Limits...
Distribution, Amplitude, Incidence, Co-Occurrence, and Propagation of Human K-Complexes in Focal Transcortical Recordings(1,2,3).

Mak-McCully RA, Rosen BQ, Rolland M, Régis J, Bartolomei F, Rey M, Chauvel P, Cash SS, Halgren E - eNeuro (2015)

Bottom Line: KCs were marked manually on each channel, and local generation was confirmed with decreased gamma power.Locally generated KCs were found in all sampled areas, including cingulate, ventral temporal, and occipital cortices.These results open a novel view where KCs overall are universal cortical phenomena, but each KC may variably involve small or large cortical regions and spread in variable directions, allowing flexible and heterogeneous contributions to sleep homeostasis and memory consolidation.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Neurosciences, University of California, San Diego , San Diego, California 92093.

ABSTRACT
K-complexes (KCs) are thought to play a key role in sleep homeostasis and memory consolidation; however, their generation and propagation remain unclear. The commonly held view from scalp EEG findings is that KCs are primarily generated in medial frontal cortex and propagate parietally, whereas an electrocorticography (ECOG) study suggested dorsolateral prefrontal generators and an absence of KCs in many areas. In order to resolve these differing views, we used unambiguously focal bipolar depth electrode recordings in patients with intractable epilepsy to investigate spatiotemporal relationships of human KCs. KCs were marked manually on each channel, and local generation was confirmed with decreased gamma power. In most cases (76%), KCs occurred in a single location, and rarely (1%) in all locations. However, if automatically detected KC-like phenomena were included, only 15% occurred in a single location, and 27% occurred in all recorded locations. Locally generated KCs were found in all sampled areas, including cingulate, ventral temporal, and occipital cortices. Surprisingly, KCs were smallest and occurred least frequently in anterior prefrontal channels. When KCs occur on two channels, their peak order is consistent in only 13% of cases, usually from prefrontal to lateral temporal. Overall, the anterior-posterior separation of electrode pairs explained only 2% of the variance in their latencies. KCs in stages 2 and 3 had similar characteristics. These results open a novel view where KCs overall are universal cortical phenomena, but each KC may variably involve small or large cortical regions and spread in variable directions, allowing flexible and heterogeneous contributions to sleep homeostasis and memory consolidation.

No MeSH data available.


Related in: MedlinePlus

Observed manual KCs versus those expected under the  hypothesis of independent occurrence, for representative nights for each subject. A, KC events observed for each subject were classified into two groups: KCs with one participating channel (light orange); and KCs with two or more participating channels (dark orange). B, KC events derived from simulating an expected distribution for each subject were classified into the same two groups: one channel (light purple); and two or more channels (dark purple). For all subjects, there are more KCs that occur only in one channel expected under the  hypothesis than were actually observed. These distributions are tested statistically using a Fisher’s exact test, and these results are listed in Table 3.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4596022&req=5

Figure 6: Observed manual KCs versus those expected under the hypothesis of independent occurrence, for representative nights for each subject. A, KC events observed for each subject were classified into two groups: KCs with one participating channel (light orange); and KCs with two or more participating channels (dark orange). B, KC events derived from simulating an expected distribution for each subject were classified into the same two groups: one channel (light purple); and two or more channels (dark purple). For all subjects, there are more KCs that occur only in one channel expected under the hypothesis than were actually observed. These distributions are tested statistically using a Fisher’s exact test, and these results are listed in Table 3.

Mentions: To test whether these observed KC co-occurrences were significant, we simulated a set of expected KC data under the hypothesis that the KC times are unrelated in different channels. This expected distribution was created by randomizing the inter-KC intervals within channels of the observed KCs and grouping the latencies over channels within 200 ms windows, as was done with the observed KCs. We examined observed and expected KCs occurring in one channel versus two or more channels for each subject (Fig. 6). In the observed KCs, there are KCs that occur only in one channel (Fig. 6A, light orange). However, there are many more KCs that occur only in one channel for the distribution (Fig. 6B, light purple). Conversely, there is a larger number of KCs involving two or more channels for the observed data (Fig. 6A, dark orange) than under the hypothesis (Fig. 6B, dark purple). To statistically verify this observation, a Fisher’s exact test was performed, comparing the number of observed KCs to the distribution in one channel versus two or more channels. For 20 nights over eight subjects, this was highly significant (p < 0.05, Bonferroni corrected; Table 3). For 4 nights over two subjects (3 of 4 nights for one subject and 1 of 3 nights for the other subject), it was not. Due to the variable number of channels for each subject, the “one channel” versus “two or more channels” comparison was the only comparison testable for all nine subjects. In eight subjects, it was possible to repeat this test using expected and observed data for one channel and three or more channels; for seven of these subjects at least 1 night was significant at p < 0.05 (Bonferroni corrected), while the night of the eighth subject was not significant. For six subjects, it was possible to test up to four or more channels; in this case, we find all nights for five subjects to be significant at p < 0.05 (Bonferroni corrected) and not in 1 night in one subject. This analysis demonstrates that KCs co-occur across the cortex.


