States Analysis#

[16]:

import os
import sys
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt
from matplotlib.ticker import FormatStrFormatter
sys.path.append('..\\..\\neurokin\\')
from neurokin.kinematic_data import KinematicDataRun
from neurokin.utils.neural import processing, importing
from neurokin.utils.experiments import neural_states_helper, neural_correlates_plot, spider_factory
from neurokin.locomotion_states import NeuralCorrelatesStates

[3]:

experiment_path = "./data/neural_correlates_cervical_dataset"
outputfolder = "./output_neural_correlates/"

NFFT = 2**12
NOV = int(NFFT/4)
TIME_CUTOFF = 1.5


pda = ["NWE00053", "NWE00054", "NWE00130", "NWE00160", "NWE00161", "NWE00162", "NWE00163", "NWE00164"]
skip_animals = ["NWE00053", "NWE00054", "NWE00052"]

CHANNEL_DICT = {"NWE00052": 6,
                "NWE00053": 1,
                "NWE00054": 1,
                "NWE00089": 1,
                "NWE00090": 1,
                "NWE00092": 1,
                "NWE00093": 1,
                "NWE00130": 3,
                "NWE00131": 2,
                "NWE00158": 3,
                "NWE00159": 3,
                "NWE00166": 3,
                "NWE00160": 3,
                "NWE00161": 3,
                "NWE00162": 3,
                "NWE00163": 3,
                "NWE00164": 3}

experiment_structure_path = experiment_path + "/experiment_structure.yaml"
conditions = ["baseline", "cervical125", "lumbar40"]

[4]:

load_saved = False
save_computed_datasets = False

ncs = NeuralCorrelatesStates(timeslice=TIME_CUTOFF,
                            experiment_structure_filepath=experiment_structure_path,
                            skip_subjects=skip_animals)


#ncs.create_events_dataset(experiment_path, verbose=False)
ncs.events_dataset = pd.read_pickle("./raw_states.pkl")

[5]:

#ncs.create_raw_neural_dataset(experiment_path,
#                                  stream_names=["LFP1", "NPr1", "EOG1"],
#                                  ch_of_interest=CHANNEL_DICT)
ncs.raw_neural_correlates_dataset = pd.read_pickle("./raw_neural.pkl")
ncs.fs=24414.1

[6]:

ncs.raw_neural_correlates_dataset.head(5)

[6]:

	date	subject	condition	run	event_gait	event_nlm_rest	event_nlm_active	event_fog_rest	event_fog_active	fs
0	230525	NWE00160	baseline	09	[]	[[0.00028505633, 0.0002514952, 0.00024686247, ...	[]	[[-5.251616e-05, -2.5121648e-05, -2.7678883e-0...	[]	24414.0625
1	230525	NWE00160	baseline	10	[]	[]	[]	[]	[]	NaN
2	230525	NWE00160	baseline	18	[]	[]	[]	[]	[]	NaN
3	230525	NWE00160	cervical40	03	[]	[]	[]	[]	[]	NaN
4	230525	NWE00160	cervical40	17	[[3.5766127e-05, -1.108076e-05, 1.40058355e-05...	[]	[]	[]	[]	24414.0625

[7]:

ncs.create_psd_dataset(NFFT, NOV, zscore=True)

[8]:

ncs.psds_correlates_dataset.head(5)

[8]:

	date	subject	condition	run	fs	event_gait	event_nlm_rest	event_nlm_active	event_fog_rest	event_fog_active
0	230525	NWE00160	baseline	09	24414.0625	None	[[10.539083, 41.68482, 11.901685, 0.85658735, ...	None	[[9.614545, 40.348915, 8.879055, 4.4661875, 4....	None
1	230525	NWE00160	baseline	10	NaN	None	None	None	None	None
2	230525	NWE00160	baseline	18	NaN	None	None	None	None	None
3	230525	NWE00160	cervical40	03	NaN	None	None	None	None	None
4	230525	NWE00160	cervical40	17	24414.0625	[[4.4795284, 38.52379, 18.883919, 4.775262, 3....	None	None	None	None

[9]:

stats = ncs.plot_prep_states_distribution(pda, condense=True, stat="std")

