Mathematician review: Pappas-PT (US 5,556,418, "Method and Apparatus for Pulsed Magnetic Induction")

Reviewer role: PhD mathematician (Step 2). Checks derivations, notation, units, hidden assumptions. Physics verdicts belong to the physicist; where I use physics I use only textbook relations and say so.

Source basis: the one project document is the patent (7 pages, text layer plus page images) [read]. I read the deep read (deep/US5556418-dd083a.md) and checked its key passages against the extracted text (_text/US5556418-dd083a.txt, lines ~195, 259-272, 321-324). I also used the Investigator's leads (sites.md, research/ notes, and the Physiology of the Cell PDF in research/files/, which I grepped for the power and pulse figures). Calculations were done with numpy in the project venv. No separate web calls were made in this pass: every [memory] fact below that a verdict depends on is high-confidence textbook physics (stated as such). The Investigator's [web] items are carried with their tags and are not re-verified by me unless stated.

Reader notes (symbols used): V = volts, E = electric field (V/m), B = magnetic field (tesla, T), f = frequency, a = cell radius, R = resistance (ohm), Q = quality factor (how many oscillations a ringing circuit survives), eV = electron-volt, h = Planck's constant, kT = thermal energy.

Corrections to Step 1

1. claims.md claim 3 is stated as the patent's logic, but the patent's two numbers are not "1-100 times"-consistent as written. The patent prints 10^6 V/m for the membrane field and 10^6 V/m for air breakdown (the deep read correctly established this is the patent's own text, not an OCR error [read: page image p.4]). Step 1 notes this but I add two further errors the deep read did not flag, both arithmetic: - Membrane field: 70 mV across a real membrane thickness of about 5 nm is 0.070 / 5e-9 = 1.4e7 V/m [memory, high: bilayer thickness 4-10 nm]. Across the patent's "several Angstrom" (3 Å) it is 2.3e8 V/m. Neither is 10^6. - Air: "10 kV/cm" = 10^6 V/m is arithmetically right, but real air breakdown is about 30 kV/cm = 3e6 V/m [memory, high]. The patent's value is low by about 3 times. The deep read's remark that the membrane field is "about 100x the air value" under the patent's own thickness is correct (2.3e8 / 1e6 = 230) and I keep it. Step 1's framing of the 1-100x boost as "fitting" is a reconciliation the patent does not make; I treat it as the reviewer's suggestion, not the inventor's argument.
2. claims.md claim 2 says bursts are "at tens to hundreds of MHz in the figures." The only figure data are the scope's statistic line: current 100.143 MHz, min 1.317 MHz, max 240.3 MHz, average 110.0 MHz, identical on FIG. 2 and FIG. 3 [read: page image p.3]. Those are the scope's frequency-measurement statistics over an unstated gate, not a spectrum. A minimum of 1.3 MHz and a maximum of 240 MHz is a spread of about 180 times, which is not what a single "characteristic eigenfrequency" would give. Step 1 reads this correctly in the deep read but claims.md condenses it to "tens to hundreds of MHz".
3. Claim 5's wording "restoring transmembrane potential is therapeutic" is the patent's assertion with no data (deep read section 6 confirms none). Step 1 is right; I only add that "the pulses can also pasteurize by pressure increase" is a separate claim (see claim 3 below) and should not share a verdict with the potential-restoration claim.
4. The deep read's inference that FIG. 2's base curve is about 0.5-1 MHz is mine-to-check, and it conflicts with the Investigator's numbers. The Investigator's independent descriptions (Viti 2023, Keilani 2025) give a base frequency of about 240 kHz and pulse about 50 us [web, via sites.md, not re-verified]. FIG. 2's 10 us window with bursts at about 1 us spacing would be a first-microsecond fragment of a ring that is much faster than 240 kHz, or a different instrument setting. The two are not necessarily in conflict (different devices and years) but the patent cannot be used to fix the base frequency. Unknown.
5. Claim 5/6 unit error stands. Claim 5 "diameter ... 6 square millimeters" mixes length and area [read: p.7]; claim 6 "diameter at least a fraction of one centimeter" conflicts with the specification's preferred 30 cm coil [read: p.6]. Not a physics error; a drafting inconsistency that I list for the record.

