Center of the Skin Electrodynamic Introscopy, Kiev, Ukraine
Corresponding author details:
Yuri Babich
Centre of the Skin Electrodynamic Introscopy
Kiev,Ukraine
Copyright:
© 2019 Babich Y. This is an
open-access article distributed under the
terms of the Creative Commons Attribution
4.0 international License, which permits
unrestricted use, distribution, and
reproduction in any medium, provided the
original author and source are credited.
Our study aimed to reveal functional landscape/ heterogeneity of malignant and benign
tumors at background of visibly healthy tissue as it seen in electrical bioimpedance
parameters. The earlier developed method of the Skin Electrodynamic Introscopy (SEI),
which had ever firstly already enabled adequate imaging the Skin Electrodynamic
Landscape (SEL), proved also to be very sensitive in assessing the 2D skin dynamics in
response to weak external stimuli. SEI uses a spectral electrical impedance method of
measurement (modulus and phase of the impedance in the bandwidth of 2kHz-1MHz)
which provides information on events happening at cellular and subcellular levels at 32 ×
64 mm2 area with spatial resolution 1 mm. Nonthermal microwaves, weak magnetic field of
therapeutic intensity, and hypoxia were used as test influences. Analysis of the
multiparameter SEL of 7 cases of melanoma, 28 nevi and >50 cases of healthy skin revealed
novel phenomena. In all melanomas, immergence of test-induced clusters of coherent
activity, in-phase and out-of-phase synchronization of the SEL oscillations, accompanied by
effects of inter-cluster interactions and propagations with a velocity up to 1 cm/min has
been registered. The impact of magnetic field, in particular, was accompanied by the
transitory effects of the global synchronizations. Their localization and intensity
corresponded to visible tumor signs and, significantly exceeding them in details and
prevalence, thus created thus a diverse landscape of hypo/hypersensitive structures.
Such phenomena, but not so very pronounced and not so clearly associated with the
lvisible nevi, were observed only in 2-3 cases, and never in health. These effects are
presumably associated with cooperative mechanisms of short- and long-range of
intercellular interactions. It seems quite predictable that these effects can emerge not
only in the skin, but also in any tissues. The discovered phenomena may be helpful
specifically for real-time mapping the tumor geometry and its affected surroundings,
thus providing biofeedback and additional opportunities for cancer research, diagnosis,
personalized treatment and surgery, as well as in various biomedical applications.
Tumor; Clusters; Synchronization; Electrical impedance; Malignanization; Melanoma;
Hypoxia; Intercellular interactions; Imaging; Diagnostics; Magnetic field; Microwaves
Most studies of tumor heterogeneity have focused on genetic heterogeneity rather than on phenotypic heterogeneity that arises due to environmental selection forces in the tumor. However, despite our growing knowledge about cellular heterogeneity in cancer, we are far from understanding the dynamics that operate among the heterogeneous subpopulations and their role in disease progression and therapeutic responses [1]. Cancer can be described by a small number of functional ingredients, despite the complexity of the pathology [2], but yet, the cooperative behavior of cancer cells and clones, that lead to the development of tumors, mostly remains as a theory with only a few experimental demonstrations and no mechanistic dissection of such cooperative interactions [3,4]. Several such studies conducted over the past few years have provided important insights into mechanisms that maintain heterogeneity and clonal cooperation within the tumors and how these can affect tumor behavior [1,4,5].
Recent perspectives propose that the breakdown of intercellular cooperation could depend on ‘fields’ and other higher-level phenomena, and could be even mutations independent. Indeed, the field would be the context, allowing (or preventing) genetic mutations to undergo an intra-organism process analogous to natural selection. The complexities surrounding somatic evolution call for integration between multiple incomplete frameworks for interpreting intercellular cooperation and its pathologies [6].
It is widely accepted that skin is a very complex, heterogeneous system which comprises different epidermal and dermal cell populations. In healthy skin these processes are tightly balanced, so the competition does not change the homeostatic limits. However, even in a healthy skin, recently, clones of cancer-driving mutations have been discovered. Around 140 mutations per square centimeter were found in the six focus genes [7]. Subsequently, it was demonstrated that mutant cells can have a competitive advantage over neighboring cells [8-10]. In a case of further tumorigenesis, the corresponding modification of processes of competition and collaboration leads to gradual loss of tissue architecture and distortion of its functional portrait. It would be therefore important to reveal and monitor these topological features.
