I’ve performed a searchlight analysis with Python using machine learning techniques, and I’ve obtained the coordinates of specific voxels based on a threshold. Is there a way to obtain the labels of the corresponding brain regions for these coordinates?(e.g. Right Cerebral Cortex…)
array([[20, 12, 25],
[20, 13, 21],
[20, 13, 24],
...,
[20, 66, 26],
[20, 66, 29],
[20, 67, 23]], dtype=int64)
First, I performed with GPT then GPT gave me the code, but not work.
from nilearn import datasets, image
# Harvard-Oxford Atlas
atlas = datasets.fetch_atlas_harvard_oxford('cort-maxprob-thr25-2mm')
# Create an empty list to store the region names for the coordinates
mapped_regions = []
# Load the Harvard-Oxford Atlas image
atlas_img = atlas.maps
# Define the reference image, which should have the same resolution as your bold data
# Resample the atlas image to the resolution of the reference image
atlas_img_resampled = image.resample_to_img(atlas_img, bold)
# Iterate through the coordinates
for coord in above_threshold_coords:
x, y, z = coord # Coordinates directly from above_threshold_coords
# Extract the region value from the resampled atlas image
region_value = atlas_img_resampled.get_fdata()[x, y, z]
# Get the region name from the atlas labels
region_name = atlas.labels[int(region_value)]
# Append the region name to the list
mapped_regions.append(region_name)
# The mapped_regions list now contains the names of brain regions corresponding to the coordinates.
print(mapped_regions)
IndexError Traceback (most recent call last)
Cell In[78], line 25
22 region_value = atlas_img_resampled.get_fdata()[x, y, z]
24 # Get the region name from the atlas labels
—> 25 region_name = atlas.labels[int(region_value)]
27 # Append the region name to the list
28 mapped_regions.append(region_name)
IndexError: list index out of range