No numeric error was found in the deep read's transcription of the resistance figures (0.5 ohm, -100 ohm) or the burst timing.

Worked checks used below

(a) Coil field. A 30 cm diameter coil (R = 0.15 m) with N = 5 turns carrying I = 1000 A (the patent's stated minimum, [read: p.6]) has centre field B = mu0 N I / (2R) = 4 pi e-7 x 5 x 1000 / 0.3 = 0.021 T. At 10 kA (the Papimi figures the Investigator cites: "several 10 kA") it is 0.21 T, consistent with the 50-150 mT the Investigator quotes [web, sites.md]. So the magnetic field scale is plausible for the apparatus (I use mu0 N I/2R, a standard Griffiths-level formula [memory, high]).

(b) Induced electric field in tissue (Faraday's law, E x 2 pi r = -dPhi/dt, so E = (r/2) dB/dt for a uniform field). Take B = 0.1 T oscillating at 240 kHz: peak dB/dt = 2 pi f B = 1.5e5 T/s. At r = 0.10 m, E = 7.5 kV/m = 7.5e3 V/m.

(c) Membrane voltage that a field E induces on a spherical cell of radius a (Schwan's result, low-frequency limit): delta V_m = 1.5 E a cos(theta) [memory, high]. With E = 7.5e3 V/m and a = 10 um: 1.5 x 7.5e3 x 1e-5 = 0.11 V. So tissue-level fields of only a few kV/m already change the membrane voltage by tens to 100 mV (if the field is slow compared with the membrane charging time, which for a 100 MHz burst it is not).

(d) Energy per photon at 100 MHz: h f = 6.6e-34 x 1e8 = 6.6e-26 J = 4.1e-7 eV. Compare thermal energy at body temperature kT = 0.027 eV, and a chemical activation or endothermic energy of order 1 eV.

(e) Plasma frequency of an electron gas: f_p = 8.98 kHz x sqrt(n_e), with n_e in cm^-3 [memory, high]. For f_p = 100 MHz, n_e = (1e8 / 8980)^2 = 1.2e8 cm^-3. A real arc has n_e of 1e14-1e17 cm^-3 (f_p of 90 GHz and above) [memory, high].

Claim-by-claim

Claim 1. Plasma arc gives transient voltage and current 10-100 times the supplied potential and average current, via arc resistance going from infinite to <0.5 ohm to about -100 ohm ("infinite" then negative "gain").

a. Internal consistency. The argument is a ratio of resistances. The text: from infinity before ignition to under 0.5 ohm after ignition, "equivalent, theoretically, to a gain of infinity/0.5, which equals plus infinity, and to a subsequent negative gain of 0.5/-100 < 0" [read: page image p.5]. - A ratio of two resistances is a dimensionless number with no stated relation to a ratio of voltages, currents or powers. For a circuit with source voltage V and total resistance R, current is V/R; changing R from infinity to 0.5 ohm changes the current from 0 to V/0.5. That is a switch closing, not gain. Call the "gain" R_before/R_after. The current ratio is I_after/I_before = R_before/R_after only in the trivial sense that the starting current is zero, so the ratio is infinite by division by zero. It carries no information. - "0.5/-100 < 0" is a ratio of the new resistance to the next one. It has no physical meaning either as gain: it is negative simply because one number has a minus sign. A negative resistance in the static sense (V/I < 0) would mean the element delivers power continuously, which an arc does not; an arc has a negative DIFFERENTIAL resistance (dV/dI < 0 on its characteristic) with V/I still positive [memory, high]. The patent does not distinguish static from differential resistance. A -100 ohm differential resistance in series with a 0.5 ohm circuit makes the circuit unstable and drives growing oscillations until the nonlinearity limits them, which is the standard arc-oscillator (Duddell singing arc) behaviour. That is real and well known. It does not give a gain in energy. - The patent itself concedes the energy accounting: "The energy increase of these electrons or ions is achieved by a corresponding decrease in the energy of the remaining electrons or ions" [read: p.4]. That sentence is energy-conserving. The next sentence of the summary ("significant excess of ... effective power is produced") is not compatible with it unless "excess" means "excess over the instantaneous source level", which is a different thing. - Hidden assumption: that "effective power" and "average current" can be compared across a duty cycle of 0.001% (1e-5, [read: p.6]). With that duty cycle, peak/average for ANY pulse train is about 1e5. So "transient current 10-100 times the average current" is automatically true of the apparatus described and carries no excess-energy content. Likewise a ringing LC circuit with quenched arc (the stated waveform: damped sinusoid with bursts at extrema [read: p.6]) overshoots the charging voltage by up to a factor of 2 in the simplest series-LC step response and by more where inductive kick or a step-up transformer coupling is present [memory, high]. 10-100x transient voltage is therefore unremarkable as a statement about voltage; as a statement about energy it is unsupported. - No numbers anywhere to test: no capacitance, charging voltage, inductance, peak current or stored energy in the patent (deep read section 3). The central quantitative claim cannot be checked against its own source. The later Papimi descriptions supply numbers (40 kV, 60-96 J per pulse, tens of kA) [web, via sites.md]. With 60 J and 40 kV the stored capacitance is C = 2E/V^2 = 2 x 60 / (4e4)^2 = 7.5e-8 F = 75 nF (my arithmetic). Those are the manufacturer's figures for a later device, not part of this patent. Verdict on (a): the mathematics is not derived; the one argument given (ratio of resistances) is a non sequitur as a gain; the weak reading (peak current and voltage far above average) is true by duty cycle alone.

