def get_all_features_hrva(data_sample, sample_rate=100, rpeak_method=0):
"""
:param data_sample:
:param sample_rate:
:param rpeak_method:
:return:
"""
if rpeak_method in [1, 2, 3, 4]:
detector = PeakDetector()
peak_list = detector.ppg_detector(data_sample, rpeak_method)[0]
else:
rol_mean = rolling_mean(data_sample,
windowsize=0.75,
sample_rate=100.0)
peaks_wd = detect_peaks(data_sample,
rol_mean,
ma_perc=20,
sample_rate=100.0)
peak_list = peaks_wd["peaklist"]
rr_list = np.diff(peak_list) * (1000 / sample_rate) #1000 milisecond
nn_list = get_nn_intervals(rr_list)
nn_list_non_na = np.copy(nn_list)
nn_list_non_na[np.where(np.isnan(nn_list_non_na))[0]] = -1
time_domain_features = get_time_domain_features(rr_list)
frequency_domain_features = get_frequency_domain_features(rr_list)
geometrical_features = get_geometrical_features(rr_list)
csi_cvi_features = get_csi_cvi_features(rr_list)
return time_domain_features,frequency_domain_features,\
geometrical_features,csi_cvi_features
def split_segment(s, split_type=0, duration=30.0,
overlaping=1,sampling_rate=100,
peak_detector=7, wave_type='ppg'):
"""
Save segment waveform and plot (optional) to csv and image file.
Input is a segment with timestamps.
Parameters
----------
s :
array-like represent the signal.
split_type :
0: split by time
1: split by beat
duration :
the duration of each segment if split by time in seconds, default = 30 (second)
the number of complex/beat in each segement if split by beat, default = 30(beat/segment)
sampling_rate :
device sampling rate
peak_detector :
The type of peak detection if split the segment by beat.
wave_type:
Type of signal. Either 'ppg' or 'ecg'
Returns
-------
>>>from vital_sqi.common.utils import generate_timestamp
>>>s = np.arange(100000)
>>>timestamps = generate_timestamp(None,100,len(s))
>>>df = pd.DataFrame(np.hstack((np.array(timestamps).reshape(-1,1),
np.array(s).reshape(-1,1))))
>>>split_segment(df,overlaping=3)
"""
assert check_signal_format(s) is True
if split_type is 0:
chunk_size = int(duration * sampling_rate)
chunk_step = int(overlaping * sampling_rate)
chunk_indices = [
[int(i), int(i + chunk_size)] for i in
range(0, len(s), chunk_size - chunk_step)
]
else:
if wave_type == 'ppg':
detector = PeakDetector(wave_type='ppg')
peak_list, trough_list = detector.ppg_detector(s,detector_type=peak_detector)
else:
detector = PeakDetector(wave_type='ecg')
peak_list, trough_list = detector.ecg_detector(s, detector_type=peak_detector)
chunk_indices = [
[peak_list[i], peak_list[i+duration]] for i in range(0,
len(peak_list),int(duration-overlaping))
]
chunk_indices[0] = 0
milestones = pd.DataFrame(chunk_indices)
segments = cut_segment(s, milestones)
return segments, milestones
def msq_sqi(s, peak_detector_1=7, peak_detector_2=6, wave_type='ppg'):
"""
MSQ SQI as defined in Elgendi et al
"Optimal Signal Quality Index for Photoplethysmogram Signals"
with modification of the second algorithm used.
Instead of Bing's, a SciPy built-in implementation is used.
The SQI tracks the agreement between two peak detectors
to evaluate quality of the signal.
