Patterns of bioelectrical impedance vector analysis: learning from electrocardiography and forgetting electric circuit models
2002; Elsevier BV; Volume: 18; Issue: 6 Linguagem: Inglês
10.1016/s0899-9007(02)00771-2
ISSN1873-1244
Autores Tópico(s)Hemodynamic Monitoring and Therapy
ResumoBoth electrocardiography and bioelectrical impedance analysis (BIA) aim to transform electrical properties of tissues into clinical information. Electrocardiography is based on several established patterns relating electrical measurements to heart disorders. 1,2 Conventional BIA is based on electric models supporting quantitative estimates of body compartments.3–9 In BIA literature, an electrical model (e.g., series, parallel, Cole’s, and Hanai’s models) is used as an electrical equivalent, 1 a circuit that electrically behaves like the original, is expressed through mathematical equations, and represents anatomic structures or physical processes (e.g., 75 trillion cells, three to six body compartments, cellular/vascular fluid shifts, etc.). According to their dictionary definitions, a model is “A representation of the supposed structure of something” and a pattern is “A set of forms used as a guide in making things.” In the clinical setting, operational patterns based on direct laboratory data are more useful than complex, explanatory, or descriptive models of phenomena. The electrocardiogram (ECG) is a graphic recording through surface electrodes (as in BIA) of electric potentials generated by the heart. Cardiac depolarization and repolarization waves are interpreted with few, time-averaged vectors. An ECG wave (vector) is the complex spatial and temporal summation of electric potentials from multiple myocardial fibers conducted to the surface of the body. Abnormalities of individual waveforms are defined with respect to reference values of healthy subjects. An ECG is interpreted with diagnostic ECG patterns, which are a combination of waveforms associated with specific cardiac disorders in clinical validation studies (e.g., bundle branch blocks, myocardial ischemia, infarction, etc.). 2 Calculations of heart volume, ischemic mass volume, infarction volume, etc., from ECG waveforms through electrical models and prediction equations are not used in the clinical setting. 1,2 The National Institutes of Health in 1994 3 and an independent panel of BIA experts in 19974 continued to recommend (in vain) to manufacturers that they provide both measured data (resistance [R], reactance [Xc], impedance [Z], and phase angle) and prediction equations in their software. The expert panel recommended the use of multifrequency BIA (i.e., R extrapolated to zero and infinite frequency) in estimating total body water (TBW; from R ), extracellular water (ECW; from R0), and intracellular water (ICW; TBW ECW) in altered fluid distribution, and a limited use of single-frequency BIA, parallel model, in estimating body cell mass (BCM). Estimation of fat-free mass (FFM) was considered acceptable only with normal fluid status, and derived estimation of fat mass (FM) was considered inappropriate. Single-frequency BIA, series model, was ranked as the least useful BIA technique.4 In 2000, Grimnes and Martinsen 1 in their book reported an extensive update of bioimpedance models and techniques making inconsistent previous recommendations of the expert panels. Relevant discrepancies can be summarized in the following state
Referência(s)