Figure 5. Characterization of brain state-specific pupil–fMRI relationships.

(A) Pupil information content maps generated by integrating PCA spatial maps with weights of linear regression models trained on cluster-specific trials. In all clusters, negative weights were found in the somatosensory cortex, the cerebellum, and posterior parts of the thalamus. All clusters had positive weights in anterior thalamic, preoptic and hypothalamic nuclei, subiculum, parts of the hippocampus and in the region containing the tuberomammillary nucleus. Clusters 1–3 displayed positive weights in neuromodulatory brainstem regions, substantia nigra, and ventral tegmental area, as well as the entorhinal cortex. The cingulate cortex and retrosplenial cortex and supramammillary nucleus were positive in clusters 2–4. Marked with gray are frames plotted in (B). (B) Cluster-specific spatial patterns are portrayed on slices selected from A (marked with gray rectangles). Characteristic to cluster 1 were positive weights in the dopaminergic substantia nigra and ventral tegmental area as well as in their efferent projections in the nucleus accumbens and caudate-putamen. Positive weighting was also found in the periaqueductal gray and brainstem laterodorsal tegmental and parabrachial nuclei, as well as in the superior colliculus. Cluster 2 was characterized by the strongest positive weights in hypothalamic regions, lateral in particular. Brainstem areas containing the arousal-regulating locus coeruleus, laterodorsal tegmental, and parabrachial nuclei, as well as the septal area and the olfactory tubercle displayed high positive weights. In cluster 3, as in cluster 2, the area containing the locus coeruleus, laterodorsal tegmental, and parabrachial nuclei showed positive linkage with pupil dynamics. The highest cluster 3 values were located in preoptic and other hypothalamic areas, as well as in stria terminalis carrying primarily afferent hypothalamic fibers, caudate-putamen, and globus pallidus. In cluster 4, the neuromodulatory region showing the strongest positive weights was the caudal raphe. The anterior parts of the brainstem displayed negative weighting. Characteristic to cluster 4 were high weights in the thalamus and in the hippocampus and the subiculum forming the hippocampal formation. Masked regions (white) did not pass the false discovery rate corrected significance threshold (p=0.01). Abbreviations: Ce – cerebellum, CgCx – cingulate cortex, CP – caudate-putamen, GP – globus pallidus, Hp – hippocampus, Hy – hypothalamus, LC – locus coeruleus, LDT – laterodorsal tegmental nuclei, LH – lateral hypothalamus, NA – nucleus accumbens, OTu – olfactory tubercle, PAG – periaqueductal gray, PB – parabrachial nuclei, PO – preoptic nuclei, Ra – raphe, Sb – subiculum, SC – superior colliculus, Se – septal area, SN – substantia nigra, ST – stria terminalis, Th – thalamus, VTA – ventral tegmental area.

Figure 5—figure supplement 1. The spatial map based on cluster 1 trials highlights regions from which pupil-related information was decoded.

Masked regions (white) did not pass the false discovery rate corrected significance threshold (p=0.01). Abbreviations: Ce – cerebellum, CP – caudate-putamen, ECx – entorhinal cortex, Hp – hippocampus, Hy – hypothalamus, LDT – laterodorsal tegmental nuclei, LH – lateral hypothalamus, NA – nucleus accumbens, PAG – periaqueductal gray, PB – parabrachial nuclei, Sb – subiculum, SC – superior colliculus, SCx – somatosensory cortex, SN – substantia nigra, Th – thalamus, TuM – tuberomammillary nucleus, VTA – ventral tegmental area.

Figure 5—figure supplement 2. The spatial map based on cluster 2 trials highlights regions from which pupil-related information was decoded.

Masked regions (white) did not pass the false discovery rate corrected significance threshold (p=0.01). Abbreviations: B9 – B9 serotonergic cells, Ce – cerebellum, CgCx – cingulate cortex, ECx – entorhinal cortex, Hp – hippocampus, Hy – hypothalamus, LC – locus coeruleus, LDT – laterodorsal tegmental nuclei, LH – lateral hypothalamus, OTu – olfactory tubercle, PB – parabrachial nuclei, PPT – pedunculopontine tegmental nuclei, Ra – raphe, RsCx – retrosplenial cortex, Sb – subiculum, SCx – somatosensory cortex, Se – septal nuclei, SN – substantia nigra, SuM – supramammillary nucleus, Th – thalamus, TuM – tuberomammillary nucleus, VTA – ventral tegmental area.

Figure 5—figure supplement 3. The spatial map based on cluster 3 trials highlights regions from which pupil-related information was decoded.

Masked regions (white) did not pass the false discovery rate corrected significance threshold (p=0.01). Abbreviations: Ce – cerebellum, CgCx – cingulate cortex, CP – caudate-putamen, ECx – entorhinal cortex, GP – globus pallidusHp – hippocampus, LC – locus coeruleus, LDT – laterodorsal tegmental nuclei, LH – lateral hypothalamus, PAG – periaqueductal gray, PB – parabrachial nuclei, PO – preoptic nuclei, Ra – raphe, RF – reticular formation, RsCx – retrosplenial cortex, Sb – subiculum, SCx – somatosensory cortex, SN – substantia nigra, ST – stria terminalis, SuM – supramammillary nucleus, Th – thalamus, TuM – tuberomammillary nucleus, VTA – ventral tegmental area.

Figure 5—figure supplement 4. The spatial map based on cluster 4 trials highlights regions from which pupil-related information was decoded.

Masked regions (white) did not pass the false discovery rate corrected significance threshold (p=0.01). Abbreviations: Ce – cerebellum, CgCx – cingulate cortex, Hp – hippocampus, LH – lateral hypothalamus, Ra – raphe, RsCx – retrosplenial cortex, Sb – subiculum, SCx – somatosensory cortex, SuM – supramammillary nucleus, Th – thalamus, TuM – tuberomammillary nucleus.