Bioelectricity is a basic phenomenon associated with cellular and subcellular structures [11]. Most of the subcellular biomolecules (e.g. DNA, RNA, tubulin, actin, septin etc.) are either charged and hence surrounded by counter-ions or endowed with high electric dipole moments that can engage in dipole-dipole interactions and polarize electrically their local environment. Living organisms are replete with both moving and oscillating electric charges and can thus be regarded as complex electrochemical and mechanical systems [12].
Since cells and all types of tissues communicate electrically, it seems logical that the methods of studying these processes should be appropriate, i.e., in particular, providing dynamic visualization of the Skin Electrodynamic Landscape (SEL). Thus, research into electrical manifestation of spatio-temporal features of tumor and its micro- and macro-environment appears to us as a vital technical and biophysical and medical challenge.
In recent decades, the method of Electrical Impedance Spectroscopy (EIS) has been intensively developed, particularly in the field of diagnosing skin cancer [13-24]. It provides valuable information in vivo on the intercellular and intracellular environment, as well as on the integrity of cell membranes. The electrical impedance analysis of single cells can provide information on cells’ pathological condition in various environments [25]. However, existing approaches on reading out the SEL do not provide a sufficiently high spatial resolution in large enough areas of the skin to obtain a coherent /unfragmented electrical portrait. However, even with these limitations, it becomes possible to detect small tumors and some tumor-related changes [26].
Our pioneering developments in the visualization and research into the phenomenology of the Skin Electrodynamic Landscape (SEL) began in the mid-1980s, when, with the aid of our early original experimental setup, firstly in the world, adequate static and, shortly, dynamic spectral images of the skin electrical landscape have been obtained. Studying the influence of various physicochemical factors (mechanical, nonthermal microwaves, magnetic fields, pharmaceuticals, and hypoxia) onto modification of the initial SEL in healthy and sick volunteers, we soon discovered a number of fundamental new spatio-temporal phenomena [27-33].
Summarizing, it can be said that SEL in healthy individuals has
predominantly chaotic low-amplitude dynamics, for modification
of which strong enough (up to painful/invasive) external stimuli
are required. On the contrary, in a case of sickness/allergy of the
same subject, the initial SEL becomes deeper and more structured;
and even vanishingly weak stimuli can trigger a process of still
more high-amplitude synchronous structuring in the form of
pulsating or wavelike structures with characteristic velocity
about the range of calcium waves of intercellular signaling. Such
distinctly expressed functional heterogeneity of the SEL was
observed in all cases of malignant melanoma (MM) (7 cases),
and only in 3 (in a poorly expressed form) of 28 cases of benign
neoplasms (nevi), and, under the same experimental conditions,
never in healthy subjects (>50). This paper firstly presents a more
detailed analysis of the discovered phenomena (both initial and
test-induced ones) of functional clusterization of the MM area.
Nonthermal microwaves (MW), magnetic field (MF), and hypoxia
were used as testing stimuli. Two methodological approaches are
examined: steady state measurements and complex spatial and/or
temporal measurements.
Here we give only brief information about the method, which was described in more detail in [30]. In this study, we used the hypoxic test similarly to [30], but now the focus has been on the effects caused by MW and MF.
The sensory head of the SEI setup is a matrix of stainless steel 32 × 64 electrodes (304, Sandmeyer Steel Co., USA), 0.6 × 0.6 mm2 each, making thus spatial resolution 1 mm (Figure 1). The measurement circuit includes: an active/running electrode, multiplexer, sinusoidal current generator (3 and 10 µA of 2 kHz and 1 MHz, respectfully), and a large (~10 cm2 ) ground electrode (Figure 2). The large electrode is equivalent to an imaginary one being located interstitially in the most conductive deep layer of skin, i.e. dermis (connective tissue and intercellular liquid), enabling thus namely cross-sectional mode of measurement. Here we analyze only 4 out of 6 simultaneously measured parameters of impedance Z=|Z|ejθ: modules (|Zk |, |ZM|), and phase angles (θk , θM). Subscripts k and M mean kHz and MHz of the spectral impedances. The frequency span 2kHz-1MHz enables to make some distinction between inter- and extracellular data since at the high frequency cellular membrane became transparent. The duration of a single measurement cycle is 4 ms/pixel, and accordingly the duration of the frame scan is ~ 8s for the full frame (64 × 32) or ~ 4s for its half (32 × 32). To register rapid test-induced changes in SEL (the MW and MF stages), the half-matrix scan mode was used. The measurement range: |Z| - from 0.2 to 35 k Ohm, θ – from -90º to +90º. Calibration was carried out at the resistance-capacitance (RC) circuits and showed: mean magnitude relative error 0.2%, mean phase absolute error 0.5º. The spatial non-uniformity of such RC direct transform, caused mainly by the on-resistance scatter of the analogue switches of multiplexer (±20 Ohms, i.e. ~0.1-0.4% of the mean |Z|) was considered insignificant for assessment SEL both: in absolute values and - a fortiori- its relative dynamics. In order to minimize and stabilize the contact impedance, electroconductive gel BEEG Super cream (Cerracarta, Italy) was used. The phantom measurements (pure gelatin 25% w/v in physiological solution) showed the error values in the difference image of two neighboring frames: Δ|Zk | = mean ± Ϭ = -0.01 ± 0.025; Δθk = mean ± Ϭ = 0.00 ± 0.008. The noise, originated from the fluctuations of the skin-electrode interface, has mainly such a pulsed character, that it can significantly interfere only to the dynamics of 1 or 2 neighboring pixels, but not to the analyzed multipixel structures. In addition, the probability of repetition of the peak values ripples in the same pixel in subsequent frames is small, which is also confirmed by the below dispersion maps for each stage of the experiment (15 frames), from which it follows that the topology of dispersion maxima is not accidental, but reflects initial features of the SEL. So, presenting the SEL images, we deemed it advisable not to cut off the values of the ± 2Ϭ level, but, when necessary (rarely), to use just a weakened palette colors at the level of about ± 1σ.
• The ”Threshold-1” (production of “Vidguk”, Kiev, approved by the Ministry of Health of the USSR for application in 1989) uses a noise generator of extremely low intensity (<0.1 mW/cm2 directly at the output) in the microwave band 54-75 GHz [34- 36]. Intensive works on of the biological effects of microwaves began in f. USSR early in the 1980s, when an unusually high effectiveness of their action onto a person in periods, when their normal functioning is disturbed, had been established [35]. The ”Threshold-1” was used as a basic one in developing method of information-wave therapy. The principal difference of this method from other similar ones is the use not of a part, but of the entire range of microwaves. It is believed that in a broad spectrum of radiation, all (or almost all) vibrations with physiologically significant information superimposed on them are necessary to restore the information homeostasis in the affected organs and systems [35,36]. Specifically, with the aid of ”Threshold-1”, the microwave-induced wavelike phenomena of the SEL had been firstly registered [28]. Now, in the first experiment, the device was (accidental, not intentionally) used twice: 1st time- at about 1 cm distance from the border of the scan area, i.e. with repeatedly reduced the already extremely low-intensive power flow, and the 2d time - directly to the border.
• The ordinary stationary magnet (37 × 50 × 25 mm) created the nonuniform magnetic field at the scanning zone in the range of 20-30 mT. Two versions of magnetic impact were tested: (i) a relatively slow unipolar increase and decrease of the MF marked as MF↓, and (ii) a reversing MF marked as MF↑↓. In the first case, the south pole of the permanent magnet was slowly brought (within a few seconds) close to the scan zone at a distance of about 1 cm, and slowly removed it in 1 min. In the second case, the north, south of the same magnet was quickly brought to the edge of the scan area, and, after about 30s, it was turned over twice within the next 30 seconds, after which it was quickly removed.
