE 12

DataCamp Biomedical Image Analysis in Python

Images of all shapes and sizes

im_ts[time, row, col, ch]

slide-13
SLIDE 13

DataCamp Biomedical Image Analysis in Python

N-dimensional images are stacks of arrays

import imageio import numpy as np im1=imageio.imread('chest-000.dcm') im2=imageio.imread('chest-001.dcm') im3=imageio.imread('chest-002.dcm') im1.shape (512, 512) vol = np.stack([im1, im2, im3]) vol.shape (3, 512, 512)

slide-14
SLIDE 14

DataCamp Biomedical Image Analysis in Python

Loading volumes directly

imageio.volread():

read multi-dimensional data directly assemble a volume from multiple images

import os

  • s.listdir('chest-data')

['chest-000.dcm', 'chest-001.dcm', 'chest-002.dcm', ..., 'chest-049.dcm'] import imageio vol = imageio.volread('chest-data') vol.shape (50, 512, 512)

slide-15
SLIDE 15

DataCamp Biomedical Image Analysis in Python

Shape, sampling, and field of view

Image shape: number of elements along each axis Sampling rate: physical space covered by each element Field of view: physical space covered along each axis

import imageio vol = imageio.volread('chest-data') # Image shape (in voxels) n0, n1, n2 = vol.shape n0, n1, n2 (50, 512, 512) # Sampling rate (in mm) d0, d1, d2 = vol.meta['sampling'] d0, d1, d2 (2, 0.5, 0.5) # Field of view (in mm) n0 * d0, n1 * d1, n2 * d2 (100, 256, 256)

slide-16
SLIDE 16

DataCamp Biomedical Image Analysis in Python

Let's practice!

BIOMEDICAL IMAGE ANALYSIS IN PYTHON

slide-17
SLIDE 17

DataCamp Biomedical Image Analysis in Python

Advanced plotting

BIOMEDICAL IMAGE ANALYSIS IN PYTHON

Stephen Bailey

Instructor

slide-18
SLIDE 18

DataCamp Biomedical Image Analysis in Python

To plot N-dimensional data slice it!

slide-19
SLIDE 19

DataCamp Biomedical Image Analysis in Python

Plotting multiple images at once

plt.subplots: creates a figure canvas

with multiple AxesSubplots objects.

import imageio vol = imageio.volread('chest-data') fig, axes = plt.subplots(nrows=1, ncols=3) axes[0].imshow(vol[0],cmap='gray') axes[1].imshow(vol[10],cmap='gray') axes[2].imshow(vol[20],cmap='gray') for ax in axes: ax.axis('off') plt.show()

slide-20
SLIDE 20

DataCamp Biomedical Image Analysis in Python

Plotting multiple images at once

slide-21
SLIDE 21

DataCamp Biomedical Image Analysis in Python

Non-standard views

Axial

import imageio vol = imageio.volread('chest-data') view_1v2 = vol[pln, :, :] view_1v2 = vol[pln]

slide-22
SLIDE 22

DataCamp Biomedical Image Analysis in Python

Non-standard views

Coronal

import imageio vol = imageio.volread('chest-data') view_1v2 = vol[pln, :, :] view_1v2 = vol[pln] view_0v2 = vol[:, row, :]

slide-23
SLIDE 23

DataCamp Biomedical Image Analysis in Python

Non-standard views

Sagittal

import imageio vol = imageio.volread('chest-data') view_1v2 = vol[pln, :, :] view_1v2 = vol[pln] view_0v2 = vol[:, row, :] view_0v1 = vol[:, :, col]

slide-24
SLIDE 24

DataCamp Biomedical Image Analysis in Python

Modifying the aspect ratio

Pixels may adopt any aspect ratio:

im = vol[:,:,100] d0, d1, d2 = vol.meta['sampling'] d0, d1, d2 (2, 0.5, 0.5) asp = d0 / d1 asp 3 plt.imshow(im, cmap='gray', aspect=asp) plt.show()

slide-25
SLIDE 25

DataCamp Biomedical Image Analysis in Python

Modifying the aspect ratio

slide-26
SLIDE 26

DataCamp Biomedical Image Analysis in Python

Let's practice!

BIOMEDICAL IMAGE ANALYSIS IN PYTHON