b. Consistency with experiments. Arcs with negative differential resistance, relaxation oscillators and ringing spark gaps are standard experimental physics (spark-gap transmitters, Tesla coils) [memory, high]. They conserve energy: output energy per pulse is no greater than energy stored in the capacitor bank. Nothing in the patent measures input and output energy. Consistent for the weak form; not tested for the strong form.

c. Tested? In the patent: only two oscillograms of the output coil with no vertical scale (deep read section 4). The inventor cites his own 1989-1991 conference papers (Einsiedeln 1989, Colorado Springs 1990, IECEC Boston 1991) as the basis for the 10-100x claim [read: p.4]. The IECEC 1991 paper ("Energy Creation in Electrical Sparks and Discharges: Theory and Direct Experimental Evidence", vol. 4, pp. 416-423 [web, via sites.md]) has not been obtained, so the measurement behind the claim is unseen. Inventor's later claims go further: "20% excess heat on the probe alone" (New Energy News June 1998, p.28, file in research/files) and a ">1 Giga Watts" peak (Physiology of the Cell, 2006) [read, Investigator's files]. No independent replication of any excess-energy claim was found by the Investigator. Checking the later GW claim for self-consistency (my arithmetic): 80 J delivered at 1 GW lasts 80/1e9 = 80 ns, so a 1 GW peak is compatible with 80 J only if the pulse is under about 100 ns. Papimi sessions of "888 pulses at about 80 J each" [read: Physiology PDF] correspond to 888 x 80 J = 71 kJ, at 2-3 pulses per second 5-7 minutes. That is internally consistent. The GW figure is about a peak power in a short burst, not average power, and says nothing about gain.

d. If true (strong, energy-gain form): conservation of energy would fail in a closed circuit with a capacitor store. The weak form (voltage and current spikes above average) requires no change to mainstream physics.

Verdict (split): - Weak form (arc-oscillator spikes above the DC supply level and above the time-average current): Established as ordinary circuit behaviour; the patent adds no data. - Strong form (net excess voltage, current or power beyond the stored energy, "infinite gain" through resistance ratios, and the inventor's later "20% excess heat"): Refuted by calculation as an argument (a resistance ratio is not a power gain; the patent's own sentence on p.4 states energy is redistributed; with duty cycle 1e-5 the peak/average ratio needs no gain). As an experimental claim of net excess energy it is Never properly tested (no input-output energy measurement on record, the 1991 paper not seen).