[10]:

labels = ["Gait", "Akinesia",  "\n Stationary \n movement"]

N = 3
theta = spider_factory.spider_factory(N, frame='polygon')

fig, ax = plt.subplots(figsize=(5, 5),
                            subplot_kw=dict(projection='radar'))

zorder = 1
data = stats[(stats["group"]==True) & (stats["condition"]=="baseline")]
spider_factory.plot_spider_single_trace(ax, data, color="crimson", theta=theta)
spider_factory.plot_spider_single_trace(ax, stats[(stats["group"]==False) & (stats["condition"]=="baseline")], color="royalblue", theta=theta)


ax.set_ylim(0, 80)
ax.set_yticklabels([])
ax.set_xticklabels([])
ax.set_rgrids([10, 30, 50, 70])
ax.set_varlabels(labels)

#plt.savefig("boilerplate_example_states.svg" ,transparent=True)

../_images/notebooks_DEMO_states_analysis_9_0.png

[11]:

ncs.freqs = pd.read_pickle("./freqs.pkl")

[12]:

max_freq = 100
min_freq = 5

idx_max = processing.find_closest_index(ncs.freqs, max_freq)
idx_min = processing.find_closest_index(ncs.freqs, min_freq)

[13]:

df = ncs.plot_prep_psds_dataset(test_sbj_list=pda, condense=True)

[14]:

df[df["group"] == True].head(5)

[14]:

	subject	condition	event_gait	event_nlm	event_fog	group
0	NWE00130	baseline	[7.700155, 35.852634, 18.280584, 4.251594, 3.7...	[7.1273823, 34.738632, 13.82882, 6.276072, 5.2...	[6.6896896, 34.363136, 12.691377, 4.172866, 5....	True
1	NWE00130	cervical125	[5.428129, 34.610428, 15.265413, 2.9339435, 2....	[7.0085664, 35.285408, 13.780792, 3.5994895, 3...	[6.409625, 35.018906, 12.921191, 4.4829187, 4....	True
2	NWE00130	cervical40	[6.8879814, 31.38297, 16.258255, 3.4334824, 2....	[7.0990953, 34.377922, 14.673093, 3.4678595, 2...	[6.3135886, 29.972769, 11.827717, 4.269376, 5....	True
9	NWE00160	baseline	[8.911066, 36.37505, 18.393803, 5.495706, 3.72...	[10.099365, 40.38185, 15.080096, 3.4323914, 1....	[9.614545, 40.348915, 8.879055, 4.4661875, 4.5...	True
10	NWE00160	cervical125	[6.141581, 36.417072, 16.629639, 3.972267, 2.1...	[7.6479807, 33.596912, 19.159771, 4.5740914, 3...	None	True

[15]:

fig, axs = plt.subplots(1,3, sharey=True, figsize=(10,5))
colors = ["crimson" ,"forestgreen", "royalblue", ]
states = ["event_fog", "event_nlm", "event_gait",]
titles = ["Akinesia", "Stationary \n movement", "Gait"]


# Plot Parkinson's animals
for i, ax in enumerate(axs):
    neural_correlates_plot.plot_psd_single_state(ax, df,  group=True,
                      condition="baseline",
                      state=states[i],
                      freqs=ncs.freqs,
                      color=colors[i],
                      idx_min=idx_min,
                      idx_max=idx_max)
    ax.set_xlabel("Freqs [Hz]", fontsize=15)
    ax.set_title(f"{titles[i]}")

# Plot Sham animals
for i, ax in enumerate(axs[1:]):
    i = i+1
    neural_correlates_plot.plot_psd_single_state(ax, df,  group=False,
                      condition="baseline",
                      state=states[i],
                      freqs=ncs.freqs,
                      color="k",
                      idx_min=idx_min,
                      idx_max=idx_max)

ylim = 15 #zscore

axs[0].set_ylabel("PSD [AU]", fontsize=15)
ax.set_ylim(0, ylim)
for a in axs:
    a.spines[['right', 'top']].set_visible(False)

../_images/notebooks_DEMO_states_analysis_14_0.png

[ ]: