├── PPG_ACC_plots.PNG ├── README.md ├── clinical_application.ipynb ├── datasets └── troika │ └── training_data │ ├── DATA_01_TYPE01.mat │ ├── DATA_02_TYPE02.mat │ ├── DATA_03_TYPE02.mat │ ├── DATA_04_TYPE01.mat │ ├── DATA_04_TYPE02.mat │ ├── DATA_05_TYPE02.mat │ ├── DATA_06_TYPE02.mat │ ├── DATA_07_TYPE02.mat │ ├── DATA_08_TYPE02.mat │ ├── DATA_10_TYPE02.mat │ ├── DATA_11_TYPE02.mat │ ├── DATA_12_TYPE02.mat │ ├── REF_01_TYPE01.mat │ ├── REF_02_TYPE02.mat │ ├── REF_03_TYPE02.mat │ ├── REF_04_TYPE01.mat │ ├── REF_04_TYPE02.mat │ ├── REF_05_TYPE02.mat │ ├── REF_06_TYPE02.mat │ ├── REF_07_TYPE02.mat │ ├── REF_08_TYPE02.mat │ ├── REF_10_TYPE02.mat │ ├── REF_11_TYPE02.mat │ ├── REF_12_TYPE02.mat │ └── Readme.pdf └── pulse_rate.ipynb /PPG_ACC_plots.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/PPG_ACC_plots.PNG -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Motion Compensated Pulse Rate Estimation 2 | 3 | ![](PPG_ACC_plots.PNG) 4 | 5 | **Summary** 6 | 7 | This repository includes two python notebooks: 8 | - *pulse_rate.ipynb*: Python code for analyzing wrist-worn PPG signal data and estimating a pulse rate while including data from a three-axis accelerometer. The algorithm is design to detect the correct pulse rate in the presence of severe movement artifacts. 9 | - *clinical_application*.ipynb: Heart rate trend investigation using CAST dataset. 10 | 11 | Project completed as part of Udacity AI for Healthcare course (2020). 12 | 13 | **Code Description** 14 | 15 | The Python code may be used in two ways: 16 | 17 | 1) Analyze algorithm performance on [TROIKA](https://ieeexplore.ieee.org/document/6905737) dataset. 18 | 19 | The function *RunPulseRateAlgorithm()* expects two files as input: 20 | - data_fl: .mat file containing PPG and X, Y, Z accelerometer data from Troika dataset (or in Troika format) 21 | - ref_fl: .mat file containing ground truth heart rates from Troika dataset (or in Troika format) 22 | 23 | RunPulseRateAlgorithm() will compute the heart rate and confidence estimates for 8 second windows of the PPG and accelerometer data, and calculate the error (difference) between the heart estimates and the ground truth heart rate for each 8 second window provided in the reference file. 24 | 25 | The function *Evaluate()* will calculate an overall mean absolute error at 90% availability (see Algorithm Performance section below), assuming the Troika data is located at *./datasets/troika/training_data*. 26 | 27 | 2) Predict heart rate from new input PPG and Accelerometer data. 28 | 29 | The function *AnalyzeWindow()* accepts numpy arrays of ppg, and accelerometer data in the x, y, and z axis, along with a sampling rate. A precondition of the function is that the incoming ppg and acceleromter data was sampled at the provided sampling rate and is aligned in time. While the algorithm was only tested with 8 second windows of PPG and accelerometer data and a sampling rate of 125 Hz, the AnalyzeWindow function may be applied to a longer window sizes or alternate sampling rates. AnalyzeWindow() returns a tuple of (BPM prediction, confidence) for the provided window of ppg and accelerometer data using the provided sample rate. 30 | 31 | **Data Description** 32 | 33 | The data used to design and test this algorithm was taken from the [TROIKA](https://ieeexplore.ieee.org/document/6905737) dataset. The dataset includes 12 subjects ages 18 to 35 with simultaneously recorded two-channel wrist-mounted PPG signals, wrist-mounted three-axis accelerometer signals, and one-channel chest-mounted ECG. The subjects ran on a treadmill at increasing rates to range from resting heart rate to intense exercise heart rate. To build a more complete dataset, it would be beneficial to include a greater number of participants that represent the broader population demographics for age and gender. Additionally, it would be useful to record PPG and accelerometer data from more points on the body that could be used in future wearable devices such as the feet, legs, upper arms, and head. This could provide a more complete picture of how accelerometer signals affect PPG data across the body. Alternative exercises could be recorded as well, such as cylcing, swimming, tennis, and weight lifting, each of which may produce different physiological responses or signal patterns. 