Claim 2. The plasma emits sub-microsecond bursts (10-15 per train, duty cycle about 0.001%) at its characteristic radio eigenfrequencies, delivered to the body via an induction coil.

a. Internal consistency. Timing numbers differ across the document: description "less than 10^-6 s" [read: page image p.6], claim 14 "about 10^-4 to 10^-6", claim 16 "less than 10^-4" [read: page image p.7]. These are compatible only if "less than 10^-6" is read as inside "10^-4 to 10^-6", which it is not (below 10^-6 is outside the range). Minor internal inconsistency in claims versus text. Duty cycle check (deep read, my agreement): 10-15 bursts of up to 1 us is about 15 us per low-frequency cycle; for a duty cycle of 1e-5 the repetition period would be 15e-6 / 1e-5 = 1.5 s, matching the "few Hertz" charging. I also checked it the other way: at 2-3 pulses per second the fraction of time on is 15 us x 3 / 1 s = 4.5e-5, so the 0.001% (1e-5) figure is right to within a factor of 4. Consistent. "Eigenfrequency characteristic of the gas": an eigenfrequency of a gas plasma is the electron plasma frequency, f_p = 8.98 kHz x sqrt(n_e) (n_e in cm^-3). The oscillogram frequencies of 1.3-240 MHz (mean about 110 MHz) correspond to n_e of order 1e8 cm^-3 for 100 MHz (calculation (e)), orders of magnitude below any arc (1e14 and above, f_p of tens of GHz). Moreover f_p depends on electron density, not on the gas species, so it does not identify "nitrogen" or "air" in any way. A circuit with 0.1-1 uH of stray inductance and tens of pF of stray capacitance rings at 100 MHz (1/(2 pi sqrt(LC)); L = 100 nH, C = 25 pF gives 100 MHz) [memory, high: LC resonance formula]. The observed frequencies are therefore at least as well explained by the parasitic LC of the probe, leads and electrodes as by any plasma property. That is a hypothesis, not a measurement; the patent shows no test (for example, a different gas with the same geometry). The scope statistics are identical on FIG. 2 and FIG. 3 (deep read), so they are not tied to a particular burst; they cannot be used to say the bursts have an eigenfrequency of 100 MHz.

b. Consistency with experiment. Sparks and arcs do emit broadband radio-frequency noise and ring at circuit resonances [memory, high]. Gas-specific lines are in the optical/UV range, not radio.

c. Tested? Waveform shown on two oscillograms; the Investigator's independent papers describe the same damped oscillation with megahertz-to-gigahertz peaks at extrema (Keilani 2025: "In the maxima and minima of the damped oscillation, high-frequency oscillation peaks in the megahertz to gigahertz range arise") [web, via sites.md]. These are authors describing manufacturer data, not independent measurements. No test links the burst frequencies to the gas.

d. If true: nothing in mainstream physics would have to change for bursts to exist; only the "characteristic of the gas" part would.

Verdict (split): - Bursts at the extrema of a damped LC ring, sub-microsecond, repetitive at a few Hz: Established for this class of device (the patent's own oscillograms plus later descriptions). - That the burst frequencies are gas-specific "eigenfrequencies" (nitrogen vs air vs argon): Never properly tested (no comparison of gases; a circuit-resonance explanation is not excluded). Calculation (e) shows the plasma-frequency reading gives an implausible electron density.

Claim 3. Because bursts are shorter than the time needed to break down air, their fields can exceed the dielectric strength of air by 1-100 times without ionizing it, so cell-scale fields of order 10^6 V/m are possible non-invasively.

