Rédiger un PREreview

Narrative review · Preprint · Version 1.2 · Not peer reviewed · Submitted to IEEE Reviews in Biomedical Engineering, 18 September 2026; under editorial consideration, not accepted.

Electrical stimulation, pulsed electromagnetic fields, photobiomodulation, therapeutic ultrasound and controlled heating are developed, regulated, reviewed and taught as separate literatures. They share one structure — a source delivers energy across a coupling interface into tissue — and therefore one vulnerability: if the quantity that actually reaches the tissue is not reported, the literature cannot be replicated, pooled or refuted.

This review compares the five families on a single axis: what quantity is delivered, whether it is reported, and whether the completeness of that reporting tracks the strength of the evidence base.

What each field solved, and what it left open. Hyperthermia has a single agreed dose metric whose reference temperature its own originators described as arbitrarily chosen. Photobiomodulation has a measured, wavelength-specific, non-monotonic dose–response — the optimal radiant exposure differs by a factor of 10–100 between 810 nm and 980 nm in the same cell type — alongside a published audit finding its own clinical trial reporting too incomplete for replication. Electrical stimulation has mature international standards for safety limits and almost none for delivered dose. Acoustic energy has the class’s most rigorous sham-controlled trial, and its result was null.

Contents. A generalised transfer chain with the exposure quantity named for each modality · the five modality sections · a cross-modality comparison of reporting completeness and evidence position, in which every row carries a genuine negative result · a graded assessment of mechanistic standing · and a thirteen-item minimum reporting set applicable to any physical-energy intervention. Six tables, three figures, 42 verified references.

What this preprint does not do. It makes no efficacy claim for any modality and no claim of novelty for the comparison itself; the prior-art search supporting that statement, and its limits, are reported in §11.2. No device, technology or framework associated with the author appears anywhere in the manuscript except in the competing-interests declaration — deliberately, because the paper’s value depends on being independent of any particular device.

Status. Version 1.0 was graded RED by an adversarial internal review and corrected. The change log at §12.4 lists fourteen corrections. The most consequential: a genuine phase 3 negative trial of tumour-treating fields was missing from the comparison table while a design limitation stood in its place; the tumour-treating-fields dose–outcome analysis was presented without its confounding; a category-F mechanism was written as category A in the mechanism section; and the physical relationship for charge per phase was stated without the voltage-controlled case, in which delivered charge is not computable from the settings at all.

Competing interests. The author is an electronics engineer, founder and Director of Rohera Healthcare & Technology Pvt. Ltd., Pune, India, and a named inventor on a patent family relating to electrical stimulation technology. He therefore holds a financial and intellectual interest in the electrical-stimulation literature reviewed in §3. The mitigations applied to the manuscript, and the four places where an adversarial review found that interest had produced a softer reading, are set out in §11.4.

Funding. Self-funded. This work received no external, grant or institutional funding; the costs were borne by the author and his companies, and the company provided no funding specific to this manuscript and had no role in its design, conduct, analysis, drafting or the decision to submit. Self-funding is declared here because it is itself a competing interest: no external sponsor reviewed, constrained or independently verified this work.

Version 1.2 (this version). A second adversarial internal pass over version 1.1 found a further set of defects, all corrected here and listed in §12.5. The most serious: the paper described its own method inaccurately, calling purposive selection “systematic” and claiming a deposited search log that did not exist (the actual log is now printed in §11.5); §3.3’s treatment of frequency was wrong in three ways, including being refuted by the paper’s own §3.4; and §8.4’s central association rests on two axes graded by the same author from the same reading — measurement dependence, not confounding — and is regraded from B to F accordingly. The paper now also states plainly that no author is independent of the company.

Journal submission status (18 September 2026). A version of this manuscript, condensed to the journal's page limit, has been submitted to IEEE Reviews in Biomedical Engineering as a Methodological Review. It is under editorial consideration; it has not been peer reviewed and has not been accepted. Statements elsewhere on this record that the work had not been submitted to a journal were accurate when written and are superseded by this notice.

Copyright notice (required by IEEE policy on electronic posting). Copyright may be transferred without notice, after which this version may no longer be accessible.