34 | 35 | **Algorithm Description** 36 | 37 | The heart rate prediction algorithm takes advantage of the physiology of blood flow through the ventricles of the wrist. Light emitted by an LED on the PPG sensor is reflected less when the ventricles contract and more blood is present, and light is reflected more when the blood returns to the heart and fewer red blood cells are present. Arm movement will also affect the blood levels in the wrist, so periodic motion such as an arm swinging back and forth can also be detected by the PPG signal. 38 | 39 | The algorithm identifies the strongest frequency components of both the PPG signal and the accelerometer signals to determine which frequency to pick as the heart rate. The algorithm follows these stages: 40 | 41 | 1) Apply bandpass filter to PPG and accelerometer signals to filter out frequencies outside of the 40-240 BPM range. 42 | 43 | 2) Aggregate the X, Y, and Z channels of the accelerometer signal into a signal magnitude signal 44 | 45 | 2) Tranform the time domain PPG and accelerometer signal to magnitude frequency representations by taking the absolute value of their Fast Fourier Transforms. 46 | 47 | 3) Using the frequency representations of PPG and accelerometer signals, find the peaks with the largest magnitudes, and choose one to be the predicted heart rate frequency. 48 | 49 | - If the highest magnitude peak of both signals is different, choose the highest magnitude peak of the the PPG signal as the heart rate frequency prediction. 50 | - If the highest magnitude peak of both signals is the same, this may mean that the arm swing signal is overpowering the pulse rate, so choose the next highest magnitude peak of the PPG signal as the heart rate frequency prediction. 51 | - If each of the highest magnitude peaks of the PPG signal are too close to the peaks of the accelerometer signal, the arm swing frequency could be the same as the pulse rate frequency, so use the highest magnitude peak of the PPG as the heart rate frequency prediction, even though the accelerometer signal has the same peak). 52 | 53 | 4) Convert the chosen peak frequency to a final **BPM Prediction**, and calculate a **Confidence Value** for the chosen frequency by computing the ratio of energy concentrated near that frequency compared to the full signal. 54 | 55 | The BPM Prediction and Confidence Value outputs from this algorithm are not gauranteed to be correct. Confidence values are only used to determine which outputs are very poor, i.e. a low confidence value implies very low signal to noise ratio, since only a small amount of energy is concentrated by the predicted peak. High confidence values do not necessarily imply the that algorithm is significantly more correct, only that the peak at that location is responsible for much more of the signal. Common failure modes include when accelerometer data has random movement spikes or when the three channels combine into a non-periodic signal. 56 | 57 | **Algorithm Performance** 58 | 59 | The algorithm performance was evaluated against the TROIKA reference data. All PPG data was compared against ECG "ground truth" BPM for associated time windows. ECG measures electric potentials across the heart and is considered to be a much more reliable method for obtaining pulse rate than PPG since these electrical signals are not susceptible to the same levels of movement-related noise. Using the confidence estimates to compare prediction quality, the bottom 10% of predictions were discarded while the remaining predictions were evaluated against the ground truth. The final calculated performance metric for the dataset at 90% availability was a mean absolute error (**MAE**) of **13.625 BPM**. This performance may be verified by executing the *Evaluate()* function. 60 | 61 | This algorithm may not perform as well at >90% availability, and does not gaurantee that consecutive or overlapping time windows have similar confidence values, so it is possible that prediction confidence could vary greatly over time if this algorithm was applied to a real-time heart rate analysis scenario. This algorithm also assumes that accelerometer motion will largely result from swinging arms during running on a treadmill and will therefore be consistently periodic. This may not be the case for other activities, such as tennis or basketball. 