a. Internal consistency. Faults found: - Membrane field and air breakdown are given the same number, 10^6 V/m, and the argument is that the external field must exceed air's strength [read: page image p.4]. As shown above, the membrane field is 1.4e7-2.3e8 V/m, and air breakdown is 3e6 V/m. - The premise is a category mistake. The patent equates "the field needed to affect a membrane" with "the field inside the membrane" (70 mV over its thickness). An external field does not need to equal the membrane field. For a cell of radius a in a uniform external field, the membrane voltage is delta V_m = 1.5 E a cos(theta). The ratio of membrane field to external field is therefore about (1.5 a / d) where d is membrane thickness: with a = 10 um and d = 5 nm this is 1.5 x 1e-5 / 5e-9 = 3000. To shift V_m by 70 mV the external field needed is E = 0.07 / (1.5 x 1e-5) = 4.7e3 V/m (not 10^6 V/m). So the "prohibitive 10^6 V/m" requirement disappears: fields of a few kV/m are enough at low frequency. This is the standard cell-electroporation relation (Schwan) [memory, high]. The patent's Nordenstrom comparison, "several volts per meter" [read: p.4], is itself a DC electrode current field; it affects cells by ion drift, a different mechanism. - The pass-through into tissue by induction needs no air breakdown consideration at all. The fields come from Faraday's law inside the body. From calculation (b), 0.1 T at 240 kHz at 10 cm radius gives E = 7.5 kV/m, which is 400 times below air breakdown (3e6 V/m). So "beating the dielectric strength of air" is neither needed nor achieved by the coil; the breakdown concern applies only inside the plasma chamber and near coil insulation. - The idea that a short pulse can transiently exceed the static breakdown field before an avalanche forms is real physics (statistical and formative time lag; nanosecond pulses sustain fields several times the static value) [memory, high]. The 10x-100x multiple is not supported by that physics at the scale needed: sustaining 100 x 3e6 = 3e8 V/m in air for a nanosecond is not achievable in practice (surface and field-emission breakdown limit around 1e8 V/m in vacuum gaps) [memory, medium; not load-bearing for the verdict]. - Dose: the claim says the field penetration depth is "proportional to the intensity of the applied field ... can reach any desired depth by increasing the instantaneous power" [read: p.5]. For a loop in a lossy medium, the induced field falls with distance and with skin depth at the frequency; at 100 MHz the skin depth in muscle is a few cm [memory, medium]. "Any depth" is not derived. - "Cold pasteurization" by a pressure increase inside cells is stated without a model [read: p.4]. The established route to non-thermal microbial kill by pulsed fields is pulsed electric field (PEF) treatment at 1e6-5e6 V/m (10-50 kV/cm) for microsecond pulses [memory, high]; the induction fields in (b) are about 200 times smaller than that.

b. Consistency with experiment. Pulsed electric fields do permeabilise and kill cells, but at the field strengths and durations above, delivered with electrodes. Magnetic induction at the levels in (a) and (b) is the well-studied territory of pulsed electromagnetic field (PEMF) devices and magnetic stimulation: tissue fields of order 1-100 V/m in transcranial magnetic stimulation at ~0.1 to 2 T [memory, medium] do depolarise nerves. Consistent with fields sufficient for nerve stimulation, but not for the claimed effects on the cancer cell potential.

c. Tested? Leisner et al. 2002 (J Vet Med A 49:33, PMID 11913824; Hebrew University) used a PAP-IMI device in a rat ulna fracture model: radiographic healing was slower, more fibrous callus; "The results of this study do not support the claim that PEMF generated by the PAP-IMI stimulate osteogenesis" [web via sites.md and research/files/pubmed-11913824-Leisner2002.txt]. That tests a bone-healing outcome, not the specific claim of non-ionising super-air fields at a cell. Nobody has measured the field in tissue for this device in a source I have seen.

d. If true as stated: nothing in mainstream physics changes; air breakdown physics already allows brief overvoltage. The claim's premise (needing 10^6 V/m at the cell) is what is mistaken.

Verdict (split): - The numerical premise (that membrane influence needs 10^6 V/m, so one must beat air): Internally inconsistent (membrane field and air strength both printed as 10^6; real values 1.4e7-2.3e8 and 3e6) and Refuted by calculation (external field needed is about 5 kV/m, not 10^6 V/m). - Brief overvoltage in air without breakdown: Established as a general phenomenon; not shown for this device. - Pasteurization/kill of degenerate cells by pressure from induction bursts: Never properly tested for this device; the induced field estimate is about 200 times below known PEF thresholds, so the mechanism is not supported by the numbers I can compute; the inventor gives no model.

Claim 4. Cell constituents resonate with the plasma eigenfrequencies and selectively absorb the energy, driving selected endothermic chemical reactions (for example, nitrogen plasma excites nitrogen atoms in tissue).