Distribution, Amplitude, Incidence, Co-Occurrence, and Propagation of Human K-Complexes in Focal Transcortical Recordings(1,2,3).

Mak-McCully RA, Rosen BQ, Rolland M, Régis J, Bartolomei F, Rey M, Chauvel P, Cash SS, Halgren E - eNeuro (2015)

Observed manual KCs versus those expected under the  hypothesis of independent occurrence, for representative nights for each subject. A, KC events observed for each subject were classified into two groups: KCs with one participating channel (light orange); and KCs with two or more participating channels (dark orange). B, KC events derived from simulating an expected distribution for each subject were classified into the same two groups: one channel (light purple); and two or more channels (dark purple). For all subjects, there are more KCs that occur only in one channel expected under the  hypothesis than were actually observed. These distributions are tested statistically using a Fisher’s exact test, and these results are listed in Table 3.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4596022&req=5

Figure 6: Observed manual KCs versus those expected under the hypothesis of independent occurrence, for representative nights for each subject. A, KC events observed for each subject were classified into two groups: KCs with one participating channel (light orange); and KCs with two or more participating channels (dark orange). B, KC events derived from simulating an expected distribution for each subject were classified into the same two groups: one channel (light purple); and two or more channels (dark purple). For all subjects, there are more KCs that occur only in one channel expected under the hypothesis than were actually observed. These distributions are tested statistically using a Fisher’s exact test, and these results are listed in Table 3.
Mentions: To test whether these observed KC co-occurrences were significant, we simulated a set of expected KC data under the hypothesis that the KC times are unrelated in different channels. This expected distribution was created by randomizing the inter-KC intervals within channels of the observed KCs and grouping the latencies over channels within 200 ms windows, as was done with the observed KCs. We examined observed and expected KCs occurring in one channel versus two or more channels for each subject (Fig. 6). In the observed KCs, there are KCs that occur only in one channel (Fig. 6A, light orange). However, there are many more KCs that occur only in one channel for the distribution (Fig. 6B, light purple). Conversely, there is a larger number of KCs involving two or more channels for the observed data (Fig. 6A, dark orange) than under the hypothesis (Fig. 6B, dark purple). To statistically verify this observation, a Fisher’s exact test was performed, comparing the number of observed KCs to the distribution in one channel versus two or more channels. For 20 nights over eight subjects, this was highly significant (p < 0.05, Bonferroni corrected; Table 3). For 4 nights over two subjects (3 of 4 nights for one subject and 1 of 3 nights for the other subject), it was not. Due to the variable number of channels for each subject, the “one channel” versus “two or more channels” comparison was the only comparison testable for all nine subjects. In eight subjects, it was possible to repeat this test using expected and observed data for one channel and three or more channels; for seven of these subjects at least 1 night was significant at p < 0.05 (Bonferroni corrected), while the night of the eighth subject was not significant. For six subjects, it was possible to test up to four or more channels; in this case, we find all nights for five subjects to be significant at p < 0.05 (Bonferroni corrected) and not in 1 night in one subject. This analysis demonstrates that KCs co-occur across the cortex.

Bottom Line: KCs were marked manually on each channel, and local generation was confirmed with decreased gamma power.Locally generated KCs were found in all sampled areas, including cingulate, ventral temporal, and occipital cortices.These results open a novel view where KCs overall are universal cortical phenomena, but each KC may variably involve small or large cortical regions and spread in variable directions, allowing flexible and heterogeneous contributions to sleep homeostasis and memory consolidation.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Neurosciences, University of California, San Diego , San Diego, California 92093.

ABSTRACT
K-complexes (KCs) are thought to play a key role in sleep homeostasis and memory consolidation; however, their generation and propagation remain unclear. The commonly held view from scalp EEG findings is that KCs are primarily generated in medial frontal cortex and propagate parietally, whereas an electrocorticography (ECOG) study suggested dorsolateral prefrontal generators and an absence of KCs in many areas. In order to resolve these differing views, we used unambiguously focal bipolar depth electrode recordings in patients with intractable epilepsy to investigate spatiotemporal relationships of human KCs. KCs were marked manually on each channel, and local generation was confirmed with decreased gamma power. In most cases (76%), KCs occurred in a single location, and rarely (1%) in all locations. However, if automatically detected KC-like phenomena were included, only 15% occurred in a single location, and 27% occurred in all recorded locations. Locally generated KCs were found in all sampled areas, including cingulate, ventral temporal, and occipital cortices. Surprisingly, KCs were smallest and occurred least frequently in anterior prefrontal channels. When KCs occur on two channels, their peak order is consistent in only 13% of cases, usually from prefrontal to lateral temporal. Overall, the anterior-posterior separation of electrode pairs explained only 2% of the variance in their latencies. KCs in stages 2 and 3 had similar characteristics. These results open a novel view where KCs overall are universal cortical phenomena, but each KC may variably involve small or large cortical regions and spread in variable directions, allowing flexible and heterogeneous contributions to sleep homeostasis and memory consolidation.

No MeSH data available.


Related in: MedlinePlus