A Whole-Body Sensory-Motor Gradient is Revealed in the Medial Wall of the Parietal Lobe

Zeharia, Hofstetter et al. (2019) — A new homunculus in the precuneus

An fMRI mapping study that resurrects a forgotten prediction. In Epilepsy and the Functional Anatomy of the Human Brain (1954), Penfield and Jasper inferred, from cortical stimulation in epileptic patients, that a second whole-body sensory representation — a “supplementary sensory” area, mirror to the supplementary motor area (SMA) — should sit in the medial wall of the parietal lobe, in the precuneus. They drew it in dashes and disclaimed it (“our observations are not yet numerous enough to justify conclusions as to form or exact position”), and the field let it lapse. Zeharia et al. test it directly and find it.

This is the “Maps” folder source, and it earns its place in the wiki by being the other kind of body map — a cortical somatotopic homunculus, not a felt bodily sensation map. Holding those two senses of “body map” apart is a job this paper does well (see the distinction developed on both concept pages).

What they did

• 16 healthy right-handed subjects (8 women, 25–35) moved 20 body parts in a fixed toes-to-tongue order (after Penfield & Boldrey 1937), blindfolded, in the scanner. A periodic (“phase-encoded” / traveling-wave) design — the same body-part sequence cycled repeatedly — let them recover, for each voxel, the phase of the response, i.e. which body part that voxel prefers. To control for order/attention effects, “toes-to-tongue” and “tongue-to-toes” runs were averaged.
• A separate event-related experiment (randomized single-body-part movements) gave a GLM cross-check that spectral analysis cannot: it exposes the overlap between body-part representations.
• Linear regression of phase value against x/y/z Talairach coordinates quantified the gradient’s axes and direction.
• A separate group of 46 subjects provided resting-state data for body-part-specific functional connectivity (seed ROIs defined by phase value).

The result

A whole-body gradient in the medial parietal wall (Fig. 1B, gradient “1”), anterior-to-posterior, dorsal-to-ventral, toes-to-tongue, lying in mirror orientation to the SMA gradient with the M1 leg area between them — exactly the geometry Penfield and Jasper sketched (Fig. 1A vs. their revised model, Fig. 1F). It is visible in single subjects with and without spatial smoothing (so not a smoothing artifact), and the y-axis regression is strong (R = −0.74 to −0.82 across hemispheres). The event-related GLM confirms leg/hand/face representation in the same territory, with the honest wrinkle that the face representation is weaker than leg and hand, and hand/face peaks overlap and stretch the full anterior-posterior length.

The connectivity story is the paper’s most interesting move for this wiki. Rather than carving the precuneus by anatomy (as Margulies et al. 2009 did — anterior sensorimotor, middle cognitive, posterior visual), they carve it by body-part preference and ask where each part connects. The answer recovers Margulies’ anterior→posterior motor→cognitive→visual shift as a by-product of somatotopy: leg (anterior) → motor cortex + SMA + insula; hand (middle) → parietal + premotor + cingulate; face (posterior) → occipital visual cortex + MT + LO. And a novel shared thread: every body part connects to the extrastriate body area (EBA), the region that lights up when you see a body part — hinting that this motor/sensory map is wired to visual body-part representation.

Why an interoception wiki keeps it

Bluntly: this is not an interoception paper. The homunculus it maps is exteroceptive/somatomotor — body surface and movement — not visceral. Three threads justify ingesting it rather than flagging it as off-topic:

1. The two-body-maps distinction. The wiki’s bodily-sensation-maps page is about felt topographies of emotion (Nummenmaa’s emBODY maps). This paper is about a cortical topography of the body surface. They share the word “map” and nothing else — and the wiki has already flagged how easily “fingerprints”/“maps” vocabulary gets conflated across literatures. The cortical-somatotopy page exists to keep the line sharp.
2. The precuneus is a DMN hub. The wiki meets the precuneus repeatedly — in Farb (2010)‘s midline self-referential network, in Lutz (2008)‘s “mentation” network, in Barrett’s internal model. This paper shows the same structure also carries a somatotopic body map. The precuneus is not only a self-processing/mentalizing region; part of it is body-mapped. See precuneus.
3. The insula homunculus, and the authors’ own closing bet. The paper’s discussion links this new precuneus homunculus to the insula homunculus the same group reported (Zeharia et al. 2012) — a topographic body map in the interoceptive cortex — and speculates that the two form a putative “node linking body and mind.” They explicitly invoke mind-body interventions that “enhance emotional well-being… by focusing our attention on the body,” coupling self-referential and body-related representation. That is the wiki’s insula / mindfulness / interoception thread, reached from the somatomotor side. It is speculation (the authors say so — “future studies must study these questions in a more direct and controlled manner”), and it is recorded here as speculation, not evidence.

What it does not show

No visceral or interoceptive measurement of any kind; no felt experience; no clinical or emotional variable. The “sensory vs motor” ambiguity is unresolved by design. And the insula/precuneus “body-mind node” is a closing conjecture, not a result — the connectivity analysis links leg-precuneus voxels to the insula, but says nothing about interoceptive function. Treat the interoceptive relevance as thematic and analogical, foregrounded wherever this study is cited.

Provenance notes

• DOI 10.1523/JNEUROSCI.0727-18.2019 is printed on the paper (no derivation needed).
• Zeharia and Hofstetter contributed equally (joint first authors). Senior/corresponding author Amir Amedi (Hebrew University; known for sensory-substitution and cortical-plasticity work) and co-authors Tamar Flash (Weizmann, motor control) are recorded here without researcher pages — first appearance in the wiki, single paper, held to the co-author convention (as with Wiens under Katkin, Jia under Tong). Flag for a page if the Amedi/Flash body-mapping line recurs first-hand.
• The phase-encoded/periodic (“traveling-wave”) fMRI design and its spectral analysis are described inline rather than given a method page — a specialized neuroimaging technique that would be an orphan on one study. Noted here so a future ingest with a second phase-encoded study can promote it.