Melody processing recruits a distinct generator in anterior Heschl’s gyrus

Evidence

The shorter melody experiment presented at the start of this paper shows how the classic energy-onset generator in PT can be separated from the melody generator in the anterior part of HG. The transients obtained from the melody generator contain both N1m and P2m components, and both components are larger in melodic sequences than they are in tonic sequences. Moreover, the N1m peak delay is shorter in melodic sequences. Reanalysis of the TR data from the four-bar melody experiment suggests that the N1m and P2m components of the response in HG can be separated with a 4-dipole source model, and that melody processing is more readily observed in the shape and magnitude of the P2m response. Finally, the sustained field originates from a generator in HG not far from the melody generator; however, the SF itself does not distinguish between the tonic and melodic sequences. — Dijk, Pim van, Physiology, Psychoacoustics and…, 4728

Recently, Patterson et al. ( 2015 ) extended earlier imaging research on melody processing (Patterson et al. 2002 ; Rupp and Uppenkamp 2005 ; Uppenkamp and Rupp 2005 ) by presenting listeners with four-bar musical phrases played by a French horn. There were four notes per bar; they had the same amplitude and the inter-note gaps where minimized to minimize the energy onset component of the responses to the second and succeeding notes 2–4 (Krumbholz et al. 2003 ). In the first and third bars, the pitch was restricted to the tonic of the four-bar phrase; the second bar contained a novel, random tonal melody and the fourth bar contained a variation of that melody. Magnetoencephalographic recordings revealed that each note produced a prominent transient response (TR) in auditory cortex with a P1m, an N1m and a P2m. The structure of the four-bar phrase was reflected in the magnitude of the TR. In all bars, the TR to the first note was somewhat larger than that to the remaining three notes of that bar. In the melody bars, the P2m component of the TR was larger than in bars where the pitch was fixed. The stability of the TR magnitude to notes 2–4 of each bar made it possible to do a four-dipole analysis and show that the activity is distributed between two sources in each hemisphere, one in the medial part of Heschl’s gyrus (HG) and the other slightly more lateral and posterior on the front edge of planum temporale (PT). The source waves showed that melody processing was largely confined to the P2m component of the TR in the medial part of HG. — Dijk, Pim van, Physiology, Psychoacoustics and…, 4704

Categorization

• XREF: Connects to broader literature on functional segregation within auditory cortex (parabelt/core organization), and to the reader’s interest in where spectrotemporal analysis yields to pattern/melody processing.
• DEFINE: Clarifies PT (planum temporale) vs anterior HG (Heschl’s gyrus) as distinct generator sites for energy onset vs melodic processing.
• corroborated: 2 independent passages support this claim — Dijk, Pim van, Physiology, Psychoacoustics and…, 4728; Dijk, Pim van, Physiology, Psychoacoustics and…, 4704