62 | -------------------------------------------------------------------------------- /clinical_application.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Part 2: Clinical Application\n", 8 | "\n", 9 | "Using the CAST dataset, investigate the trend that average resting ehart rate increase up until middle age then decrease into old age.\n", 10 | "\n", 11 | "### Dataset (CAST)\n", 12 | "\n", 13 | "The data from this project comes from the [Cardiac Arrythmia Suppression Trial (CAST)](https://physionet.org/content/crisdb/1.0.0/), which was sponsored by the National Heart, Lung, and Blood Institute (NHLBI). CAST collected 24 hours of heart rate data from ECGs from people who have had a myocardial infarction (MI) within the past two years.[1] This data has been smoothed and resampled to more closely resemble PPG-derived pulse rate data from a wrist wearable.[2]\n", 14 | "\n", 15 | "1. **CAST RR Interval Sub-Study Database Citation** - Stein PK, Domitrovich PP, Kleiger RE, Schechtman KB, Rottman JN. Clinical and demographic determinants of heart rate variability in patients post myocardial infarction: insights from the Cardiac Arrhythmia Suppression Trial (CAST). Clin Cardiol 23(3):187-94; 2000 (Mar)\n", 16 | "2. **Physionet Citation** - Goldberger AL, Amaral LAN, Glass L, Hausdorff JM, Ivanov PCh, Mark RG, Mietus JE, Moody GB, Peng C-K, Stanley HE. PhysioBank, PhysioToolkit, and PhysioNet: Components of a New Research Resource for Complex Physiologic Signals (2003). Circulation. 101(23):e215-e220.\n", 17 | "\n", 18 | "-----" 19 | ] 20 | }, 21 | { 22 | "cell_type": "markdown", 23 | "metadata": {}, 24 | "source": [ 25 | "### Code\n", 26 | "#### Imports" 27 | ] 28 | }, 29 | { 30 | "cell_type": "code", 31 | "execution_count": 3, 32 | "metadata": {}, 33 | "outputs": [], 34 | "source": [ 35 | "import glob\n", 36 | "import os\n", 37 | "\n", 38 | "import numpy as np\n", 39 | "import pandas as pd" 40 | ] 41 | }, 42 | { 43 | "cell_type": "markdown", 44 | "metadata": {}, 45 | "source": [ 46 | "#### Load the dataset\n", 47 | "\n", 48 | "The dataset is stored as [.npz](https://docs.scipy.org/doc/numpy/reference/generated/numpy.savez.html) files. Each file contains roughly 24 hours of heart rate data in the 'hr' array sampled at 1Hz. The subject ID is the name of the file. You will use these files to compute resting heart rate.\n", 49 | "\n", 50 | "Demographics metadata is stored in a file called 'metadata.csv'. This CSV has three columns, one for subject ID, age group, and sex. You will use this file to make the association between resting heart rate and age group for each gender.\n", 51 | "\n", 52 | "Find the dataset in `../datasets/crisdb/`" 53 | ] 54 | }, 55 | { 56 | "cell_type": "code", 57 | "execution_count": 4, 58 | "metadata": {}, 59 | "outputs": [], 60 | "source": [ 61 | "hr_filenames = glob.glob('/data/crisdb/*.npz')" 62 | ] 63 | }, 64 | { 65 | "cell_type": "markdown", 66 | "metadata": {}, 67 | "source": [ 68 | "#### Load Metadata\n", 69 | "Load the metadata file into a datastructure that allows for easy lookups from subject ID to age group and sex." 70 | ] 71 | }, 72 | { 73 | "cell_type": "code", 74 | "execution_count": 5, 75 | "metadata": {}, 76 | "outputs": [], 77 | "source": [ 78 | "metadata_filename = '/data/crisdb/metadata.csv'\n", 79 | "\n", 80 | "with open(metadata_filename, 'r') as f:\n", 81 | " metadata = pd.read_csv(f)" 82 | ] 83 | }, 84 | { 85 | "cell_type": "code", 86 | "execution_count": 6, 87 | "metadata": {}, 88 | "outputs": [ 89 | { 90 | "data": { 91 | "text/html": [ 92 | "
\n", 93 | "\n", 106 | "\n", 107 | " \n", 108 | " \n", 109 | " \n", 110 | " \n", 111 | " \n", 112 | " \n", 113 | " \n", 114 | " \n", 115 | " \n", 116 | " \n", 117 | " \n", 118 | " \n", 119 | " \n", 120 | " \n", 121 | " \n", 122 | " \n", 123 | " \n", 124 | " \n", 125 | " \n", 126 | " \n", 127 | " \n", 128 | " \n", 129 | " \n", 130 | " \n", 131 | " \n", 132 | " \n", 133 | " \n", 134 | " \n", 135 | " \n", 136 | " \n", 137 | " \n", 138 | " \n", 139 | " \n", 140 | " \n", 141 | " \n", 142 | " \n", 143 | " \n", 144 | " \n", 145 | " \n", 146 | " \n", 147 | "
