Author: Chithra Srinivasan
To run the program follow the format: python analysis.py
Prompts:
- filename is the name of the nii file
- x1,y1,z1 is the voxel of the seed 1
- x2,y2,z2 is the voxel of the seed 2
- choice is 1 or 2 - 1 is for finding correlation between two seeds - 2 is for finding significant correlation
Classification: python classification.py (Choose one of the techniques. "0" for all the techniques to run)
Regression: python regression.py (Choose one of the techniques. Eg., 1, 2 or 5. "0" for all the techniques to run)
You would need the following python packages installed: numpy, scipy, pandas, nibabel, matplotlib
cmh_01_func.nii
cmh_02_func.nii
cmh_03_func.nii
mrc_01_func.nii
mrc_02_func.nii
mrc_03_func.nii
The coordinates are specified as values to the keys: Eg., [28,33,9], [30,28,9] These coordinates are defined as 'points' in analysis.py
{'ventral rostral putamen':
{'orbital frontal cortex': np.array([28,33,9]),
'Insula': np.array([45,21,13]),
'Subcollosal cortex': np.array([30,28,9]),
'Thalamus_+': np.array([24,13,15]),
'Thalamus_-': np.array([37,13,16]),
'Dorsolateral Prefrontal cortex': np.array([45,32,20])
},
'Nucleus Accumbens':
{'Middle Temporal Gyrus': np.array([45,3,16]),
'Inferios frontal gyrus': np.array([49,28,14]),
'Superior parietal lobule': np.array([36,0,25]),
'Thalamus': np.array([36,17,15]),
'Supramarginal gyrus': np.array([14,10,25]),
'Lateral occipital cortex': np.array([27,0,24])
},
'Dorsal Caudate':
{'inferior temporal gyrus': np.array([49,1,11]),
'frontal pole': np.array([24,33,9]),
'putamen': np.array([39,27,14]),
'accumbens': np.array([30,28,14]),
'planum temporale_-': np.array([43,12,17]),
'occipital cortex': np.array([29,0,18]),
'planum temporale_+': np.array([17,0,17]),
'superior frontal gyrus': np.array([37,29,28])
},
'Dorsal caudal putamen':
{'Insula_-': np.array([50,21,17]),
'Insula_+': np.array([16,24,13]),
'Insula_=': np.array([45,19,14]),
'planum polare': np.array([13,20,16]),
'dorsolateral prefrontal cortex': np.array([45,31,20])
},
'Dorsal rostral putamen':
{'Frontal operculum cortex': np.array([16,26,13]),
'Insula': np.array([45,19,14]),
'Orbital frontal cortex': np.array([27,33,10]),
'Anterior cingulate cortex': np.array([28,40,16]),
'Middle temporal gyrus': np.array([52,6,14]),
'Middle frontal gyrus': np.array([45,31,20])
},
'Ventral caudate':
{'Accumbens, putamen': np.array([37,28,11]),
'Accumbens, caudate': np.array([30,27,13]),
'middle frontal gyrus': np.array([43,23,5]),
'Thalamus_-': np.array([27,13,16]),
'Insula': np.array([43,15,18]),
'Thalamus_+': np.array([35,16,16]),
'Occipital cortex': np.array([27,0,18]),
'supramargival gyrus': np.array([18,10,17])
}
}
(by Shuo) The new regression code (with plotting) is added, to draw the prediction errors treating all features as output (one by one), in two classes and for all regression models.