Inter-individual and inter-site neural code conversion without shared stimuli

TL;DR

The paper tackles the challenge of functional alignment across individuals without requiring shared stimuli, introducing a content loss-based neural code converter that uses hierarchical DNN features to map source brain activity into a target brain space. This converter enables accurate inter-individual and inter-site decoding and high-quality image reconstructions, with performance comparable to traditional brain-loss–driven methods and robust to non-overlapping stimuli and different decoders. The approach demonstrates data-efficient inter-subject decoding, scalable cross-dataset inter-site reconstruction, and generalizable representations that transfer to alternative DNN readouts, suggesting broad applicability beyond visual reconstruction. These results offer a scalable framework for reusing public neuroimaging data and open avenues for brain-to-brain communication and cross-modality applications while reducing data-collection costs. The work advances functional alignment by leveraging content-based latent representations rather than paired brain responses, enabling flexible, cross-subject analyses across datasets and sites.

Abstract

Inter-individual variability in fine-grained functional brain organization poses challenges for scalable data analysis and modeling. Functional alignment techniques can help mitigate these individual differences but typically require paired brain data with the same stimuli between individuals, which is often unavailable. We present a neural code conversion method that overcomes this constraint by optimizing conversion parameters based on the discrepancy between the stimulus contents represented by original and converted brain activity patterns. This approach, combined with hierarchical features of deep neural networks (DNNs) as latent content representations, achieves conversion accuracy comparable to methods using shared stimuli. The converted brain activity from a source subject can be accurately decoded using the target's pre-trained decoders, producing high-quality visual image reconstructions that rival within-individual decoding, even with data across different sites and limited training samples. Our approach offers a promising framework for scalable neural data analysis and modeling and a foundation for brain-to-brain communication.