Parameters
----------
s : sequence
A signal with peaks.
peak_detector_1 : array of int
Type of the primary peak detection algorithm, default = Billauer
peak_detect2 : int
Type of the second peak detection algorithm, default = Scipy
Returns
-------
msq_sqi : number
MSQ SQI value for the given signal
"""
if wave_type == 'ppg':
detector = PeakDetector(wave_type='ppg')
peaks_1, trough_list = detector.ppg_detector(
s, detector_type=peak_detector_1)
peaks_2 = detector.ppg_detector(s,
detector_type=peak_detector_2,
preprocess=False)
else:
detector = PeakDetector(wave_type='ecg')
peaks_1, trough_list = detector.ecg_detector(
s, detector_type=peak_detector_1)
peaks_2 = detector.ecg_detector(s,
detector_type=peak_detector_2,
preprocess=False)
if len(peaks_1) == 0 or len(peaks_2) == 0:
return 0.0
peak1_dom = len(np.intersect1d(peaks_1, peaks_2)) / len(peaks_1)
peak2_dom = len(np.intersect1d(peaks_2, peaks_1)) / len(peaks_2)
return min(peak1_dom, peak2_dom)
def msq(x):
"""Mean signal quality"""
# Library
from vital_sqi.common.rpeak_detection import PeakDetector
# Detection of peaks
detector = PeakDetector()
peak_list, trough_list = detector.ppg_detector(x, 7)
# Return
return sq.standard_sqi.msq_sqi(x, peaks_1=peak_list, peak_detect2=6)
def get_peak_error_features(data_sample,
sample_rate=100,
rpeak_detector=0,
low_rri=300,
high_rri=2000):
rules = ["malik", "karlsson", "kamath", "acar"]
try:
wd, m = hp.process(data_sample, sample_rate, calc_freq=True)
except:
try:
wd, m = hp.process(data_sample, sample_rate)
except:
error_dict = {rule + "_error": np.nan for rule in rules}
error_dict["outlier_error"] = np.nan
return error_dict
if rpeak_detector in [1, 2, 3, 4]:
detector = PeakDetector(wave_type='ecg')
peak_list = detector.ppg_detector(data_sample,
rpeak_detector,
preprocess=False)[0]
wd["peaklist"] = peak_list
wd = calc_rr(peak_list, sample_rate, working_data=wd)
wd = check_peaks(wd['RR_list'],
wd['peaklist'],
wd['ybeat'],
reject_segmentwise=False,
working_data=wd)
wd = clean_rr_intervals(working_data=wd)
rr_intervals = wd["RR_list"]
rr_intervals_cleaned = remove_outliers(rr_intervals,
low_rri=low_rri,
high_rri=high_rri)
number_outliers = len(np.where(np.isnan(rr_intervals_cleaned))[0])
outlier_ratio = number_outliers / (len(rr_intervals_cleaned) -
number_outliers)
error_sqi = {}
error_sqi['outlier_error'] = outlier_ratio
interpolated_rr_intervals = interpolate_nan_values(rr_intervals_cleaned)
for rule in rules:
nn_intervals = remove_ectopic_beats(interpolated_rr_intervals,
method=rule)
number_ectopics = len(np.where(np.isnan(nn_intervals))[0])
ectopic_ratio = number_ectopics / (len(nn_intervals) - number_ectopics)
error_sqi[rule + "_error"] = ectopic_ratio
return error_sqi
def get_all_features_hrva(s, sample_rate=100, rpeak_method=0,wave_type='ecg'):
"""
Parameters
----------
data_sample :
Raw signal
rpeak_method :
return: (Default value = 0)
sample_rate :
(Default value = 100)
Returns
-------
"""
# if rpeak_method in [1, 2, 3, 4]:
# detector = PeakDetector()
# peak_list = detector.ppg_detector(data_sample, rpeak_method)[0]
# else:
# rol_mean = rolling_mean(data_sample, windowsize=0.75, sample_rate=100.0)
# peaks_wd = detect_peaks(data_sample, rol_mean, ma_perc=20,
# sample_rate=100.0)
# peak_list = peaks_wd["peaklist"]
if wave_type=='ppg':