subjectagesex
0e198a20-24Male
1e198b20-24Male
2e028b30-34Male
3e028a30-34Male
4e061b30-34Male
\n", 148 | "
" 149 | ], 150 | "text/plain": [ 151 | " subject age sex\n", 152 | "0 e198a 20-24 Male\n", 153 | "1 e198b 20-24 Male\n", 154 | "2 e028b 30-34 Male\n", 155 | "3 e028a 30-34 Male\n", 156 | "4 e061b 30-34 Male" 157 | ] 158 | }, 159 | "execution_count": 6, 160 | "metadata": {}, 161 | "output_type": "execute_result" 162 | } 163 | ], 164 | "source": [ 165 | "metadata.head(5)" 166 | ] 167 | }, 168 | { 169 | "cell_type": "markdown", 170 | "metadata": {}, 171 | "source": [ 172 | "#### Compute Resting Heart Rate\n", 173 | "For each subject we want to compute the resting heart rate while keeping track of which age group this subject belongs to. An easy, robust way to compute the resting heart rate is to use the lowest 5th percentile value in the heart rate timeseries." 174 | ] 175 | }, 176 | { 177 | "cell_type": "code", 178 | "execution_count": 19, 179 | "metadata": {}, 180 | "outputs": [], 181 | "source": [ 182 | "def AgeAndRHR(metadata, filename):\n", 183 | "\n", 184 | " # Load the heart rate timeseries\n", 185 | " hr_data = np.load(filename)['hr']\n", 186 | " \n", 187 | " # Compute the resting heart rate from the timeseries by finding the lowest 5th percentile value in hr_data\n", 188 | " rhr = np.percentile(hr_data, 5)\n", 189 | "\n", 190 | " # Find the subject ID from the filename.\n", 191 | " subject = filename.rsplit('/')[-1].rstrip('.npz')\n", 192 | "\n", 193 | " # Find the age group for this subject in metadata.\n", 194 | " age_group = metadata.loc[metadata['subject'] == subject]['age'].values[0]\n", 195 | " \n", 196 | " # Find the sex for this subject in metadata.\n", 197 | " sex = metadata.loc[metadata['subject'] == subject]['sex'].values[0]\n", 198 | "\n", 199 | " return age_group, sex, rhr\n", 200 | "\n", 201 | "df = pd.DataFrame(data=[AgeAndRHR(metadata, filename) for filename in hr_filenames],\n", 202 | " columns=['age_group', 'sex', 'rhr'])" 203 | ] 204 | }, 205 | { 206 | "cell_type": "code", 207 | "execution_count": 20, 208 | "metadata": {}, 209 | "outputs": [ 210 | { 211 | "data": { 212 | "text/html": [ 213 | "
\n", 214 | "\n", 227 | "\n", 228 | " \n", 229 | " \n", 230 | " \n", 231 | " \n", 232 | " \n", 233 | " \n", 234 | " \n", 235 | " \n", 236 | " \n", 237 | " \n", 238 | " \n", 239 | " \n", 240 | " \n", 241 | " \n", 242 | " \n", 243 | " \n", 244 | " \n", 245 | " \n", 246 | " \n", 247 | " \n", 248 | " \n", 249 | " \n", 250 | " \n", 251 | " \n", 252 | " \n", 253 | " \n", 254 | " \n", 255 | " \n", 256 | " \n", 257 | " \n", 258 | " \n", 259 | " \n", 260 | " \n", 261 | " \n", 262 | " \n", 263 | " \n", 264 | " \n", 265 | " \n", 266 | " \n", 267 | " \n", 268 | "
age_groupsexrhr
060-64Female89.302326
165-69Male65.641026
245-49Male56.928458
360-64Male68.571429
450-54Male84.395604
\n", 269 | "
" 270 | ], 271 | "text/plain": [ 272 | " age_group sex rhr\n", 273 | "0 60-64 Female 89.302326\n", 274 | "1 65-69 Male 65.641026\n", 275 | "2 45-49 Male 56.928458\n", 276 | "3 60-64 Male 68.571429\n", 277 | "4 50-54 Male 84.395604" 278 | ] 279 | }, 280 | "execution_count": 20, 281 | "metadata": {}, 282 | "output_type": "execute_result" 283 | } 284 | ], 285 | "source": [ 286 | "df.head(5)" 287 | ] 288 | }, 289 | { 290 | "cell_type": "markdown", 291 | "metadata": {}, 292 | "source": [ 293 | "#### Plot Resting Heart Rate vs. Age Group\n", 294 | "We'll use [seaborn](https://seaborn.pydata.org/) to plot the relationship. Seaborn is a thin wrapper around matplotlib that enables higher-level statistical plots.\n", 295 | "\n", 296 | "We will use [lineplot](https://seaborn.pydata.org/generated/seaborn.lineplot.html#seaborn.lineplot) to plot the mean of the resting heart rates for each age group along with the 95% confidence interval around the mean. More about making plots that show uncertainty [here](https://seaborn.pydata.org/tutorial/relational.html#aggregation-and-representing-uncertainty)." 