Blood Pressure Estimation From Photoplethysmography by Considering Intra- and Inter-Subject Variabilities: Guidelines for a Fair Assessment
Carregando...
Citações na Scopus
0
Tipo de produção
article
Data de publicação
2023
Título da Revista
ISSN da Revista
Título do Volume
Editora
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
Autores
COSTA, Thiago Bulhoes Da Silva
LIMA, Daniel Mario De
Citação
IEEE ACCESS, v.11, p.57934-57950, 2023
Resumo
Cardiovascular diseases are the leading causes of death, and blood pressure (BP) monitoring is essential for prevention, diagnosis, assessment, and treatment. Photoplethysmography (PPG) is a low-cost opto-electronic technique for BP measurement that allows the acquisition of a modulated light signal highly correlated with BP. There are several reports of methods to estimate BP from PPG with impressive results; in this study, we demonstrate that the previous results are excessively optimistic because of their train/test split configuration. To manage this limitation, we considered intra- and inter-subject data arrangements and demonstrated how they affect the results of feature-based BP estimation algorithms (i.e., XGBoost, LightGBM, and CatBoost) and signal-based algorithms (i.e., Residual U-Net, ResNet-18, and ResNet-LSTM). Inter-subject configuration performance is inferior to intra-subject configuration performance, regardless of the model. We also showed that, using only demographic attributes (i.e., age, sex, weight, and subject index number), a regression model achieved results comparable to those obtained in an intra-subject scenario.Although limited to a public clinical database, our findings suggest that algorithms that use an intra-subject setting without a calibration strategy may be learning to identify patients and not predict BP.
Palavras-chave
Blood pressure, photoplethysmography, wearables
Referências
- Alian AA, 2014, BEST PRAC RES-CL ANA, V28, P395, DOI 10.1016/j.bpa.2014.08.006
- Allen J, 2007, PHYSIOL MEAS, V28, pR1, DOI 10.1088/0967-3334/28/3/R01
- Athaya T, 2021, SENSORS-BASEL, V21, DOI 10.3390/s21051867
- Brophy E., 2021, SENSORS-BASEL, V21, P1
- Chen TQ, 2016, KDD'16: PROCEEDINGS OF THE 22ND ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, P785, DOI 10.1145/2939672.2939785
- Chowdhury MH, 2020, SENSORS-BASEL, V20, DOI 10.3390/s20113127
- Dhillon MS, 2019, HDB CUFFLESS BLOOD P, P43, DOI [10.1007/978-3-030-24701-0_5, DOI 10.1007/978-3-030-24701-0_5]
- Dias F. M., 2022, P COMPUT CARDIOL, V498, P1
- Drawz P. E., HYPERTENSION
- El-Hajj C, 2021, BIOMED SIGNAL PROCES, V70, DOI 10.1016/j.bspc.2021.102984
