Biology subjects

Caisachana Guevara, L. M.

Publications and source records attributed to Caisachana Guevara, L. M..

3 recordsLinked to original sources

bioRxiv2026

Temporo-Occipital and Medial Temporal Networks Underlying Object-Location Learning

Object-location memory (OLM), a fundamental component of spatial memory, is essential for everyday functioning yet declines with aging and neurodegenerative disorders. Functional neuroimaging studies have consistently implicated medial temporal lobe (MTL) structures, including the hippocampus and parahippocampal gyrus, as well as medial and lateral temporo-occipital regions, in object-location associative learning. However, the neural mechanisms and functional network interactions supporting OLM acquisition remain largely unknown. To address this gap, we examined twenty healthy adults (18-45 years) performing an object-location learning task during functional MRI. As a first main finding, task-related functional connectivity analyses revealed enhanced coupling between MTL structures, ventral visual regions, and temporo-occipital cortices during OLM learning. Secondly, stronger connectivity within this network was associated with higher learning accuracy, highlighting the behavioral relevance of coordinated cortical-MTL interactions. Thirdly, these connectivity patterns were not static but evolved across learning stages: interregional coupling was strongest during early learning and progressively attenuated as performance stabilized, suggesting sharpening and increased efficiency of network interactions with learning. Together, our results provide mechanistic insight into OLM acquisition at the functional network level and offer an evidence-based framework for identifying target networks to enhance spatial memory through non-invasive brain stimulation.

neuroscience↗

bioRxiv2025

Deep learning to overcome human error and bias in electrode position extraction in tDCS-fMRI studies

Combining transcranial direct current stimulation (tDCS) with fMRI enables investigation of stimulation effects, while structural MRI verifies electrode positioning, critical for focal montages where minor misplacements reduce target current dose. Currently, no fully automated methods exist for precise electrode extraction from MRI. We developed an Attention U-Net-based deep learning approach to automate electrode detection in focal tDCS-fMRI studies, using structural pointwise-encoding time reduction with radial acquisition (PETRA) MRI scans from a multicenter trial (N = 392 images; 1,568 electrodes, https://www.memoslap.de/en/home/). Performance was compared to manual and semi-automated methods for a 3x1 montage (three cathodes around a central anode). The network achieved robust segmentation (Dice Score = 0.76, Hausdorff distance = 36.76 mm), identifying all electrodes in 95% of cases (323/340). These metrics compared network-segmented electrodes to highly accurate, manually segmented electrodes, which are also called "ground truth". Linear mixed-effects models on 52 "ground truth" images showed deep learning outperformed manual and semi-automated methods, aligning best with ground truth. Fully/semi-automated methods comparison of 290 images showed highest agreement (ICC = 0.990, bias = 1.3 mm), while manual extraction exhibited larger biases (-5.77 to -7.06 mm) and systematic errors. The automated approach overcomes manual limitations by improving precision, eliminating human variability and bias, and enabling scalability for large studies. It sets a new standard for electrode verification in focal tDCS, particularly for studies requiring precise localization. The model is open-source, with future refinements discussed

neuroscience↗

bioRxiv2025

Functionally Relevant and Reliable Brain Stimulation Targets for Enhancement of Novel Word-Learning

Linking word-forms and their meanings is central to language learning. Transcranial direct current stimulation (tDCS), has shown potential to enhance this process, but with variable effects. This study aimed to (1) identify reliable and functionally relevant tDCS target brain regions to enhance novel-word learning and (2) assess test-retest reliability (TRR) of behavioral and imaging outcomes. Twenty healthy individuals completed two functional magnetic resonance imaging (fMRI) sessions using parallel task versions. Participants learned picture-pseudoword associations across six learning blocks. Behavioral learning was analyzed using linear-mixed-models. Whole-brain and region-of-interest (ROI) analyses examined learning-related activity changes and their behavioral relevance. TRR was assessed using intraclass correlation coefficients (ICCs). Participants successfully acquired the novel-word forms, indexed by increased accuracy and faster latency across stages. Behavioral outcomes showed good-to-excellent TRR. The task elicited robust language-learning related activity and activity changes across stages were correlated with learning success. Task-related activity was variable, but voxels within significant clusters ([~]81%) and most ROIs showed moderate-to-excellent consistency. Power analyses confirmed a sufficient sample size for detecting the reported ICCs. Current modeling suggested that focal-tDCS can induce neurophysiologically relevant electrical field strength in the identified target regions. Hence, we identified accessible, reliable and functionally relevant cortical targets for enhancing novel-word learning. TRR results support the usefulness of the paradigm for future concurrent tDCS-fMRI research. Our study also outlines a general path towards optimization of brain stimulation studies by implementing an empirically informed approach for selecting reliable and relevant target regions and implementation of reliable experimental and imaging paradigms.

neuroscience↗