297 | ] 298 | }, 299 | { 300 | "cell_type": "code", 301 | "execution_count": null, 302 | "metadata": {}, 303 | "outputs": [], 304 | "source": [ 305 | "# !pip install seaborn==0.9.0" 306 | ] 307 | }, 308 | { 309 | "cell_type": "code", 310 | "execution_count": 9, 311 | "metadata": {}, 312 | "outputs": [ 313 | { 314 | "data": { 315 | "image/png": "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\n", 316 | "text/plain": [ 317 | "" 318 | ] 319 | }, 320 | "metadata": { 321 | "needs_background": "light" 322 | }, 323 | "output_type": "display_data" 324 | } 325 | ], 326 | "source": [ 327 | "import seaborn as sns\n", 328 | "from matplotlib import pyplot as plt\n", 329 | "\n", 330 | "labels = sorted(np.unique(df.age_group))\n", 331 | "df['xaxis'] = df.age_group.map(lambda x: labels.index(x)).astype('float')\n", 332 | "plt.figure(figsize=(12, 8))\n", 333 | "sns.lineplot(x='xaxis', y='rhr', hue='sex', data=df)\n", 334 | "_ = plt.xticks(np.arange(len(labels)), labels)" 335 | ] 336 | }, 337 | { 338 | "cell_type": "markdown", 339 | "metadata": {}, 340 | "source": [ 341 | "### Clinical Conclusion\n", 342 | "\n", 343 | "For females, we see a notable increase in resting heart rate from ages 35-39 to age 40-44, then a slight decline from 40-44 to 55-59 before a steeper decline from 55-59 to ages higher than 70. We also see a larger confidence interval from ages 40-44 to 55-59, indicating that female resting heart ranges could easily be around 10 RHR higher or lower than average.\n", 344 | "\n", 345 | "For males, we see a minor increase in heart rate around age 45-49, and then slight decline into old age.\n", 346 | "\n", 347 | "Compared to males, females have a higher average resting heart rate from ages 40-44 to 65-69. Additionally, females have a much wider confidence interval around their average heart rate values compared to men. The trend of increasing heart rate into middle age before decreasing heart rate into old age is much more visible in women.\n", 348 | "\n", 349 | "There may be several possible reasons these results. First, women are generally expected to have a higher heart rate than men because females hearts are typically smaller than male hearts and must beat faster to provide similar output of blood per beat. Women also have a slightly faster physiological heart rhythm, causing higher heart rates ([journal source](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4190707/)). Menopausal women may also experience changes in heart rate. \n", 350 | "\n", 351 | "It is important to verify that the dataset is representative of the broader population for both men and women. As shown in the bar charts below, females are clearly underrepresented compared to males in the CAST dataset. This could likely be the reason why the confidence interval for women is much wider then it is for men. Additionally, the age distributions for both men and women are both left skewed with most samples taken around middle age (50s to 60s), while younger age groups are underrepresented. This could be expected because the CAST dataset is collected from people who have had a myocardial infarction (MI) within the past two years, which are more common in middle and older aged people. We might improve our results if we chose a larger dataset that is more representative of the broader population age and sex demographics, and compared equal counts of males and females." 352 | ] 353 | }, 354 | { 355 | "cell_type": "code", 356 | "execution_count": 16, 357 | "metadata": {}, 358 | "outputs": [ 359 | { 360 | "data": { 361 | "text/plain": [ 362 | "" 363 | ] 364 | }, 365 | "execution_count": 16, 366 | "metadata": {}, 367 | "output_type": "execute_result" 368 | }, 369 | { 370 | "data": { 371 | "image/png": "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\n", 372 | "text/plain": [ 373 | "" 374 | ] 375 | }, 376 | "metadata": { 377 | "needs_background": "light" 378 | }, 379 | "output_type": "display_data" 380 | } 381 | ], 382 | "source": [ 383 | "figure, axes = plt.subplots(1, 2, figsize=(15,4))\n", 384 | "\n", 385 | "df.groupby('sex').size().plot(kind='bar', ax=axes[0])\n", 386 | "df.groupby(['age_group', 'sex']).size().unstack(level=1).plot(kind='bar', ax=axes[1])" 387 | ] 388 | }, 389 | { 390 | "cell_type": "markdown", 391 | "metadata": {}, 392 | "source": [ 393 | "We were unable to fully validate for both sexes the trend that average resting heart rate increases up until middle age and then decreases into old age. Our dataset did not contain adequate samples for younger ages, so we did not see an increase in heart rate for men although we did see an increase for women between age groups 35-39 and 40-44. For both sexes, we were able to see in the plots that average resting heart rate decreases from middle age into old age. However, the age and sex demographics of the CAST dataset are not sufficient to fully validate the trend." 