AUTHOR=Liu Menghui , Wan Chunxiao , Wang Chunyan , Li Xinyi 

TITLE=Predicting upper limb motor dysfunction after ischemic stroke: a functional near-infrared spectroscopy-based nomogram model

JOURNAL=Frontiers in Neurology

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2025.1524851

DOI=10.3389/fneur.2025.1524851

ISSN=1664-2295

ABSTRACT=BackgroundThis study aimed to identify independent risk factors associated with upper limb motor functional recovery in ischemic stroke patients 3 months post-stroke and to construct a predictive model based on functional near-infrared spectroscopy (fNIRS) data.MethodsThe study included 114 patients with ischemic stroke, divided into a training group (n = 80) and a validation group (n = 34). Variables related to the FMA-UE score 3 months later were selected from fNIRS data using LASSO regression, and independent risk factors were determined through logistic regression analysis. A nomogram was constructed based on these factors to predict the probability of upper limb motor dysfunction scores after stroke, and the model’s discriminative ability was assessed using the area under the ROC curve (AUC), as well as the clinical net benefit was evaluated using decision curve analysis (DCA).ResultsThe LASSO regression ultimately selected seven variables for the assessment of motor dysfunction post-stroke, of which five were identified as independent risk factors. The five independent fNIRS risk factors associated with upper limb motor functional recovery are A_A_dxy_DLPFC_to_Temporal: The number of brain functional connectivity edges from the affected side dorsolateral prefrontal cortex (DLPFC) to the affected side temporal lobe under deoxygenated hemoglobin monitoring level, A_UA_oxy_DLPFC_to_PSMC: The number of brain functional connectivity edges from the affected side DLPFC to the unaffected side primary somatosensory motor cortex (PSMC) under oxyhemoglobin monitoring level, A_UA_total_Temporal_to_DLPFC: The number of brain functional connectivity edges from the affected side temporal lobe to the unaffected side DLPFC under total hemoglobin monitoring level, UA_UA_dxy_Temporal_to_Frontopolar: The number of brain functional connectivity edges from the unaffected side temporal lobe to the unaffected side frontopolar cortex under deoxygenated hemoglobin monitoring level, and UA_UA_total_PSMC_to_PMC: The number of brain functional connectivity edges from the unaffected side PSMC to the unaffected side premotor cortex (PMC) under total hemoglobin monitoring level. The AUC of the ROC curve for the nomogram was 0.971 in the training dataset and 0.804 in the testing dataset, demonstrating good discriminative ability. DCA results indicated that the model showed good clinical net benefit in both the validation and development cohorts.ConclusionThis pilot study successfully constructed a predictive model based on fNIRS data to forecast the risk factors for upper limb motor functional recovery 3 months after ischemic stroke, providing a valuable tool for clinical decision-making and treatment planning.