A follow up to Star Face — a survey of existing evidence bearing on the STAR FACE census, and a proposal
by Grego (Gergely Földvári)
STAR FACE is a sieve model demonstrated on a polyhedron, introducing the term summetry, the discrete unity of sum and symmetry. The sieve descends from a population well over the number of atoms in the Solar System to 6,223,974 magic readings, the red cells in a single microlitre of blood, and terminates in a singularity: one numbering, unique up to the symmetry of the solid. Each of the twelve four-faced pyramids stands for one diagnostic reading; when two or more pyramids carry the same four values, one reading masks another or more that could be distinct and clear. This paper does four things. It separates the several unrelated senses the word masking carries in medicine. It corrects the statistic that a clinical comparison requires, replacing the census by shape with a census by the depth of the deepest mask. It surveys what the literature already contains at each masking depth, and finds the diagnostic odyssey of rare disease to be the only body of evidence reaching past depth two. And it sketches, for others better qualified to finish it, the survey that would test the model's one genuinely untested claim.
The book's census groups numberings by shape, counting the blocks of pyramids that share a reading. That is the correct combinatorial record, and it is not the statistic a clinician needs. A numbering holding one masked triple hides three readings behind one appearance; one holding a masked pair hides two. Grouped by shape both count as one block, and the clinical meaning is destroyed. Regrouped by the size of the largest mask:
| Largest mask | Numberings | Share |
|---|---|---|
| 1 reading, all clear | 3,538,104 | 56.846381 % |
| 2 readings coincide | 2,644,702 | 42.492176 % |
| 3 readings coincide | 41,151 | 0.661169 % |
| 4 readings coincide | 17 | 0.000273 % |
| total | 6,223,974 | 100 % |
Before any comparison can be made, one obstacle has to be cleared. In medicine masking means at least six different things, and the confusion is the reason the evidence has never been gathered in one place. Three of them are this model's sense: diagnostic overshadowing, where the mask is a prior label; pharmacological masking, where the mask is the therapy; and symptom-disease pairing, where a benign label conceals a dangerous disease. Three are not: blinding in a clinical trial, the tone played into one ear during audiometry, and the concealment of autistic or attention-deficit traits.
Doctors already have a name for what happens when a patient's existing label swallows their new symptoms. But nobody ever counts. No study states how many other conditions were hiding behind the label. Four literatures describe one phenomenon, none cites the others, and none asks how many readings coincide.
People eventually diagnosed with a rare disease are given, on average, two or three misdiagnoses before the correct one. The EURORDIS Rare Barometer surveyed 6,507 people living with 1,675 rare diseases across 41 countries and found an average diagnosis time of 4.7 years. A patient carrying two or three successive wrong labels before the true one is a masked reading of depth three or four. This is the zebra of the book's own aphorism, and here it is counted.
The honest limitation is stated in the paper: census masking is simultaneous, the odyssey is sequential. And every dataset that displays deep masking is drawn from cases already known to have been hard — a confirmed rare disease, a referral to a quaternary centre, a death followed by autopsy, a subsequent adverse event. The census is unselected. Correcting for that selection brings the rare-disease evidence into the same order of magnitude as the census figure of 0.661169 per cent.
Nothing found contradicts the census. Nothing confirms it either. The paper sets out, in full and ready to copy, the physician survey whose categories are taken from the census rather than invented: for a consecutive series of presentations, how many conditions remained genuinely compatible, whether the deepest ambiguity stood alone, and what concealed what.
The author is a musician, not a physician. It is way beyond his competence to even remotely suggest that such a survey would be complete or best worded, and it would be a grave mistake not to emphasise that it is a mere sketch, a skeleton, a kind of kaleidoscope to take as a basis: a viewpoint from which a proper and professional study may be developed.
The zebra is hiding, and patients are dying because of the masking depths. It is time to employ the data processing powers available to mankind: to screen, to sieve, to save them.
Star Face is the model this paper follows, with its companion animations at grego.hu/star. Summetry sets out the arithmetic the model rests on. See also The Tabakgasse Polyhedron and GooglyCube.