Evidence level: B
“脉象=血管几何”的断言,只有把波形切成基波与各阶谐波、逐条核对能量占比时,才算真正被检验——这正是 1-D 血流模型闭环验证的下一层级。
正文
上一篇文章回答了”血管几何能否决定脉搏波形”——能。1-D 血流模型在几何参数完全实测的条件下,能以 <4% 的压力误差复现 37 条动脉网络的脉搏波形(PMID 17640653)。但”波形长得像”只是表象验证:波形可以在时域上高度相似,却在频域能量分布上系统性地偏离。中医脉诊读的不是波形”像不像”,而是谐波之间的相对能量关系——王唯工的脉搏频谱分析正是把桡动脉压力波做傅里叶分解,用前 10 个谐波的相对振幅(C1-C10)对应脏腑(PMID 31275406)。因此,把血管几何到脉象的因果链推进到”逐谐波能量分布”的闭合检验,才是”血管几何决定脉象”的最强判决实验。
为什么必须做这一层?傅里叶分析的数学性质决定了:时域波形相似 ≠ 频域能量分布一致。两个形状几乎重合的波,可能在第二、第三谐波上有 20%-30% 的能量差异,而这些差异恰恰是脏腑辨证所依赖的指纹。已有临床证据表明,谐波能量确实携带疾病信息:一篇 2019 年研究用 FFT 把 83 例受试者的桡动脉波分解为 10 个谐波,发现冠心病组 C1 升高、C5 降低,把谐波指标并入传统危险因素后,ROC 曲线下面积从 0.788 升到 0.856(PMID 31275406);2026 年的 SYNTAX 评分研究进一步用左右手谐波差异(ΔCn/ΔPn/ΔDn)构建判别模型,预测冠脉病变严重度(PMID 40396324)。谐波能量分布是有生理对应物的信号——那么它能不能被”血管几何”单独预测出来?
1-D 模型的”谱级”验证已有技术底座。患者特异模型已被反复证明能把波形预测到高精度:Reymond 等用 MR 血管造影重建个体动脉树几何,患者特异模型在具体位点显著优于通用模型(PMID 21622820);Willemet 等在下肢旁路术患者身上把 1D-0D 模型与体内实测对照,压力与速度波形的平均相对均方根误差仅 9.6% 与 16.0%(PMID 23701843);2020 年一项 62 人研究用 0D-1D 耦合模型反演个体化参数,个体压力波 MSE 低至 7.1 mmHg²,颈动脉血流速度与超声实测相关系数 0.911(PMID 32142412);2022 年 22 人 EECP 研究同样证实个性化 0D-1D 模型能预测个体血流动力学(PMID 36379202)。但所有这些验证用的都是”波形误差”——把预测波与实测波逐点比对——而不是”谐波能量误差”。 一篇波形与频谱误差不一定同向的隐患,正是本文要补的缺口。
还有一个必须正视的模型自由度问题。1-D 模型要预测谐波能量,输入的几何参数(长度、半径、分支位置、管壁弹性)必须被无创测准。Epstein 等在 2015 年证明:把 55-67 条动脉的分布式模型压缩到 21-37 段,只要保留总阻力与总顺应性,波形形状几乎不变(PMID 25888513)——这说明几何简化的容错度对”波形”高、对”逐谐波能量”未必高,谐波能量对反射位点(胸主动脉、肾动脉)的精确位置更敏感。这也意味着:逐谐波闭合检验不仅验证”血管几何决定脉象”,还反向验证”现有无创几何测量协议是否足够精细”。
可执行的检验设计由此清晰起来:① 同一受试者用超声/MRI 重建主传导动脉几何(长度/半径/分支/弹性);② 用 1-D 模型预测桡动脉压力波并做 FFT 分解;③ 同一位置用压平眼压计/光电脉搏实测波形同样分解;④ 输出一张谐波级误差矩阵——第 1-10 阶谐波的预测相对能量比 vs 实测相对能量比、相位差、以及各谐波能量的归一化均方根误差;⑤ 跨个体重复,报告误差随谐波阶次/血管段数/几何精度的衰减曲线。两种结局都是知识:若高谐波能量可被几何精确预测→”血管几何决定脉象频谱”获谱级证据,脉诊的物理载体闭环完整;若波形相似但谐波能量系统性偏离→说明还存在几何之外的输入(管壁弹性个体化、微循环边界条件),脏腑谐波映射的机制解释需重修。无论哪种,都能把”同病不同脉/同脉不同病”从现象学推进到可计算的因果层面。
证据与局限
已验证: 患者特异 1-D 模型能高精度预测压力/血流波形(PMID 21622820, PMID 23701843, PMID 32142412);几何压缩保留波形形状(PMID 25888513);谐波相对能量携带冠心病等疾病判别信息(PMID 31275406, PMID 40396324)。
部分验证: 1-D 模型预测波形误差低至个位数百分比,但已有研究全部以波形而非逐谐波能量为比对目标;谐波能量对反射位点精度的敏感性尚未被系统量化。
待验证: 预测频谱 vs 实测频谱的逐谐波能量分布误差矩阵(各阶谐波相对能量比/相位差/归一化误差);误差随几何测量精度、血管段数的变化规律;逐谐波闭合在个体间的一致性(”血管几何指纹→脉象频谱指纹”跨人群稳定复现)。
相关阅读
证据等级: B(1-D 模型波形预测能力有 A 级直接支撑,谐波能量携带疾病信息也有直接文献;但”逐谐波能量分布的完整闭合检验”本身无先例,属研究设计提案)
参考文献: PMID 25888513, PMID 21622820, PMID 23701843, PMID 32142412, PMID 36379202, PMID 31275406, PMID 40396324
免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。
Harmonic Energy Distribution: Can Vascular Geometry Fully Determine Pulse Wave Spectra?
Evidence level: B