detector = PeakDetector(wave_type='ppg')
peak_list, trough_list = detector.ppg_detector(s, detector_type=rpeak_method)
else:
detector = PeakDetector(wave_type='ecg')
peak_list, trough_list = detector.ecg_detector(s, detector_type=rpeak_method)
rr_list = np.diff(peak_list) * (1000 / sample_rate) # 1000 milisecond
nn_list = get_nn_intervals(rr_list)
nn_list_non_na = np.copy(nn_list)
nn_list_non_na[np.where(np.isnan(nn_list_non_na))[0]] = -1
time_domain_features = get_time_domain_features(rr_list)
frequency_domain_features = get_frequency_domain_features(rr_list)
geometrical_features = get_geometrical_features(rr_list)
csi_cvi_features = get_csi_cvi_features(rr_list)
return time_domain_features, frequency_domain_features, geometrical_features, csi_cvi_features
def get_split_rr_index(segment_seconds,sequence):
"""
handy
Return the index of the splitting points
:param segment_seconds: the length of each cut split (in seconds)
:param sequence:
:return:
"""
detector = PeakDetector()
indices = [0]
for i in range(0, int(np.ceil(len(sequence) / segment_seconds))):
chunk = sequence[int(segment_seconds * i):
int(segment_seconds * (i + 1) + 60)]
peak_list, trough_list = detector.ppg_detector(chunk)
if len(trough_list)>0:
indices.append(int(trough_list[-1]+segment_seconds * i))
else:
indices.append(int(segment_seconds * (i+1)))
return indices
def get_all_features_heartpy(data_sample, sample_rate=100, rpeak_detector=0):
"""
Parameters
----------
data_sample :
Raw signal
sample_rate :
(Default value = 100)
rpeak_detector :
(Default value = 0)
Returns
-------
"""
# time domain features
td_features = [
"bpm", "ibi", "sdnn", "sdsd", "rmssd", "pnn20", "pnn50", "hr_mad",
"sd1", "sd2", "s", "sd1/sd2", "breathingrate"
]
# frequency domain features
fd_features = ["lf", "hf", "lf/hf"]
try:
wd, m = hp.process(data_sample, sample_rate, calc_freq=True)
except Exception as e:
try:
wd, m = hp.process(data_sample, sample_rate)
except:
time_domain_features = {k: np.nan for k in td_features}
frequency_domain_features = {k: np.nan for k in fd_features}
return time_domain_features, frequency_domain_features
if rpeak_detector in [1, 2, 3, 4]:
detector = PeakDetector(wave_type='ecg')
peak_list = \
detector.ppg_detector(data_sample, rpeak_detector, preprocess=False)[0]
wd["peaklist"] = peak_list
wd = calc_rr(peak_list, sample_rate, working_data=wd)
wd = check_peaks(wd['RR_list'],
wd['peaklist'],
wd['ybeat'],
reject_segmentwise=False,
working_data=wd)
wd = clean_rr_intervals(working_data=wd)
rr_diff = wd['RR_list']
rr_sqdiff = np.power(rr_diff, 2)
wd, m = calc_ts_measures(wd['RR_list'],
rr_diff,
rr_sqdiff,
working_data=wd)
m = calc_poincare(wd['RR_list'],
wd['RR_masklist'],
measures=m,
working_data=wd)
try:
measures, working_data = calc_breathing(wd['RR_list_cor'],
data_sample,
sample_rate,
measures=m,
working_data=wd)
except:
measures['breathingrate'] = np.nan
wd, m = calc_fd_measures(measures=measures, working_data=working_data)
time_domain_features = {k: m[k] for k in td_features}
frequency_domain_features = {}
for k in fd_features:
if k in m.keys():
frequency_domain_features[k] = m[k]
else:
frequency_domain_features[k] = np.nan
# frequency_domain_features = {k:m[k] for k in fd_features if k in m.keys}
# frequency_domain_features = {k:np.na for k in fd_features if k not in m.keys}
return time_domain_features, frequency_domain_features
def ectopic_sqi(
data_sample,
rule_index=0,
sample_rate=100,
rpeak_detector=0,
wave_type='ppg',
low_rri=300,
high_rri=2000,
):
"""
Evaluate the invalid peaks (which exceeds normal range)
base on HRV rules: Malik, Karlsson, Kamath, Acar
Output the ratio of invalid
Parameters
----------
data_sample :
rule_index:
0: Default Outlier Peak
1: Malik
2: Karlsson
3: Kamath
4: Acar
(Default rule is Malik)
sample_rate :
(Default value = 100)
rpeak_detector :
(Default value = 0)
To explain other detector options
low_rri :
(Default value = 300)
high_rri :
(Default value = 2000)
Returns
-------
"""
rules = ["malik", "karlsson", "kamath", "acar"]
try:
wd, m = hp.process(data_sample, sample_rate, calc_freq=True)
except:
try:
wd, m = hp.process(data_sample, sample_rate)
except:
error_dict = {rule + "_error": np.nan for rule in rules}
error_dict["outlier_error"] = np.nan
return error_dict
# if rpeak_detector in [1, 2, 3, 4]:
if wave_type == 'ecg':
detector = PeakDetector(wave_type='ecg')
peak_list = detector.ecg_detector(data_sample, rpeak_detector)[0]
else:
detector = PeakDetector(wave_type='ppg')
peak_list = detector.ppg_detector(data_sample,
rpeak_detector,
preprocess=False)[0]
wd["peaklist"] = peak_list
wd = calc_rr(peak_list, sample_rate, working_data=wd)
wd = check_peaks(wd['RR_list'],
wd['peaklist'],
wd['ybeat'],
reject_segmentwise=False,
working_data=wd)
wd = clean_rr_intervals(working_data=wd)
rr_intervals = wd["RR_list"]
rr_intervals_cleaned = remove_outliers(rr_intervals,
low_rri=low_rri,
high_rri=high_rri)
number_outliers = len(np.where(np.isnan(rr_intervals_cleaned))[0])
outlier_ratio = number_outliers / (len(rr_intervals_cleaned) -
number_outliers)
if rule_index == 0:
return outlier_ratio
# error_sqi = {}
# error_sqi['outlier_error'] = outlier_ratio
interpolated_rr_intervals = interpolate_nan_values(rr_intervals_cleaned)
rule = rules[rule_index]
nn_intervals = remove_ectopic_beats(interpolated_rr_intervals, method=rule)
number_ectopics = len(np.where(np.isnan(nn_intervals))[0])
ectopic_ratio = number_ectopics / (len(nn_intervals) - number_ectopics)
return ectopic_ratio
def segment_PPG_SQI_extraction(signal_segment,
sampling_rate=100,
primary_peakdet=7,
secondary_peakdet=6,
hp_cutoff_order=(1, 1),
lp_cutoff_order=(20, 4),
template_type=1):
"""
Extract all package available SQIs from a single segment of PPG waveform.
Return a dataframe with all SQIs and cut points for each segment.
Parameters
----------
signal_segment : array-like
A segment of raw signal. The length is user defined in compute_SQI()
function.
sampling_rate : int
Sampling rate of the signal
primary_peakdet : int
Selects one of the peakdetectors from the PeakDetector class. The primary
one is used to segment the waveform.
secondary_peakdet : int
Selects one of the peakdetectors from the PeakDetector class. The
secondary peakdetector is used to compute MSQ SQI.
