Write a PREreview

Version of record: version 1.2, DOI 10.5281/zenodo.22828656. This version is first-revision text and is not the version under review at the journal. See Submission status below.

Version 1.4 (22 September 2026). Two corrections. No value, table entry, figure, reference, grading or conclusion changes. Both were found by a check run across the three manuscripts of this programme rather than within any one of them.

§11.4 claimed that “no technology associated with the author appears anywhere in the paper.” The affiliation line names the author’s patent family, and this manuscript’s own Declarations say so two pages later — “outside this Declarations section and the author’s affiliation line, where the company and the patent family are necessarily named”. The manuscript therefore carried two statements of one fact, one scoped and one not, and the unscoped one was false by the scoped one’s own admission. §11.4 is now scoped to match. The underlying fact is unchanged and verified line by line: outside the change log, the brand name does not occur at all, the framework name occurs once and in the affiliation line, and the company name occurs three times — affiliation, competing interests, funding. The body of the paper carries none.

The same defect had been corrected in the companion manuscript on 21 September, and the correction did not travel. Two things let it through, and both are recorded in §12.7 because they are more useful than the sentence.

First, the checker’s pattern was written against the wording the sibling happened to use. It matched “anywhere in this” and this manuscript says “anywhere in the paper” — the same claim, one preposition away.

Second, and more consequential: the claim had a verdict, and the verdict was earned by a different clause. The resolution ledger that guards these manuscripts keys on the sentence, and this sentence makes three separate universal claims. One fingerprint, one verdict. The recorded evidence adjudicated the Table 3 clause and said nothing about the technology clause, which inherited a verified verdict it had never been given. Of the 77 claim-sentences surfaced across the three manuscripts, eight carry more than one distinct universal; in seven the clauses concern the same fact, and in this one they did not. The checker now fails any multi-universal sentence whose register entry does not state that every one of its claims was examined.

The second correction is minor and is recorded for completeness: the end-of-manuscript line read “End of manuscript version 1.2” in a version 1.3 file and pointed at a change-log range that no longer covered the last entry.

Submission status. Version 1.2 remains the version submitted to IEEE Reviews in Biomedical Engineering on 18 September 2026 under manuscript number RBME-00307-2026, and is the version of record with the journal until a revision is invited. This version is the first-revision text.