Note
Go to the end to download the full example code
Getting started tutorial¶
Here is a quick overview of physio
import numpy as np
import matplotlib.pyplot as plt
from pathlib import Path
import physio
physio module overview ———————-
physiohas utility functions for reading some formats like:
micromed
brainvision
Here, we use an internal file in the numpy format for the demo
raw_resp = np.load('resp1.npy')
raw_ecg = np.load('ecg1.npy')
srate = 1000.
times = np.arange(raw_resp.size) / srate
fig, axs = plt.subplots(nrows=2, sharex=True)
ax = axs[0]
ax.plot(times, raw_resp)
ax.set_ylabel('raw resp')
ax = axs[1]
ax.plot(times, raw_ecg)
ax.set_ylabel('raw ECG')
ax.set_xlim(185, 225)
(185.0, 225.0)
Analyze respiration¶
compute_respiration() is an easy function to:
preprocess the respiratory signal (resp)
compute cycle features (resp_cycles)
resp, resp_cycles = physio.compute_respiration(raw_resp, srate)
fig, ax = plt.subplots()
ax.plot(times, raw_resp)
ax.plot(times, resp)
ax.set_xlim(185, 225)
(185.0, 225.0)
Respiration cycles and features¶
resp_cycles is a dataframe containing all respiratory cycles as rows. Columns contain features like duration, amplitudes, volumes.
print(resp_cycles.shape)
print(resp_cycles.columns)
columns = ['cycle_duration', 'inspi_volume', 'expi_volume', 'total_amplitude' ]
resp_cycles[columns].plot(kind='hist', subplots=True, sharex=False, layout=(2, 2), bins=50)
resp_cycles
(30, 17)
Index(['inspi_index', 'expi_index', 'next_inspi_index', 'inspi_time',
'expi_time', 'next_inspi_time', 'cycle_duration', 'inspi_duration',
'expi_duration', 'cycle_freq', 'cycle_ratio', 'inspi_volume',
'expi_volume', 'total_amplitude', 'inspi_amplitude', 'expi_amplitude',
'total_volume'],
dtype='object')
inspi_ind = resp_cycles['inspi_index'].values
expi_ind = resp_cycles['expi_index'].values
fig, ax = plt.subplots()
ax.plot(times, resp)
ax.scatter(times[inspi_ind], resp[inspi_ind], color='green')
ax.scatter(times[expi_ind], resp[expi_ind], color='red')
ax.set_xlim(185, 225)
(185.0, 225.0)
Analyse ECG¶
compute_ecg() is an easy function to:
Preprocess the ECG signal output, which is normalized by default (ecg)
Detect R peaks (ecg_peaks)
ecg, ecg_peaks = physio.compute_ecg(raw_ecg, srate)
r_peak_ind = ecg_peaks['peak_index'].values
fig, axs = plt.subplots(nrows=2, sharex=True)
ax = axs[0]
ax.plot(times, raw_ecg)
ax.scatter(times[r_peak_ind], raw_ecg[r_peak_ind], color='#A8AE27')
ax = axs[1]
ax.plot(times, ecg)
ax.scatter(times[r_peak_ind], ecg[r_peak_ind], color='#A8AE27')
ax.set_xlim(185, 225)
(185.0, 225.0)
- ECG metrics
compute_ecg_metrics() is a simple function to compute time-domain Heart Rate Variability (HRV) metrics.
We can visualize these metrics and the RR interval distribution.
ecg_metrics = physio.compute_ecg_metrics(ecg_peaks)
fig, ax = plt.subplots()
ax.hist(np.diff(ecg_peaks['peak_time']) * 1000., bins=np.arange(0, 1400, 10), alpha=0.5)
ax.axvline(ecg_metrics['HRV_Mean'], color='orange', label='HRV_Mean')
ax.axvline(ecg_metrics['HRV_Median'], color='violet', label='HRV_Median')
ax.set_xlabel('HRV [ms]')
ax.legend()
print(ecg_metrics)
HRV_Mean 734.444717
HRV_SD 145.936125
HRV_Median 678.000000
HRV_Mad 115.642973
HRV_CV 0.198703
HRV_MCV 0.170565
HRV_Asymmetry -56.444717
HRV_RMSSD 103.553299
dtype: float64
- Cyclic deformation
deform_traces_to_cycle_template() is a tool to deform traces
to a cycle template by stretching with linear resampling to a fixed number of
points per cycle.
This is helpful to explore if features of a signal are driven by a cyclic phenomenon like respiration.
Here, we deform the signal trace by “itself” : the respiratory cycle. This leads to an average respiratory template.
Importantly, this can be done using one or several segment inside the cycle.
# here we have 3 times per cycle so 2 segments
cycle_times = resp_cycles[['inspi_time', 'expi_time', 'next_inspi_time']].values
deformed_resp_1seg = physio.deform_traces_to_cycle_template(resp, times, cycle_times,
points_per_cycle=40, segment_ratios=0.4,
output_mode='stacked')
print(deformed_resp_1seg.shape, cycle_times.shape)
# here we have 2 times per cycle so 1 segment
cycle_times = resp_cycles[['inspi_time', 'next_inspi_time']].values
deformed_resp_2seg = physio.deform_traces_to_cycle_template(resp, times, cycle_times,
points_per_cycle=40, segment_ratios=None,
output_mode='stacked')
print(deformed_resp_2seg.shape, cycle_times.shape)
fig, axs = plt.subplots(ncols=2, sharey=True)
physio.plot_cyclic_deformation(deformed_resp_1seg, segment_ratios=None, two_cycles=False, ax=axs[0])
axs[0].set_title('Deformation 2 segments')
physio.plot_cyclic_deformation(deformed_resp_2seg, segment_ratios=None, two_cycles=False, ax=axs[1])
axs[1].set_title('Deformation 1 segment')
(30, 40) (30, 3)
(30, 40) (30, 2)
Text(0.5, 1.0, 'Deformation 1 segment')
- Cyclic deformation on ECG
Let’s use the same for ECG trace
We can also use a simple vector in this case it is converted a a 1 segment case.
# This is 1 segment
cycle_times = ecg_peaks['peak_time'].values
deformed_ecg = physio.deform_traces_to_cycle_template(ecg, times, cycle_times,
points_per_cycle=300, segment_ratios=None,
output_mode='stacked')
print(deformed_ecg.shape, cycle_times.shape)
fig, ax = plt.subplots()
physio.plot_cyclic_deformation(deformed_ecg, two_cycles=True, ax=ax)
ax.set_title('Two ECG cycle averaged')
(407, 300) (408,)
Text(0.5, 1.0, 'Two ECG cycle averaged')
plt.show()
Total running time of the script: (0 minutes 2.252 seconds)