The assertion that “pulse = vessel geometry” is only rigorously validated when the waveform is decomposed into fundamental and higher-order harmonics, with energy ratios verified individually for each harmonic—this represents the next level of closed-loop validation for 1-D hemodynamic models.
Main Text
The previous article addressed whether “vessel geometry determines pulse waveform”—the answer is yes. Under conditions of complete measurement of geometric parameters, 1-D hemodynamic models can reproduce the pulse waveforms of 37 arterial networks with pressure errors below 4% (PMID 17640653). However, “visual similarity in shape” is merely a superficial verification: waveforms may be highly similar in the time domain while systematically deviating in frequency-domain energy distribution. Traditional pulse diagnosis does not read whether the waveform “looks alike”; it reads the relative energy relationships between harmonics. Wang Weigong’s pulse spectrum analysis exemplifies this approach by performing a Fourier decomposition on radial artery pressure waves, mapping the relative amplitudes of the first 10 harmonics (C1-C10) to internal organs (PMID 31275406). Therefore, advancing the causal chain from vessel geometry to pulse diagnosis to a “closed-loop verification of harmonic energy distribution” constitutes the strongest judgment experiment for “vessel geometry determines pulse.”
Why is this layer necessary? The mathematical properties of Fourier analysis dictate that waveform similarity in the time domain $\neq$ consistency in frequency-domain energy distribution. Two waves with nearly overlapping shapes may exhibit 20%-30% energy differences in the second and third harmonics, which are precisely the fingerprints relied upon for organ differentiation in clinical diagnosis. Existing clinical evidence confirms that harmonic energy carries disease information: a 2019 study used FFT to decompose radial artery waves from 83 subjects into 10 harmonics, finding elevated C1 and reduced C5 in the coronary artery disease group; incorporating these harmonic indices into traditional risk factors increased the ROC area under the curve from 0.788 to 0.856 (PMID 31275406); a 2026 SYNTAX score study further constructed a discriminative model using left-right hand harmonic differences ($\Delta$Cn/$\Delta$Pn/$\Delta$Dn) to predict the severity of coronary lesions (PMID 40396324). Harmonic energy distribution represents a signal with physiological counterparts—can it be predicted solely by “vessel geometry”?
