Abstract 19755: A More Accurate Automatic Cuff Blood Pressure Measurement Method
Most automatic cuff blood pressure (BP) measurement devices are based on oscillometry. These devices compute BP from the envelope of the oscillometric cuff pressure waveform using fixed-ratios. The fixed-ratio values are derived from population averages. As a result, the devices are often inaccurate in patients with stiff arteries. Our hypothesis is that oscillometry can be made more accurate in this important patient population by physical modeling of the phenomenon. We studied 126 patients (ages 38-85, 97 males) referred for diagnostic cardiac catheterization at the Taipei Veterans General Hospital under IRB approval. We obtained oscillometric cuff pressure waveforms for analysis using a commercial device (Omron VP1000 or Microlife WatchBP Office) and reference BP waveforms via an invasive catheter in the opposite brachial artery. We recorded these waveforms before and after sublingual nitroglycerin in many of the patients. We divided the data into a training set comprising 57 of the patient records and a testing set consisting of the remaining 69 patient records. We used the training dataset to develop a method to compute BP from the oscillometric cuff pressure waveform based on a physical model. In contrast to existing devices, the model-based method simultaneously determines BP and the arterial stiffness of the patient. In this way, the method is patient-specific rather than population-based. We used the testing dataset to evaluate the new method against the reference BP levels as well as to compare its accuracy to the commercial device. The Table illustrates the resulting BP errors. The model-based method showed, on average, 28% lower precision errors than the commercial device (statistically significant via Pitman-Morgan test with Bonferroni correction for multiple comparisons) and 41% less absolute BP errors exceeding 10 mmHg. In conclusion, the model-based method may afford more accurate automatic cuff BP measurement in patients with stiff arteries.
Author Disclosures: J. Liu: None. H. Cheng: None. C. Chen: None. S. Sung: None. J. Hahn: None. R. Mukkamala: None.
- © 2015 by American Heart Association, Inc.