a. Internal consistency. The mechanism chain is: plasma of element A emits radio at "eigenfrequencies of A"; atoms of element A in tissue absorb "due to resonance" and gain "energy surplus"; this energy "promotes an endothermal chemical reaction" [read: p.5]. Problems: - Quantum energy scale. A 100 MHz photon carries 4.1e-7 eV (calculation (d)). A bond-breaking or endothermic step needs of order 1 eV. The ratio is about 2.4 million to one. Resonance at a frequency does not change how much energy one quantum carries; nor can the energy of millions of photons be pooled into one bond, because the reactions the patent writes need the energy at one molecule within a vibrational relaxation time of picoseconds in water [memory, high]. The thermal energy of the surroundings (0.027 eV) is also 60,000 times larger than a single radio photon. A resonant line at radio frequency would be washed out by collisions and thermal motion. - Gas-specific lines. Atomic nitrogen's transitions are electronic (eV scale, optical and UV) [memory, high]; nuclear-spin resonance of 14-N is about 3.08 MHz per tesla, so at 0.1 T it is about 0.31 MHz (my arithmetic from the standard gyromagnetic ratio [memory, high]). That is a nuclear magnetic resonance (NMR) frequency and depends on the applied magnetic field and the chemical environment (chemical shifts of tens to hundreds of ppm). It is not tied to what element the plasma was made of: a plasma containing nitrogen emits no radio line at the NMR frequency. The idea of "plasma of A excites tissue A" has no mathematical derivation in the patent; the only support is a verbal analogy (like frequency, like absorption) between emitter and receiver. - Reactions (1) and (2), XAB + Y + energy A = XA + YB, XAB + Y + energy B = XB + YA [read: page image p.5], are bookkeeping of two outcomes. They are balanced in species but there is no energy value or selection rule that links "energy A" to a frequency. Energetically, endothermic means the products have more energy than reactants; the patent offers no number for either reaction or for the absorbed dose. Inventor's later restatement as NMR/EPR (US 2006/0220648 A1, [web, via sites.md]) changes the physics language but still gives no derivation of reaction selectivity from the Investigator's abstract. I did not see the full text of this later patent. - The claims (1, 17) assert "selective absorption of energy by atoms of said specific element ... due to resonance" with atoms in compounds treated as the free atoms of the plasma. Bound atoms have different resonance frequencies (chemical shifts, molecular vibrational and rotational modes), so the "similar atoms" assumption is hidden and unjustified. - The inventor's later extension to biological nuclear transmutation (Na + O -> K, "C to Fe", 30 keV energies, [web, via sites.md]) is an energy-scale leap: nuclear reactions need keV to MeV per event while radio photons carry 1e-7 eV. Not derived anywhere I could read; mass-and-charge balance checks (Na 23 + O 16 = 39 = K 39; Z 11 + 8 = 19) are numerologically neat but say nothing about mechanism or Coulomb barrier. Pertinent to record; not deriving. Fair statement of what is mathematically sound here: in an inductively coupled magnetic field, nuclear spins DO resonate at a field-dependent frequency (NMR/EPR), and MRI is an established application [memory, high]. The 2003 patent family's "multiple resonance" framing points at that. That is a different mechanism from "plasma of element X excites element X."

b. Consistency with experiment. Radio-frequency heating of tissue by dielectric loss is established; selective chemistry by radio frequency at these energy levels is not seen. Experiments in magnetic-field effects on chemistry (radical-pair mechanism) exist but act on spin states of radicals at milli-tesla fields and do not need element-matched emission [memory, medium].

c. Tested? Not in this patent. The inventor's later claims on potassium and transmutation (1999, 2006 file) are his own reports with no independent replication found by the Investigator. Viti et al. 2023 and Keilani et al. 2025 measure heart-rate variability and fatigue scales, not chemical selectivity [web via sites.md].

d. If true: absorption of a 4e-7 eV quantum would have to drive eV-scale chemistry. That would require a collective, coherent energy-pooling mechanism unknown to physics (or a nonlinear multi-photon process at power densities far beyond the 80 J per pulse delivered over 30 x 30 cm). Transmutation would require nuclear physics changes.

Verdict: Refuted by calculation for the stated mechanism (radio-frequency photon energy of order 1e-7 eV cannot supply endothermic reaction energy of order 1 eV by single-quantum resonance; element-matching between plasma and tissue is not derived). The weaker statement "pulsed magnetic fields may influence tissue by some other mechanism" is a different claim (see claim 5). Later transmutation extensions: Never properly tested (no independent controlled test on record) and, on the energy scale, Refuted by calculation unless a new mechanism is offered.

