Innovative continuous non-invasive cuffless blood pressure monitoring based on photoplethysmography technology

Supplementary material

Arterial lines calibration

Arterial lines were calibratedas follows. The soft tubing was connected to a bag containing 0.9% sodium chloride 500 ml without heparin. This fluid bag was then placed in a pressure bag which was inflated to 300 mmHg, pressuring the flush solution. The solution was flushed completely, taking great care to remove all air from the tubing to the transducer set ports. We maintained the bag pressure at 300 mm Hg. The transducer was then placed at the level of the right atrium (phlebostatic axis located in the fourth intercostal space, midaxillary line) and zeroing was performed. A square wave test was then performed as proposed by Gardner RM 1. If there was evidence or suspicion of over/underdamping phenomena or artifacts in the BP wave, attempts were made to correct it. If these abnormalities persisted, the patient was excluded.

An overdamped arterial pressure waveform was recognized by its slurred upstroke, absent dicrotic notch and loss of fine detail. Underdamped pressure waveform was recognized by oscillations above and below the baseline.

  1. Gardner RM (1981). Direct blood pressure measurement-dynamic response requirements. Anesthesiology 54:227-236.

Figure 1. Infrastructure for data collection

Infrastructure layout for data collection allowing the simultaneous recording of photoplethysmography and arterial waveforms. (1) Catheter, (2) Pulse oximetry, (3) Cables, (4) Processing modulestomonitorsignals, (5) Display unit, (6) Networkdatafrom the Intensive Care Unit or Postanesthesia Care Unit, (7) Central network station, (8) Laptop fordata collection (i.e., storing the simultaneously recorded photoplethysmographic and arterial waveforms).

Figure 2. Standard error calculation

Rangelimits for systolic arterial pressure (SAP), mean arterial pressure (MAP) and diastolic arterial pressure (DAP) and SAP, MAP and DAP standard error calculation. For example, we aimed to calculate the error for MAP and we knew that our reference oscillates between [90-93] mmHg during our 10 second frame; therefore, if the estimate provided by the system is 92 mmHg, the error will be zero; conversely, if the estimate is 95 mmHg, the error will be – 2 mmHg. Finally, if the estimate is 89 mmHg, the error will be 1 mmHg. The same process is used to calculate the error of SAP and DAP. Max value SAP: highest value of SAP; Min value SAP: lowest value of SAP; Max value MAP: highest value of MAP; Min value MAP: lowest value of MAP; Max value DAP: highest value of DAP; Min value DAP: lowest value of DAP.

Figure 3. Outline of study. Flow diagram of patients.

Figure 4. BP distributions

Histograms for the recorded values of systolic arterial pressure, mean arterial pressure and diastolic arterial pressure. These samples are fitted to a Gaussan distribution (unbroken lines).

Primary Admission Diagnoses

Diagnoses / Training cohort / Validation cohort
Surgery (%) / 54.09 / 51.1
Gastrointestinal / 28.2 / -
Hepato-biliary / 20.4 / 23.1
Thoracic / 18.7 / 30.8
Neurosurgery / 12.0 / 46.2
Vascular / 10.2 / -
Cardiac / 6.3 / -
Urology / 3.2 / -
Gynaecology / 0.7 / -
Intracraneal hemorrhage (%) / 10.3 / 10.6
Pneumonia (%) / 9.5 / -
Non-cardiogenic pulmonary edema (%) / 3.4 / 10.6
Sepsis (kidney/urinary tract) (%) / 3.2 / 14.9
Bleeding (%) / 2.7 / 4.3
Hemoptisis (%) / 2.5 / -
Obstructive pulmonary desease (%) / 1.7 / -
Pancreatitis (%) / 1.7 / -
Meningitis (%) / 1.3 / -
Seizure (%) / 1.1 / -
Overdose/Poisoning (%) / 1.0 / -
Others (%) / 7.4 / 8.5

1