Biology subjects

Fridriksson, J.

Publications and source records attributed to Fridriksson, J..

12 recordsLinked to original sources

bioRxiv2026

Head-level lesion-symptom mapping of picture naming in vision-language models

Researchers in artificial intelligence increasingly intervene on language models to study how their functions are organized, silencing weights and attention components in ways reminiscent of the brain lesions long used to map language in post-stroke aphasia. Under this program of mechanistic interpretability, a common move is to ablate an attention head and read the resulting drop in a behavior as evidence that the head implements it. How much that move can establish about where a behavior is computed remains unclear, because a head whose removal disrupts a behavior is necessary for it but need not be the site where it is computed. We examined this question for the picture naming task, taking the loss of naming (anomia) that defines aphasia as the behavior of interest. Across six vision-language models, spanning three language backbones and a range of parameter scales, we ablated each attention head in turn during single-word picture naming and measured accuracy before and after. The degree of localization varied widely across models. In LLaVA-1.6-Vicuna-13B, a single early head (layer 0, head 20) was necessary: removing it alone reduced naming accuracy from 0.99 to 0.006. The same head was not sufficient, because retaining it while ablating the other 1{,}599 heads also produced 0\% accuracy. Two Mistral-backbone models (LLaVA-Mistral-7B and Idefics2-8B) had no critical head. An early-layer dependence was present in every model but varied in strength, whereas dependence on any single head ranged from dominant to absent. Within Qwen2.5-VL, a dominant head was present in the 7B model but not the 3B model, indicating that this concentration emerged with scale rather than being fixed across a model family, and it was not explained by attention type. These results show that ablating the head whose removal disrupts naming does not establish that the head computes the behavior, that the result generalizes across models, or that the behavior localizes to a head at all.

neuroscience↗

bioRxiv2026

Transformer Language Models Reveal Distinct Patterns in Aphasia Subtypes and Recovery Trajectories

Language impairments in aphasia are characterized by various representational disruptions that may be reflected in discourse production. This research examines the capacity of transformer-based language models, particularly GPT-2, to serve as a computational framework for analyzing variations in aphasic narrative speech. A longitudinal dataset of narrative speech samples collected at six time points from individuals with aphasia (N = 47) was utilized as part of an intervention study. All transcripts were processed via the GPT-2 language model to obtain activation values from each of the 12 transformer layers. Statistically significant differences in activation magnitude across aphasia subtypes were found at every layer (all p < .001), with the most pronounced effects in the deeper layers. Pairwise Tukey HSD tests revealed consistent distinctions between Brocas aphasia and both Anomic and Wernickes aphasia, suggesting a shared activation profile between the latter two. Longitudinal tests revealed significant changes over time, especially in the final three layers (10-12). These findings suggest that transformer-based activation patterns reflect meaningful variation in aphasic discourse and could complement current diagnostic tools. Overall, GPT-2 provides a scalable tool to model representational dynamics in aphasia and enhance the clinical interpretability of deep language models.

neuroscience↗

bioRxiv2025

Mapping of critical prosodic and phonetic networks in post-stroke apraxia of speech

PurposeMany have made proposals to better diagnose and/or classify post-stroke apraxia of speech (AOS), with some arguing for the separation of AOS into behavioral subtypes. Recent studies of primary progressive AOS have promoted a separation of prosodic and phonetic subtypes, aligning with a dual-motor coordination model separating the neural substrates of prosodic and phonetic function. Motivated by the limited corroboration of these subtypes in post-stroke AOS, here we present mapping results in a cohort of stroke survivors aiming to identify distinct neural substrates for prosodic and phonetic aspects of speech motor coordination. MethodsLeft-hemisphere stroke survivors (n = 127; 64 with AOS) received speech-language evaluation and neuroimaging at the Center for the Study and Treatment of Aphasia Recovery (C-STAR). AOS severity was quantified via the Apraxia of Speech Rating Scale (ASRS). We utilized a novel lesion-symptom mapping technique with an emphasis on prediction that identifies ensembles of regions supporting performance in the prosodic and phonetic domains. ResultsAn ensemble of networks supporting prosodic function localized to dorsal and ventral (but primarily dorsal) sensorimotor cortex, as well as a distributed network of white matter pathways connecting Rolandic cortex to auditory regions and cerebellum, emphasizing the role of auditory feedback processing and laryngeal control in supporting prosodic function. A separate but partially overlapping network supporting phonetic function localized primarily to ventral Rolandic cortex and the arcuate fasciculus. ConclusionsThis work represents the first mapping of prosodic and phonetic subtypes in post-stroke AOS in a large cohort of individuals. We hope our results motivate the development of assessment and treatment techniques individually targeting prosodic and phonetic functioning to better serve individuals with AOS and facilitate clinical discussion of the disorder.

