Decoding the infant mind: Multivariate pattern analysis (MVPA) using fNIRS

Abstract

The MRI environment restricts the types of populations and tasks that can be studied by cognitive neuroscientists (e.g., young infants, face-to-face communication). FNIRS is a neuroimaging modality that records the same physiological signal as fMRI but without the constraints of MRI, and with better spatial localization than EEG. However, research in the fNIRS community largely lacks the analytic sophistication of analogous fMRI work, restricting the application of this imaging technology. The current paper presents a method of multivariate pattern analysis for fNIRS that allows the authors to decode the infant mind (a key fNIRS population). Specifically, multivariate pattern analysis (MVPA) employs a correlation-based decoding method where a group model is constructed for all infants except one; both average patterns (i.e., infant-level) and single trial patterns (i.e., trial-level) of activation are decoded. Between subjects decoding is a particularly difficult task, because each infant has their own somewhat idiosyncratic patterns of neural activation. The fact that our method succeeds at across-subject decoding demonstrates the presence of group-level multi-channel regularities across infants. The code for implementing these analyses has been made readily available online to facilitate the quick adoption of this method to advance the methodological tools available to the fNIRS researcher.

The present and future use of functional near‐infrared spectroscopy (fNIRS) for cognitive neuroscience

“…Other data‐driven approaches detect functional activity assuming that the recorded fNIRS data are a combination of task‐evoked and task‐independent components that can be separated by means of independent component analysis (ICA), principal component analysis (PCA), or task‐related component analysis . Recently, more powerful and multivariate techniques, such as the multivoxel pattern analysis (MVPA), have been borrowed from fMRI, with the capacity to discriminate task‐evoked brain activity between two or more experimental conditions . Freeware analysis toolboxes have been developed and listed on the Society for fNIRS…”

“…57 Recently, more powerful and multivariate techniques, such as the multivoxel pattern analysis (MVPA), have been borrowed from fMRI, with the capacity to discriminate task-evoked brain activity between two or more experimental conditions. 58 Freeware analysis toolboxes have been developed and listed on the Society for fNIRS. 59 All these methods rely on knowledge of the timeline of the stimuli that is predetermined and recorded in typical computer-based experiments conducted in laboratories.…”

Fearful but not happy expressions boost face detection in human infants

“…MVPA combines multiple variables (such as voxel activations) to classify trials according to stimulus category, target location, or other parameters [42]. MVPA is used with different data types including fMRI [43], EEG [44], fNIRS [45] and behavioural [46] data. One previous study used MVPA with infant fNIRS [45], and at least one used MVPA with infant EEG data [47].…”

Temporal dynamics of visual representations in the infant brain

“…Participants. Within-subject multivariate pattern classification analyses require that each included participant contribute sufficient data to train classifiers, which is challenging to achieve with infants (Emberson et al, 2017). Thus, we restricted analyses to individual participants who contributed at least 50% valid trials compared to the maximum possible number of trials (infants: at least 80 valid trials total out of a maximum of 160, or an average of 10 valid trials per condition; adults: at least 128 valid trials out of a maximum of 256, or an average of 16 valid trials per condition; range: infants 9-19, adults 15-32).…”

“…Two recent studies have used group-level MVPA-related methods to uncover representations of a small number of stimulus classes in awake infants using near infra-red spectroscopy (fNIRS; Emberson, Zinszer, Raizada, & Aslin, 2017) and fMRI: Deen et al, 2017). These studies provided an early demonstration that information-based analysis methods can map the geometry of neural representations in infant visual cortices at the group level.…”