Naturalistic language assessment and brain-network biomarkers in post-stroke aphasia for precision neurorehabilitation
- 1Clinical Research Center, Shanghai Second Rehabilitation Hospital, Shanghai University of Traditional Chinese Medicine , ,
- 2Department of Rehabilitation Medicine, Renhe Hospital, Baoshan District , ,
- 3Department of Rehabilitation Medicine, Huashan Hospital, Fudan University , ,
Abstract
Advances in naturalistic language assessment, portable neuroimaging, electrophysiology, and brain-network modeling are creating new opportunities to evaluate post-stroke aphasia under conditions closer to real-world communication. Conventional aphasia batteries and isolated linguistic tasks remain essential for diagnosis, severity classification, and treatment planning, but they provide limited information about how language-network activity supports communication in dynamic, naturalistic contexts. This integrative review, informed by a structured literature search and descriptive thematic mapping, synthesizes evidence on connected speech, discourse production, narrative comprehension, interactive communication, and contextual reading after stroke. We examine how multilevel features from these behaviors can be organized into graded, interacting dimensions of lexical-semantic access, phonological encoding, grammatical organization, discourse processing, executive-predictive control, and speech-motor output. We further review multimodal evidence from fNIRS, fMRI, EEG/ERP, lesion-symptom mapping, structural connectivity, and functional connectivity, highlighting residual left-hemisphere language networks, conditional right-hemisphere support, bilateral reorganization, domain-general control systems, and longitudinal recovery. Portable EEG and fNIRS are considered complementary approaches for repeated assessment during less constrained language tasks, with distinct temporal, spatial, and practical trade-offs. Building on this evidence, we propose an operational ecological biomarker framework that links task choice, multilevel feature extraction, multivariate behavioral profiles, neural measures, prespecified clinical questions, and longitudinal outcomes. This framework may support recovery prediction, rehabilitation stratification, neuromodulation target selection, individualized therapy planning, and monitoring of functional communication.
Related
- Automated screening of autism via a pretrained vision-language model in naturalistic caregiver-child interactions.
- A Low-Rank Space Captures Linear Alignment Between LLM and Human Neural Activity During Naturalistic Language Listening
- Brain alignment of reasoning and action representations from vision-language and action models during naturalistic gameplay
- Gut microbiota-derived tryptophan metabolites in cardiovascular disease: mechanisms, biomarkers, and precision interventions
- Highlighting the strengths of new wearable fNIRS, precision accuracy microgenetic research designs, and new NeuroDOT processing algorithm to measure change over time in clinical populations for Speech-Language and Hearing Researchers
BibTeX
@article{xia2026naturalistic,
title = {{Naturalistic language assessment and brain-network biomarkers in post-stroke aphasia for precision neurorehabilitation}},
author = {Jiayi Xia and Song Pei and Song Zhou and Yulong Bai},
journal = {Frontiers in human neuroscience},
year = {2026},
month = oct,
doi = {10.3389/fnhum.2026.1966552},
url = {https://europepmc.org/article/PMC/PMC13717245},
note = {CC BY},
}
RIS
TY - JOUR
TI - Naturalistic language assessment and brain-network biomarkers in post-stroke aphasia for precision neurorehabilitation
AU - Jiayi Xia
AU - Song Pei
AU - Song Zhou
AU - Yulong Bai
PY - 2026
DA - 2026/10/15
JO - Frontiers in human neuroscience
DO - 10.3389/fnhum.2026.1966552
UR - https://europepmc.org/article/PMC/PMC13717245
AB - Advances in naturalistic language assessment, portable neuroimaging, electrophysiology, and brain-network modeling are creating new opportunities to evaluate post-stroke aphasia under conditions closer to real-world communication. Conventional aphasia batteries and isolated linguistic tasks remain essential for diagnosis, severity classification, and treatment planning, but they provide limited information about how language-network activity supports communication in dynamic, naturalistic contexts. This integrative review, informed by a structured literature search and descriptive thematic mapping, synthesizes evidence on connected speech, discourse production, narrative comprehension, interactive communication, and contextual reading after stroke. We examine how multilevel features from these behaviors can be organized into graded, interacting dimensions of lexical-semantic access, phonological encoding, grammatical organization, discourse processing, executive-predictive control, and speech-motor output. We further review multimodal evidence from fNIRS, fMRI, EEG/ERP, lesion-symptom mapping, structural connectivity, and functional connectivity, highlighting residual left-hemisphere language networks, conditional right-hemisphere support, bilateral reorganization, domain-general control systems, and longitudinal recovery. Portable EEG and fNIRS are considered complementary approaches for repeated assessment during less constrained language tasks, with distinct temporal, spatial, and practical trade-offs. Building on this evidence, we propose an operational ecological biomarker framework that links task choice, multilevel feature extraction, multivariate behavioral profiles, neural measures, prespecified clinical questions, and longitudinal outcomes. This framework may support recovery prediction, rehabilitation stratification, neuromodulation target selection, individualized therapy planning, and monitoring of functional communication.
SN - 1662-5161
ER -