neuroscience↗

bioRxiv2024

Verbal working memory and syntactic comprehension segregate into the dorsal and ventral streams

Syntactic processing and verbal working memory are both essential components to sentence comprehension. Nonetheless, the separability of these systems in the brain remains unclear. To address this issue, we performed causal-inference analyses based on lesion and connectome network mapping using MRI and behavioral testing in 103 individuals with chronic post-stroke aphasia. We employed a rhyme judgment task with heavy working memory load without articulatory confounds, controlling for the overall ability to match auditory words to pictures and to perform a metalinguistic rhyme judgment, isolating the effect of working memory load. We assessed noncanonical sentence comprehension, isolating syntactic processing by incorporating residual rhyme judgment performance as a covariate for working memory load. Voxel-based lesion analyses and structural connectome-based lesion symptom mapping controlling for total lesion volume were performed, with permutation testing to correct for multiple comparisons (4,000 permutations). We observed that effects of working memory load localized to dorsal stream damage: posterior temporal-parietal lesions and frontal-parietal white matter disconnections. These effects were differentiated from syntactic comprehension deficits, which were primarily associated with ventral stream damage: lesions to temporal lobe and temporal-parietal white matter disconnections, particularly when incorporating the residual measure of working memory load as a covariate. Our results support the conclusion that working memory and syntactic processing are associated with distinct brain networks, largely loading onto dorsal and ventral streams, respectively.

neuroscience↗

bioRxiv2023

How can graph theory inform the dual-stream model of speech processing? a resting-state fMRI study of post-stroke aphasia

The dual-stream model of speech processing has been proposed to represent the cortical networks involved in speech comprehension and production. Although it is arguably the prominent neuroanatomical model of speech processing, it is not yet known if the dual-stream model represents actual intrinsic functional brain networks. Furthermore, it is unclear how disruptions after a stroke to the functional connectivity of the dual-stream models regions are related to specific types of speech production and comprehension impairments seen in aphasia. To address these questions, in the present study, we examined two independent resting-state fMRI datasets: (1) 28 neurotypical matched controls and (2) 28 chronic left-hemisphere stroke survivors with aphasia collected at another site. Structural MRI, as well as language and cognitive behavioral assessments, were collected. Using standard functional connectivity measures, we successfully identified an intrinsic resting-state network amongst the dual-stream models regions in the control group. We then used both standard functional connectivity analyses and graph theory approaches to determine how the functional connectivity of the dual-stream network differs in individuals with post-stroke aphasia, and how this connectivity may predict performance on clinical aphasia assessments. Our findings provide strong evidence that the dual-stream model is an intrinsic network as measured via resting-state MRI, and that weaker functional connectivity of the hub nodes of the dual-stream network defined by graph theory methods, but not overall average network connectivity, is weaker in the stroke group than in the control participants. Also, the functional connectivity of the hub nodes predicted specific types of impairments on clinical assessments. In particular, the relative strength of connectivity of the right hemispheres homologues of the left dorsal stream hubs to the left dorsal hubs versus right ventral stream hubs is a particularly strong predictor of post-stroke aphasia severity and symptomology.

neuroscience↗

bioRxiv2022

Improving 3D Edge Detection for Visual Inspection of MRI Coregistration and Alignment