394 | ] 395 | } 396 | ], 397 | "metadata": { 398 | "kernelspec": { 399 | "display_name": "Python 3", 400 | "language": "python", 401 | "name": "python3" 402 | }, 403 | "language_info": { 404 | "codemirror_mode": { 405 | "name": "ipython", 406 | "version": 3 407 | }, 408 | "file_extension": ".py", 409 | "mimetype": "text/x-python", 410 | "name": "python", 411 | "nbconvert_exporter": "python", 412 | "pygments_lexer": "ipython3", 413 | "version": "3.6.3" 414 | } 415 | }, 416 | "nbformat": 4, 417 | "nbformat_minor": 2 418 | } 419 | -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_01_TYPE01.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_01_TYPE01.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_02_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_02_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_03_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_03_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_04_TYPE01.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_04_TYPE01.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_04_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_04_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_05_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_05_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_06_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_06_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_07_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_07_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_08_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_08_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_10_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_10_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_11_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_11_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/DATA_12_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/DATA_12_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_01_TYPE01.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_01_TYPE01.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_02_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_02_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_03_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_03_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_04_TYPE01.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_04_TYPE01.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_04_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_04_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_05_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_05_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_06_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_06_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_07_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_07_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_08_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_08_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_10_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_10_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_11_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_11_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/REF_12_TYPE02.mat: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/REF_12_TYPE02.mat -------------------------------------------------------------------------------- /datasets/troika/training_data/Readme.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/KJStrand/Pulse_Rate_Estimation/b2b884bde482bd2a15fdbc5170291516442e5867/datasets/troika/training_data/Readme.pdf --------------------------------------------------------------------------------