• At the beginning of the final –relaxation- stage, a short breathhold test during ~50s was also used.
Figure 1: Photos of the SEI experimental set-up and its sensor head
(right) - flat multielectrode matrix of 32 × 64 mm2
, 32 × 64 =2048
stainless steel electrodes of 0.6 × 0.6 mm2
each.
Figure 2: Photo of the microwave generator “Threshold-1”
(«Порог-1» in Russian)and a stationary magnet.
As a whole, the experiment consisted of 10 stages (145 frames): the preliminary full-frame adaptation stage “0”: of 10 frames (~3min), following by 8 main stages of the 32 × 32 mm2 frames (“Ι”-“VΙΙΙ”, #20-120, Figure 5), and a final one-“ΙX”- of 20 full frames. The main stages, from I to VΙΙ, included the following tests:
I. No influence (10 frames, all the rest 7 stages consisted of 15 frames);
II. The microwave exposure of vanishingly low intensity - MW1 in the course of the first 10 frames;
III. No influence, relaxation;
IV. The microwave of extremely low intensity MW2 during 10 first frames;
V. Relaxation;
VI. The constant magnetic field MF↓ during first 10 frames;
VII. The reversing MF↑↓ during first 10 frames;
VIII. Relaxation
Figure 5 illustrates the frame-to-frame test-induced changes of the 4 SEL parameters (θk , θM,|Zk |,|ZM|) at the 32 × 32 area in the form of correlation graphs calculated for each subsequent frame-image matrix relative to the first one.
The high sensitivity of the SEL to the MW and reversing MF was confirmed even in the analysis of means (Figure 5). Of particular note is the unexpectedly pronounced reaction to the breath holding (IX). It is important to emphasize the clearly visible response sequence of different parameters: |ZM|→|Zk |→ θM (the θk -response seemingly have occurred simultaneously with that of |ZM|, although it is not so obvious from the smoother θk -graph). Regarding the stages I-VIII, the correlation dependencies can create a false impression that the most significant changes in the SEL are reflected in the parameters of |Zk | and |ZM|. The subsequent analysis shows that in this case, in order to identify and evaluate clustering process, the change of interand intracellular conductivity are less adequate than that of the cooperative dynamics of cell membranes (θk , θM). This conclusion suggests itself already even when comparing Figure 6 and 7. Thus, our earlier assumption [31], that the topology of such super-high SEL fluctuations was directly related with specificity the MM environment, has been unexpectedly confirmed.
Figure 8 represents an overview of the test-induced events/ changes of θk and |Zk | (in %) happening in each subsequent stage with respect to the preceding one. At the I stage (I-0, Figure 8a, a’), both difference maps θk and |Zk | showed rather chaotic character and do not provide much significant information. The first signs of the MW1- induced clusterization become noticeable in the ZMT (“3”) at the II-I map (b,b’). The noticeable clusterization happened as a result of the MW1 after effect during relaxation stage (III-II) can be clearly seen at c,c’. It is worth to emphasize the spatial discrepancy between the θk - and |Zk |-structures inside and near the ZMT, and emergence of red coherent structures antiphased to the θk -blue at the “4”and “5” sectors. There are also some significant distinctions between the θk - and |Zk |- patterns, which enable to contrast the tumor geometry, i.e.: the large blue structure at c) and the red islands at c’) around the melanoma perimeter. A noticeable response of the edge of the 7th sector should also be mentioned.
The MF↓ exposure (f,f’) seemingly inhibits the source of the wavelike θk -activity,i.e.: activity of the “2,3,5” areas. There are two areas of interesting exceptions: in the lower left corner and in “6”; the red structures indicate further propagation of the θk -crest. The θk-reaction of the scanned part of zone “7” is also clearly visible.
The |Zk |-response, similarly to d’, also has a total character. Against the background of the general rise in the |Zk |-level, a multitude of islands have manifested themselves again in the vicinity of melanoma, especially in the ZMT.
