Ollmar S, Nicander I, Ollmar J, Emtestam L
Center for Oral Biology, Karolinska Institute, Novum, Huddinge, Sweden.
Med Biol Eng Comput. 1997 Jul;35(4):415-9. doi: 10.1007/BF02534100.
Information relating to patho-physiological events in living tissues, including skin, should be found in the beta-dispersion frequency range of the electrical impedance. For intact skin, the alpha-dispersion is strongly influenced by the condition of the stratum corneum, particularly its hydration. For intact skin, these dispersions are not well separated: in a Nyquist plot, most of the relatively small beta-dispersion will be hidden behind the upper-frequency end of the alpha-dispersion arc. In the study, information obtained with a set of four indices, claimed to extract most of the information in the pertinent frequency interval, is compared with the full information measured at 31 frequencies, between 1 kHz and 1 MHz, for irritant and allergic contact reactions and nodular basal cell carcinomas, using a holographic neural network that appears to be useful for model-independent evaluation of the consequences of data-reduction procedures. Cole parameters should be avoided in the beta-range for intact skin. The indices are well supported, as long as differences from a reference site are used, and it seems that they can serve as the basis for differential diagnostics even as absolute values, although more information may be extracted from the complete spectrum.
与包括皮肤在内的活体组织中的病理生理事件相关的信息，应在电阻抗的β色散频率范围内找到。对于完整皮肤，α色散受角质层状况的强烈影响，尤其是其水合作用。对于完整皮肤，这些色散没有很好地分开：在奈奎斯特图中，大部分相对较小的β色散将隐藏在α色散弧的高频端后面。在该研究中，使用一组四个指标获得的信息（据称可提取相关频率区间内的大部分信息）与在1 kHz至1 MHz之间的31个频率下测量的完整信息进行了比较，这些信息涉及刺激性和过敏性接触反应以及结节性基底细胞癌，使用了一种全息神经网络，该网络似乎有助于对数据简化程序的结果进行独立于模型的评估。对于完整皮肤，应避免在β范围内使用科尔参数。只要使用与参考部位的差异，这些指标就有充分的依据，而且即使作为绝对值，它们似乎也可作为鉴别诊断的基础，尽管从完整频谱中可能会提取更多信息。