hp_cutoff_order : touple (int, int)
A high pass filter parameters, cutoff frequency and order
Lp_cutoff_order : touple (int, int)
A low pass filter parameters, cutoff frequency and order
template_type : int
Selects which template from the dtw SQI should be used
Returns
-------
Pandas series object with all SQIs for the given segment
"""
raw_segment = signal_segment[signal_segment.columns[1]].to_numpy()
# Prepare final dictonary that will be converted to dataFrame at the end
SQI_dict = {
'first': signal_segment['idx'][0],
'last': signal_segment['idx'][-1]
}
# Prepare filter and filter signal
filt = BandpassFilter(band_type='butter', fs=sampling_rate)
filtered_segment = filt.signal_highpass_filter(raw_segment,
cutoff=hp_cutoff_order[0],
order=hp_cutoff_order[1])
filtered_segment = filt.signal_lowpass_filter(filtered_segment,
cutoff=lp_cutoff_order[0],
order=lp_cutoff_order[1])
# Prepare primary peak detector and perform peak detection
detector = PeakDetector()
peak_list, trough_list = detector.ppg_detector(filtered_segment,
primary_peakdet)
# Helpful lists for iteration
variations_stats = ['', '_mean', '_median', '_std']
variations_acf = [
'_peak1', '_peak2', '_peak3', '_value1', '_value2', '_value3'
]
stats_functions = [('skewness', sq.standard_sqi.skewness_sqi),
('kurtosis', sq.standard_sqi.kurtosis_sqi),
('entropy', sq.standard_sqi.entropy_sqi)]
# Raw signal SQI computation
SQI_dict['snr'] = np.mean(sq.standard_sqi.signal_to_noise_sqi(raw_segment))
SQI_dict['perfusion'] = sq.standard_sqi.perfusion_sqi(y=filtered_segment,
x=raw_segment)
SQI_dict['mean_cross'] = sq.standard_sqi.mean_crossing_rate_sqi(
raw_segment)
# Filtered signal SQI computation
SQI_dict['zero_cross'] = sq.standard_sqi.zero_crossings_rate_sqi(
filtered_segment)
SQI_dict['msq'] = sq.standard_sqi.msq_sqi(y=filtered_segment,
peaks_1=peak_list,
peak_detect2=secondary_peakdet)
correlogram_list = sq.rpeaks_sqi.correlogram_sqi(filtered_segment)
for idx, variations in enumerate(variations_acf):
SQI_dict['correlogram' + variations] = correlogram_list[idx]
# Per beat SQI calculation
dtw_list = sq.standard_sqi.per_beat_sqi(sqi_func=sq.dtw_sqi,
troughs=trough_list,
signal=filtered_segment,
taper=True,
template_type=template_type)
SQI_dict['dtw_mean'] = np.mean(dtw_list)
SQI_dict['dtw_std'] = np.std(dtw_list)
for funcion in stats_functions:
SQI_dict[funcion[0] +
variations_stats[0]] = funcion[1](filtered_segment)
statSQI_list = sq.standard_sqi.per_beat_sqi(sqi_func=funcion[1],
troughs=trough_list,
signal=filtered_segment,
taper=True)
SQI_dict[funcion[0] + variations_stats[1]] = np.mean(statSQI_list)
SQI_dict[funcion[0] + variations_stats[2]] = np.median(statSQI_list)
SQI_dict[funcion[0] + variations_stats[3]] = np.std(statSQI_list)
#
return pd.Series(SQI_dict)
filter_type=filter_type, sampling_rate=sampling_rate))
print(ppg_data.signals.shape)
s = np.arange(0, 1000, 1)
fig, ax = plt.subplots()
ax.plot(s, ppg_data.signals[0][8000:9000])
plt.show()
ppg_data.update_segment_indices(sqi_sg.generate_segment_idx(segment_length=segment_length, sampling_rate=sampling_rate,
signal_array=ppg_data.signals))
print(ppg_data.segments.shape)
print(ppg_data.segments)
detector = PeakDetector()
peak_list, trough_list = detector.ppg_detector(ppg_data.signals[0][ppg_data.segments[0][0]:ppg_data.segments[0][1]], 7)
# Plot results of peak detection
s = np.arange(0, 3000, 1)
fig, ax = plt.subplots()
ax.plot(s, ppg_data.signals[0][ppg_data.segments[0][0]:ppg_data.segments[0][1]])
if len(peak_list) != 0:
ax.scatter(peak_list, ppg_data.signals[0][peak_list], color="r", marker="v")
if len(trough_list) != 0:
ax.scatter(trough_list, ppg_data.signals[0][trough_list], color="b", marker="v")
plt.show()
# Plot a single period
fig, ax = plt.subplots()
ax.plot(ppg_data.signals[0][trough_list[0]:trough_list[1]])
plt.show()