Development and clinical validation of a miniaturized finger probe for bedside hemodynamic monitoring

Figures

• 

  Figure 1. Bland-Altman and correlation plots for SBP, MBP, and DBP The two datasets, auscultated and hydrostatic challenge, are plotted with different markers.

• 

  Figure 2. Instrument overview (A) Photograph showing the device in use. (B) Cross-sectional view of the sensor setup. (C) 3D model of the device. (D) Oscillometric method. The pressure ramp signal is filtered with a pass-band filter (1–10 Hz) and Hilbert transformed. The peaks, or oscillations, are detected, and a polynomial curve is fitted to the peaks. The resulting bell-shaped curve is called an oscillometric waveform envelope (OMWE). Mean arterial pressure (MAP) is found at the point on the pressure curve that corresponds with the maximum of OMWE. DBP is found on the left side of the OMWE at 80% of maximum pressure. SBP is computed from DBP and MAP.

• 

  Figure 3. Results of consecutive measurements and arterial stiffness (A) A demonstration of consecutive measurements taken with the device. Oscillometric measurements (shown as black vertical lines and markers) were taken with the finger device approx every 2 min. The red and light red lines represent the continuous BP taken with the reference device CNAP 500. (B) Example V -Pt plots estimating arterial compliance for a young subject with low carotid-to-femoral PWV (6.3 m/s) and an elderly subject with high PWV (10.6 m/s). The slope of the curve in the young subject’s case is higher, resulting in higher ASI. (C) Figure showing the negative correlation between ASI and PWV.

• 

  Figure 4. Atrial fibrillation detection The difference between the oscillometric envelope and the amplitude of each peak is used to calculate the mean error. The algorithm was able to differentiate AF from sinus rhythm using the error between oscillometric envelope and pulse peaks.

• 

  Figure 5. Description of the system (A) Block diagram of the system. An ARM Cortex M4 based MCU is used as the controller unit. The MCU is connected to a laptop running a Python GUI for data logging. (B) Inner mechanics illustrated in a 3D view. A DC motor is connected to a worm gear assembly through a reduction gear cassette. This rotates a hinge-type press that is used to apply pressure to the finger. A pair of micro limit switches assembled on a custom circuit board is used to limit the motion.