Claim 5. Restoring transmembrane potential (cancer about 15 mV, aged or unhealthy about 50 mV, healthy about 70 mV) is therapeutic, and the pulses can also pasteurize degenerate cells or microorganisms by pressure increase.

a. Internal consistency. The numbers (15, 50, 70 mV) are cited as "prior art" without references [read: p.4]. The scale is plausible for resting potentials (about -70 mV for many cells; depolarised in proliferating or cancer cells) [memory, medium]; the patent states only magnitudes. The logic "low potential causes pain signals" and "weak bursts increase potential by ion conduction across the membrane" [read: p.4] is qualitative. No calculation of how much current crosses the membrane per pulse or how that moves the potential. A quick consistency estimate (mine): membrane capacitance is about 1 uF/cm^2 [memory, high]; changing V_m by 55 mV (15 to 70 mV) needs charge of 5.5e-8 C/cm^2, a small number, which is why even weak fields can transiently shift potential. Yet this shift relaxes within the membrane time constant (microseconds to milliseconds) as ion pumps and leaks restore the resting level, so a net lasting restoration is not demonstrated by any calculation here. The patent gives no dwell-time argument. The pressure-burst "pasteurization" is the same unquantified mechanism assessed under claim 3. Also a logical gap: claim 5 mixes two opposite directions of action (raise V_m of low-potential cells, but also burst/kill degenerate cells). The patent assigns "strong bursts" to killing and "weak bursts" to potential raising without giving threshold values for either.

b. Consistency with experiment. Pulsed electromagnetic fields have recognised effects: FDA-cleared PEMF devices for bone healing and for postoperative pain exist [memory, medium]. Small effects of pulsed fields on autonomic measures (heart rate variability, HRV) were found in sham-controlled PAPIMI trials (below). Transmembrane-potential restoration was not measured in any source I read.

c. Tested? By whom, with what result (all [web] via sites.md, not re-verified by me): - Viti et al. 2023, Int J Environ Res Public Health 20(5):3934, PMC10001461: single-blind randomised sham-controlled pilot, n = 32 (17 PAPIMI, 15 sham), chronic musculoskeletal pain; HRV time-domain indices and high-frequency power rose after one session, no change after sham; authors declare no conflict. Small sample, an autonomic marker, not pain or tissue healing, not transmembrane potential. - Viti et al. 2025, Healthcare PMC12385873: single-arm pilot, n = 30, pain score fell by median 2 points; no sham arm. The manufacturer's site promotes it (interested party). - Keilani et al. 2025, Wien Klin Wochenschr, PMC12534222: randomised, single-blinded, placebo-controlled pilot, n = 20, post-COVID fatigue, 10 sessions; improved 6-minute walk and fatigue scales; authors declare no competing interests. - Leisner et al. 2002 (rat fracture): negative for bone healing. - FDA record (interested regulator): 510(k) not cleared in 1995 for lack of clinical data; 2002 inspection and 2005 letter report adverse events including "one fatal myocardial infarction during treatment", rat deaths, and the inspector's "the science behind the device is not clear". The Seattle Times 2007 report quotes Pappas ("can cure cancer and AIDS") and attributes deaths; the paper and regulators have their own stakes. User-reported testimonials (Wallach) are the interested parties' own claims. They are cited here without weighting. No source tests the specific claim: no one measured membrane potential before and after exposure, and no controlled cancer trial exists in the record. Small positive controlled pilots exist for pain and fatigue and one negative controlled animal study for bone. Replication: positive pilot signals come from two Italian and Austrian groups with different outcomes; none is a replication of the same endpoint. A sample of n = 20-32 gives wide uncertainty (for a binomial-type endpoint, a 95% interval of about plus or minus 20 percentage points at n = 25, my arithmetic from the standard formula 1.96 sqrt(p(1-p)/n) at p = 0.5).

d. If true: no change to mainstream physics for "pulsed magnetic stimulation affects excitable tissue." Restoring tumour transmembrane potential as a cancer therapy would need biology and clinical evidence rather than new physics.

Verdict (split): - That PAPIMI-type pulses produce measurable physiological effects (HRV, pain or fatigue scores) in small controlled trials: Plausible, untested at the level that matters (small sham-controlled pilots only; no replication of the same endpoint, no mechanism shown). - That the effect works by restoring transmembrane potential (15 -> 70 mV): Never properly tested (never measured; not dismissed without testing, simply not examined). - Cancer treatment, pain "generation" explanation and cold pasteurisation claims: Never properly tested in controlled trials; the record includes unregulated use and FDA adverse-event findings (interested regulator). - Bone healing: Tested, not supported (Leisner 2002, one controlled animal study).