- El-Hajj C, 2021, BIOMED SIGNAL PROCES, V65, DOI 10.1016/j.bspc.2020.102301
- El-Hajj C, 2020, BIOMED SIGNAL PROCES, V58, DOI 10.1016/j.bspc.2020.101870
- Elgendi M, 2019, NPJ DIGIT MED, V2, DOI 10.1038/s41746-019-0136-7
- Elgendi M, 2012, CURR CARDIOL REV, V8, P14, DOI 10.2174/157340312801215782
- Friedman JH, 2001, ANN STAT, V29, P1189, DOI 10.1214/aos/1013203451
- Goldberger AL, 2000, CIRCULATION, V101, pE215, DOI 10.1161/01.CIR.101.23.e215
- Hasanzadeh N, 2020, IEEE SENS J, V20, P4300, DOI 10.1109/JSEN.2019.2961411
- He KM, 2016, PROC CVPR IEEE, P770, DOI 10.1109/CVPR.2016.90
- Hochreiter S, 1997, NEURAL COMPUT, V9, P1735, DOI [10.1162/neco.1997.9.8.1735, 10.1162/neco.1997.9.1.1, 10.1007/978-3-642-24797-2]
- Hsu YC, 2020, SENSORS-BASEL, V20, DOI 10.3390/s20195668
- Huybrechts SAM, 2011, J HYPERTENS, V29, P1577, DOI 10.1097/HJH.0b013e3283487841
- Ibtehaz N, 2020, Arxiv, DOI arXiv:2005.01669
- Johnson AEW, 2016, SCI DATA, V3, DOI 10.1038/sdata.2016.35
- Kachuee M., 2015, UCI MACHINE LEARNING
- Kachuee M, 2015, IEEE INT SYMP CIRC S, P1006, DOI 10.1109/ISCAS.2015.7168806
- Kachuee M, 2017, IEEE T BIO-MED ENG, V64, P859, DOI 10.1109/TBME.2016.2580904
- Kario K, 2019, CIRC RES, V124, P990, DOI 10.1161/CIRCRESAHA.118.313219
- Ke GL, 2017, ADV NEUR IN, V30
- Khalid SG, 2018, J HEALTHC ENG, V2018, DOI 10.1155/2018/1548647
- Kim DK, 2022, IEEE J BIOMED HEALTH, V26, P3697, DOI 10.1109/JBHI.2022.3172514
- Le T, 2020, IEEE ACCESS, V8, P212478, DOI 10.1109/ACCESS.2020.3040257
- Leitner J, 2022, IEEE J BIOMED HEALTH, V26, P218, DOI 10.1109/JBHI.2021.3085526
- Lewington S, 2002, LANCET, V360, P1903, DOI 10.1016/S0140-6736(02)11911-8
- Liang YB, 2019, J CLIN MED, V8, DOI 10.3390/jcm8010012
- Liang YB, 2018, SCI DATA, V5, DOI 10.1038/sdata.2018.76
- Lin WH, 2020, PHYSIOL MEAS, V41, DOI 10.1088/1361-6579/ab7d78
- Lozano R, 2013, LANCET, V381, P628
- Mahmud S., 2022, SENSORS-BASEL, V22, P1
- MANCIA G, 1983, CIRC RES, V53, P96, DOI 10.1161/01.RES.53.1.96
- Mark RG, 2006, COMPUT CARDIOL, V33, P13
- Martinez G, 2018, J CLIN MED, V7, DOI 10.3390/jcm7100316
- Martinez-Rios E, 2021, BIOMED SIGNAL PROCES, V68, DOI 10.1016/j.bspc.2021.102813
- McCombie Devin B, 2006, Conf Proc IEEE Eng Med Biol Soc, V2006, P3521
- Messerli FH, 2007, LANCET, V370, P591, DOI 10.1016/S0140-6736(07)61299-9
- Moody G., 2020, MIMIC 3 WAVEFORM DA
- Moody GB, 1996, COMPUT CARDIOL, P657, DOI 10.1109/CIC.1996.542622
- Mousavi SS, 2019, BIOMED SIGNAL PROCES, V47, P196, DOI 10.1016/j.bspc.2018.08.022
- Mukkamala R, 2021, HYPERTENSION, V78, P1161, DOI 10.1161/HYPERTENSIONAHA.121.17747
- Mukkamala R, 2015, IEEE T BIO-MED ENG, V62, P1879, DOI 10.1109/TBME.2015.2441951