In the time domain, in scale interframe changes, corresponding subregions of the θk -anomalies were also detected. The correlation fields depicted at Figure 11 show the degree of consistency of the dynamics of each pixel with respect to the average dynamics of all pixels, which allows revealing the cluster character of the SEL dynamics in more details. The clustering effect may have been already emerged even at the preliminary stage I (under the action of the measuring procedure) as a few blue spots exactly in ZMT, and which then multiplied and concentrated there in the II stage. The relaxation stage III (c) reveals a remarkable aftereffect of MW1, i.e. about totally synchronized response with positive correlations throughout the whole scan-area (r=0.05…1). Moreover, topology of the areas with minimum r well coincides with the areas of maximum θk -oscillations (Figure 10, a1-a3). The MW2-exposure (d) triggers fast development (within two frames, Figure 9) of distinctly antiphase structures (r ≈±1), i.e. “3” vs “4”,”5” (Figure 4 e). The MW2-aftereffect, unlike that of MW1, does not cause total synchronization, but only leads to the development of two new antiphase structures (on the border of melanoma), which apparently continue the process that arose in previous stage (Figure 11d, bottom left). The MF↓-exposure (Figure 11f) did not cause any noticeable changes. In contrast, the impact of MF↑↓ (Figure 11g) resulted in increased synchronization (r=0.4..1), i.e. higher than that of MW2 (c). Worth noting, that this time, the “3” zone was not involved in the synchronization process. The relaxation stage VIII (Figure 11h) looks more informative that the previous one: the state of high synchronization is preserved only in the inactive zones, while the closer to zero correlation r in the active ones reveals the melanoma core and other features of the scan-area (i.e.: “2,3,4,5,7”, Figure 4).
In Figure 12, a few examples of the in-phase and anti-phase dynamics of some pixels located on both sides of the ZMT border are presented. Two such cases (bold curves) indicate a distinct antiphase response of θk to MW2: the pixels 8 × 28 and 6 ×26, located on both sides of the border of ZMT zone (3, Figure 4). Besides, the p8 × 28 demonstrates a particularly marked behavior, i.e. a sound, but somewhat returnable, drop of |Zk | in response to MW1, near-stable regime during the relaxation (III), a trigger-like response to both MW2, and following resistant equanimity at this lower level during the subsequent stages V-VII. As for the |Zk | dynamics, there are two interesting pixels bordering on ZMT: (i) p6 × 23 demonstrating marked transient reaction to MW2 followed by a return to its average dynamics; and (ii) p13 × 26, which is out of phase with p6 × 26 in response to both: MW1 and MW2, and also is out of phase with the averaged response to MF↓.
Figures 13 and 14 indicate the existence of zones of coherence/ anticoherence with respect to the MW1-induced θk - and |Zk |- dynamics of some ZMT pixels. Left: temporal θk -curves (running average) of the adjacent pixels (p6 × 24, p6 × 25 and p6 × 26), and the corresponding correlation fields calculated for the joint period of stages II-III (compare with Figure 8c,d,c’,d’).
It should be recalled here that all these sub-clusters - p5 × 23, p7
× 25 (as well as p13 × 25) - are components of the time averaged |Zk
|-
structure, i.e. the |Zk
|-response to MW2, which clearly manifested
itself (as a marked red whole) adjacent and partially penetrating
into the “3” zone, Figure 8c’).
Figure 3: Photos: (i) The scanner placed onto the melanoma area
(the ground electrode is placed on the left). (ii) The melanoma
area. On the skin, one can see the imprint of the scanner’s head.
The rectangles denote the 32 × 32 zone of rapid scanning (the solid line) and a full area of investigation (the dotted line). A dashed
circle indicates a pigmentation zone of undetermined origin. The
blue arrow indicates the direction of the test influences exposures
(microwaves and MF).
Figure 4: Scheme of the most representative sub-areas of the 32
x 32 zone: 1) Contour of melanoma; 2) Its central part; 3) Zone of
maximum transformations (ZMT); 4-6) The areas surrounding the
ZMT; 7) A part of the pigmentation zone; 8) An area intermediate
between 7 and 6.
Figure 5: Graphs ofcorrelation between the all current θk
,
θM,|Zk
|,|ZM| image matrices and initial one calculated for the 32
× 32 mm2
area.