Fifteen of 17 tested responsive SEEG pairs showed compressive temporal summation; patient-level tests were underpowered (3 patients)

We tested if SEEG depth contacts show 2 temporal non-linearities of visual broadband (70-170 Hz) responses: compressive temporal summation (CTS) and suppression of the second of 2 pulses at short gaps. We used the raw files of OpenNeuro ds004194, task temporalpattern, first session, all 32 runs of the 8 patients with SEEG contacts. A plan was written before the signals were read. We used only good SEEG contacts, made bipolar pairs of adjacent contacts on the same shaft (175 pairs), and calculated broadband power. Odd trials selected the responsive pairs and even trials tested them. 19 of 175 pairs were responsive, in 3 of 8 patients (p05: 16, p07: 2, p10: 1). H1 (log-log slope of response against duration < 1): patient slopes 0.72, 0.04 and 0.22, p = 0.125 (the smallest possible p for n = 3), Holm p = 0.25. H2 (response at ISI 533 ms > response at ISI 17 ms): patient values +1.26, +0.42 and -3.23, p = 0.625. At the pair level (descriptive, pairs are not independent), 15 of 17 slopes were less than 1 (median 0.69), and the median ratio R(533 ms) / (4 R(133 ms)) was 0.64. In the grand time course, the second-pulse response was 41 % at ISI 267 ms and 70-73 % at ISI 533 ms, against a first-pulse peak of 140-160 %. Thus the pairs show CTS and recovery with ISI, but mostly in 1 patient. The data set does not have sufficient patients with visual SEEG responses for a patient-level test. Limits: no anatomical labels, 6 test trials per condition, and a 1 s window that cuts the late second-pulse response. The package has the plan, code, per-trial and per-pair results, and SHA-256 hashes of all 343 input files.

SEEG · visual cortex · broadband power · compressive temporal summation · temporal summation · OpenNeuro