Notation, units and hidden-assumption list (for the record)

• "Gain" as a ratio of resistances: dimensionless, not a power gain; division by an infinite quantity.
• "Duty cycle of 0.001%" combined with "average current": makes peak/average about 1e5 by construction.
• Units: 10 kV/cm = 1e6 V/m is correct; 70 mV per several Angstrom is 2.3e8 V/m, not 1e6; claim 5 mixes "diameter" and "square millimeters"; claim 6 versus the 30 cm coil; burst-duration ranges differ between text and claims 14 and 16.
• Hidden assumptions: (i) that membrane effect requires an external field equal to the intra-membrane field (it does not, by a factor of about 3000); (ii) that bound atoms respond at the free-atom frequency of the plasma element; (iii) that circuit or probe ringing is a "gas eigenfrequency"; (iv) that a quantum of 4e-7 eV can drive eV-scale chemistry; (v) that stated oscillogram statistics identify a spectral line.
• Missing quantities that make the patent untestable as a derivation: capacitance, charging voltage, inductance, peak current, stored energy, gas pressure, electrode gap, vertical scale of FIG. 2 and 3.

Investigator's findings kept in this review (whatever the verdict)

Patent family (WO 94/01176, EP 0648140 B1 granted 2001, RU 2126278 C1 with 24 claims, Greek priority GR920100309A; the US text prints 920200309, a likely misprint); the earlier lightning-conductor and spark-gap oscillator patents (EP 0402552 and Greek patents, 1988-1991) that predate the medical use; the 1989 Cure-Pappas-Eichler ion-transport patents US 5,068,039 and US 5,239,442 (ideal for Nordenstrom's method); the 2003 NMR/EPR family (US 7,151,372 B2); the Physiology of the Cell PDF (">1 GW", "888 pulses at about 80 J", Na -> K); the 1998 New Energy News letter ("C to Fe", "20% excess heat"); the physics publications (Nuovo Cimento B 76, 189 (1983) on the Ampere force and Biot-Savart-Lorentz forces, cited 151 times per Google Scholar; Moyssides & Pappas experiment at the Athens NTU, 1985-86); the IECEC 1991 paper not yet obtained; Liboff et al. US 5,160,591 (ion cyclotron resonance, the closest mainstream-adjacent selective-resonance idea, which does have a stated frequency-field relation f = qB/(2 pi m) [memory, high], unlike this patent); the citing patents (Horkel DE 10 2007 030325 near copy; Healing Machines; Biofields; Pulsartec); the clinical papers (Viti 2023 and 2025, Keilani 2025), Leisner 2002; the FDA and Seattle Times record; the Karampatsou memoriam with claimed use in Cuba, Vietnam and Switzerland (interested collaborator; no study cited); the continued CE-marked European availability, so the record did not go dark; the inventor's death date (11 May 2026, single Facebook source) and his children's continuation of the work. Absence of evidence worth flagging for the other reviewers: the IECEC 1991 paper and the ICCF-7 "Electrically induced fusion in the living cell" paper are the places where the numbers behind claim 1 and the strong claims would be shown; neither has been read. Any experimental weighing of claim 1's strong form turns on them.

Summary of verdicts

1. Claim 1: split. Weak form (spikes above average) Established; strong form (net energy gain via resistance ratios) Refuted by calculation as an argument and Never properly tested as an experiment.
2. Claim 2: split. Sub-microsecond bursts at extrema of a damped ring Established; gas-specific eigenfrequencies Never properly tested (plasma-frequency reading implies n_e ~ 1e8 cm^-3, implausible; circuit ringing not excluded).
3. Claim 3: Internally inconsistent premise (10^6 V/m for both membrane and air; real 1.4e7-2.3e8 and 3e6) and Refuted by calculation (needed external field about 5 kV/m); pasteurization by induction bursts Never properly tested.
4. Claim 4: Refuted by calculation (photon energy 4e-7 eV vs about 1 eV; no derivation of element-matched absorption); transmutation extension Never properly tested.
5. Claim 5: split. Small controlled physiological effects Plausible, untested (pilots only); transmembrane-potential mechanism and cancer therapy Never properly tested; bone healing Tested, not supported (Leisner 2002).