The perception of a meaningful facial pattern on a nebulous stimulus—face pareidolia—is a typical human experience. Neuroimaging and electrophysiological studies have generally shown similarities in the spatio-temporal responses to typical faces and objects eliciting face pareidolia, that is, facelike objects. However, the extent to which facelike objects engage the same neural basis as human faces remains unclear. To address this issue, we used direct measures of brain activity from intracerebral electrodes implanted in the ventral occipito-temporal cortex (VOTC) of a large group of patients (n = 44). Face selectivity was determined by contrasting a large set of naturalistic face or facelike object images with non-face object categories. High signal-to-noise ratio face-selective and facelike object-selective responses were objectively identified and quantified with frequency tagging and compared in space and time throughout the VOTC. Selective activity to facelike objects was found in all key regions of the human cortical face network, extending to the previously unexplored anterior temporal lobe (ATL). Although category-selective activity was markedly reduced for facelike objects compared with human faces, consistent with previous findings, 89% of facelike object-selective contacts spatially overlapped with human face-selective contacts, while the remaining spatially scattered contacts recorded negligible responses. Furthermore, the amplitude of the two face-selective neural signals showed high correlations across regions, recording contacts and time courses as well as concurrent early onset, challenging the view that facelike objects are interpreted as faces through feedback from higher order brain regions. Together, our findings demonstrate that the pareidolic perception of face in facelike objects engages the same ventro-temporal neural circuitry, with the same temporal dynamics, as human faces.

Intracranial EEG (iEEG) is increasingly used in many fields of human cognitive neuroscience since it offers a unique opportunity to directly record brain activity from awake humans at a high spatial and temporal resolution. However, little is known about the influence of the reference montage on the spatial and temporal characteristics of iEEG activity. Here, we compare the spatial and temporal profiles of neural activity for five reference montages (scalp reference, common average, zero reference, local Bipolar, and Laplacian) applied to a large dataset of depth electrodes (StereoElectroEncephaloGraphy, SEEG) recordings across the human ventral occipito-temporal cortex (VOTC, N individual brains = 77). Frequency-tagging is used for objective identification and quantification of both low- (<30 Hz) and high-frequency (40–160 Hz) face-selective neural activity. For low-frequency responses, similar spatial distributions and time-courses of significant face-selective contacts and of face-selective amplitudes are found across the five reference montages, although the latter two local reference montages enhance face selectivity along the fusiform gyrus until the anterior temporal lobe. However, they also reduce the right hemisphere dominance, a hallmark of face-selective neural activity, and increase the number of significant contacts in the white matter. For high-frequency responses, similar spatial distributions and time-courses of significant face-selective contacts and of face-selective amplitudes are found for all references, except for the scalp reference (SCA), which enhances face selectivity in lateral and medial regions of the anterior VOTC. However, SCA also increases the number of significant contacts in the white matter. Thus, specificities of each electrode montage should be considered before choosing an iEEG reference, according to the research question, the anatomical region, the type of analyses, and the responses frequency range.

Fast periodic visual stimulation (FPVS) allows the objective measurement of brain responses of human word discrimination (i.e., reproducible word-category-selective responses) with a high signal-to-noise ratio. This approach has been successfully employed over the last decade in a number of scalp electroencephalography (EEG) studies. Three important advances for research on word-selective brain responses were achieved in the present study: (1) we extend previous evidence of robust word-category-selective responses to the English language, (2) report results for combined EEG and MEG signals, and (3) source estimation results. English words were presented periodically (2 Hz) among different types of letter strings (10 Hz; consonant strings, non-words, pseudo-words) while recording simultaneous EEG and MEG in 25 participants who performed a simple non-linguistic colour detection task. Data were analysed in sensor and in source space. With only 4 minutes of stimulation, we observed a robust word discrimination response in each condition, even when words were embedded in sequences of word-like pseudo-words. This response was larger in non-words and largest in consonant strings. We observed left-lateralised responses in all conditions in the majority of our participants. Cluster-based permutation tests revealed that these responses were left-lateralised in sensor as well as in source space, with peaks in left posterior regions. Our results demonstrate that the FPVS approach can elicit robust English word discrimination responses in EEG and MEG within only a few minutes of recording time. Together with source estimation, this can provide novel insights into the neural basis of visual word recognition in healthy and clinical populations.