The “spectral level” validation of the 1-D model already possesses a technical foundation. Patient-specific models have been repeatedly proven to predict waveforms with high precision: Reymond et al. reconstructed individual arterial tree geometries using MR angiography, showing patient-specific models significantly outperformed general models at specific sites (PMID 21622820); Willemet et al. compared 1D-0D models against in-vivo measurements in patients undergoing lower limb bypass surgery, reporting average relative root-mean-square errors of only 9.6% for pressure and 16.0% for velocity waveforms (PMID 23701843); a 2020 study involving 62 subjects used a coupled 0D-1D model to invert individualized parameters, achieving a low individual pressure wave MSE of 7.1 mmHg² and a correlation coefficient of 0.911 between individualized blood flow velocity and ultrasound measurements (PMID 32142412); a 2022 study of 22 patients with EECP also confirmed that personalized 0D-1D models can predict individual hemodynamics (PMID 36379202). However, all these validations relied on “waveform error”—point-by-point comparison between predicted and measured waves—rather than “harmonic energy error.” The potential risk that waveform and spectral errors do not necessarily move in the same direction is precisely the gap this article aims to fill.
There is also a critical model degree-of-freedom issue to address. To predict harmonic energy, the 1-D model requires geometric parameters (length, radius, branch location, wall elasticity) to be measured non-invasively with high accuracy. Epstein et al. demonstrated in 2015 that compressing distributed models of 55-67 arteries into 21-37 segments, while retaining total resistance and compliance, results in almost unchanged waveform shapes (PMID 25888513)—this indicates that the tolerance for geometric simplification is high for “waveform shape” but not necessarily for “harmonic energy distribution,” as harmonic energy is more sensitive to the precise location of reflection sites (thoracic aorta, renal arteries). This implies that a closed-loop verification at the harmonic level not only validates “vessel geometry determines pulse” but also inversely verifies whether “current non-invasive geometric measurement protocols are sufficiently refined.”
The design for an executable verification is now clear: ① Reconstruct the geometry of main conducting arteries (length/radius/branches/elasticity) for the same subject using ultrasound/MRI; ② Use a 1-D model to predict radial artery pressure waves and perform FFT decomposition; ③ Decompose the waveform measured at the same location using a flatbed tonometer or photoplethysmography; ④ Generate a harmonic-level error matrix—comparing predicted vs. measured relative energy ratios for harmonics 1-10, phase differences, and normalized root-mean-square errors for each harmonic energy; ⑤ Repeat across individuals to report attenuation curves of error relative to harmonic order, vessel segment count, and geometric precision. Both outcomes yield knowledge: if high-order harmonic energies can be precisely predicted by geometry → “vessel geometry determines pulse spectrum” gains spectral-level evidence, closing the physical carrier loop of pulse diagnosis; if waveforms are similar but harmonic energies systematically deviate → it indicates additional inputs beyond geometry (individualized wall elasticity, microcirculatory boundary conditions), requiring a revision of the mechanistic explanation for organ-harmonic mapping. Regardless of the outcome, this approach advances “same disease, different pulse / same pulse, different disease” from phenomenology to a calculable causal level.
Evidence and Limitations
Verified: Patient-specific 1-D models can predict pressure and flow waveforms with high precision (PMID 21622820, PMID 23701843, PMID 32142412); geometric simplification preserves waveform shape (PMID 25888513); relative harmonic energy carries discriminative information for coronary artery disease and other conditions (PMID 31275406, PMID 40396324).
Partially Verified: 1-D models predict waveform errors at single-digit percentages, yet all existing studies compared against waveforms rather than harmonic energy; the sensitivity of harmonic energy to reflection site precision has not been systematically quantified.
To Be Verified: Harmonic-level energy distribution error matrix comparing predicted vs. measured spectra (relative energy ratios/phase differences/normalized errors); trends of error variation with geometric measurement precision and vessel segment count; inter-subject consistency of closed-loop verification at the harmonic level (“vessel geometry fingerprint → pulse spectrum fingerprint” stable replication across populations).
Related Reading
- Vessel Geometry Determines Pulse Spectrum: Closed-Loop Validation of 1-D Hemodynamic Models
- Quantitative Mapping Model Between Arterial Geometry and Pulse Harmonic Spectrum
- Longitudinal Stability and Plasticity of Individual Pulse Spectrum Signatures
- Research Pool P0: Environment-Life Frequency Interface
Evidence Level: B (1-D model waveform prediction capability has direct A-level support; harmonic energy carrying disease information also has direct literature; however, “complete closed-loop verification of harmonic energy distribution” itself has no precedents and constitutes a research design proposal)
References: PMID 25888513, PMID 21622820, PMID 23701843, PMID 32142412, PMID 36379202, PMID 31275406, PMID 40396324
Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.
Leave a Reply