- Nichols WW, 2011, MCDONALD'S BLOOD FLOW IN ARTERIES: THEORETICAL, EXPERIMENTAL AND CLINICAL PRINCIPLES, 6TH EDITION, P1
- Nitzan M, 2014, MED DEVICES-EVID RES, V7, P231, DOI 10.2147/MDER.S47319
- O'Brien E., 1993, J HYPERTENS, V11, pS43
- Ogedegbe G, 2010, CARDIOL CLIN, V28, P571, DOI 10.1016/j.ccl.2010.07.006
- Panwar M, 2020, IEEE SENS J, V20, P10000, DOI 10.1109/JSEN.2020.2990864
- Parati G., 2019, HDB CUFFLESS BLOOD P, P9
- Parati G, 2014, J HYPERTENS, V32, P1359, DOI 10.1097/HJH.0000000000000221
- Parati G, 2013, NAT REV CARDIOL, V10, P143, DOI 10.1038/nrcardio.2013.1
- Paviglianiti A, 2022, COGN COMPUT, V14, P1689, DOI 10.1007/s12559-021-09910-0
- Prokhorenkova L, 2018, ADV NEUR IN, V31
- Rabi DM, 2020, CAN J CARDIOL, V36, P596, DOI 10.1016/j.cjca.2020.02.086
- Rhee S, 2001, IEEE T BIO-MED ENG, V48, P795, DOI 10.1109/10.930904
- Rong M, 2021, BIOMED SIGNAL PROCES, V68, DOI 10.1016/j.bspc.2021.102772
- Ronneberger O, 2015, LECT NOTES COMPUT SC, V9351, P234, DOI 10.1007/978-3-319-24574-4_28
- Saeed M, 2002, COMPUT CARDIOL, V29, P641, DOI 10.1109/CIC.2002.1166854
- Saeed M, 2011, CRIT CARE MED, V39, P952, DOI 10.1097/CCM.0b013e31820a92c6
- Saugel B, 2014, BEST PRACT RES-CLIN, V28, P309, DOI 10.1016/j.bpa.2014.08.001
- Schlesinger O, 2020, INT CONF ACOUST SPEE, P1135, DOI [10.1109/ICASSP40776.2020.9053446, 10.1109/icassp40776.2020.9053446]
- Schrumpf F, 2021, SENSORS-BASEL, V21, DOI 10.3390/s21186022
- Segers P, 2009, ARTERY RES, V3, P79, DOI 10.1016/j.artres.2009.02.006
- Sharma M, 2017, TECHNOLOGIES, V5, DOI 10.3390/technologies5020021
- Siddique N, 2021, IEEE ACCESS, V9, P82031, DOI 10.1109/ACCESS.2021.3086020
- Slapnicar G, 2019, SENSORS-BASEL, V19, DOI 10.3390/s19153420
- Stevens SL, 2016, BMJ-BRIT MED J, V354, DOI 10.1136/bmj.i4098
- Tamura T, 2014, ELECTRONICS-SWITZ, V3, P282, DOI 10.3390/electronics3020282
- Teng XF, 2003, P ANN INT IEEE EMBS, V25, P3153, DOI 10.1109/IEMBS.2003.1280811
- Verberk W. J., 2019, HDB CUFFLESS BLOOD P, P203
- Wang GX, 2018, IEEE CIRC SYST MAG, V18, P6, DOI 10.1109/MCAS.2018.2849261
- Wang LD, 2018, J HEALTHC ENG, V2018, DOI 10.1155/2018/7804243
- Wang WN, 2022, IEEE J BIOMED HEALTH, V26, P2075, DOI 10.1109/JBHI.2021.3128383
- Whelton Paul K, 2018, J Am Soc Hypertens, V12, DOI [10.1161/HYP.0000000000000065, 10.1016/j.jash.2018.06.010]
- WHITE WB, 1993, HYPERTENSION, V21, P504, DOI 10.1161/01.HYP.21.4.504
- Xing XM, 2019, SCI REP-UK, V9, DOI 10.1038/s41598-019-45175-2
- Xing XM, 2016, BIOMED OPT EXPRESS, V7, P3007, DOI 10.1364/BOE.7.003007
- Yoon YZ, 2018, IEEE J BIOMED HEALTH, V22, P1068, DOI 10.1109/JBHI.2017.2714674
- Zhang ZX, 2018, IEEE GEOSCI REMOTE S, V15, P749, DOI 10.1109/LGRS.2018.2802944