Detecting and visualizing edges is important in several neuroimaging and medical imaging applications. For example, it is common to use edge maps to ensure the automatic alignment of low-resolution functional MRI images to match a high-resolution structural image has been successful. Specifically, software toolboxes like FSL and AFNI generate volumetric edge maps that can be particularly useful for visually assessing the alignment of datasets, overlaying the edge map of one on the other. Therefore, edge maps play a crucial role in quality assurance. Popular methods for computing edges are based on either the first derivative of the image as in FSL, or a variation of the Canny Edge detection method as implemented in AFNI. The crucial algorithmic parameter for adjustment for each of these methods relates to the image intensity. However, image intensity is relative and can be quite variable in most neuroimaging modalities. Further, the existing approaches do not necessarily generate a closed edge/surface, which can reduce the ability to determine the correspondence between a represented edge and another image. We suggest that using the second derivative (difference of Gaussian, or DoG) of the image to generate edges resolves both these issues. This method primarily operates by specifying a spatial scale of interest (which is typically known in medical imaging) rather than a contrast scale, and creates closed surfaces by definition. We describe some convenient implementation features (for both efficiency and visual quality) developed here, and we provide open source implementations of this method as both online and high performance portable code. Finally, we include this method as part of both the AFNI and FSL software packages.

neuroscience↗

bioRxiv2021

The Wernicke conundrum revisited: evidence from connectome-based lesion-symptom mapping in post-stroke aphasia

Wernickes area has been assumed since the 1800s to be the primary region supporting word and sentence comprehension. However, Mesulam et al. (2015; 2019) raised what they termed the Wernicke conundrum, noting widespread variability in the anatomical definition of this area and presenting data from primary progressive aphasia that challenged this classical assumption. To resolve the conundrum, they posited a double disconnection hypothesis: that word and sentence comprehension deficits in stroke-based aphasia result from disconnection of anterior temporal and inferior frontal regions from other parts of the brain due to white matter damage, rather than dysfunction of Wernickes area itself. To test this hypothesis, we performed lesiondeficit correlations, including connectome-based lesion-symptom mapping, in four large, partially overlapping groups of English-speaking chronic left hemisphere stroke survivors. After removing variance due to object recognition and associative semantic processing, the same middle and posterior temporal lobe regions were implicated in both word comprehension deficits and complex noncanonical sentence comprehension deficits. Connectome lesion-symptom mapping revealed similar temporal-occipital white matter disconnections for impaired word and noncanonical sentence comprehension, including the temporal pole. We found an additional significant temporal-parietal disconnection for noncanonical sentence comprehension deficits, which may indicate a role for phonological working memory in processing complex syntax, but no significant frontal disconnections. Moreover, damage to these middle-posterior temporal lobe regions was associated with both word and noncanonical sentence comprehension deficits even when accounting for variance due to the strongest anterior temporal and inferior frontal white matter disconnections, respectively. Our results largely agree with the classical notion that Wernickes area, defined here as middle superior temporal gyrus and middle-posterior superior temporal sulcus, supports both word and sentence comprehension, suggest a supporting role for temporal pole in both word and sentence comprehension, and speak against the hypothesis that comprehension deficits in Wernickes aphasia result from double disconnection.

neuroscience↗

bioRxiv2020

Neuroanatomical dissociations of syntax and semantics revealed by lesion-symptom mapping

Theories of language organization in the brain commonly posit that different regions underlie distinct linguistic mechanisms. However, such theories have been criticized on the grounds that many neuroimaging studies of language processing find similar effects across regions. Moreover, condition by region interaction effects, which provide the strongest evidence of functional differentiation between regions, have rarely been offered in support of these theories. Here we address this by using lesion-symptom mapping in three large, partially-overlapping groups of aphasia patients with left hemisphere brain damage due to stroke (N=121, N=92, N= 218). We identified multiple measure by region interaction effects, associating damage to the posterior middle temporal gyrus with syntactic comprehension deficits, damage to posterior inferior frontal gyrus with expressive agrammatism, and damage to inferior angular gyrus with semantic category word fluency deficits. Our results are inconsistent with recent hypotheses that regions of the language network are undifferentiated with respect to high-level linguistic processing.

neuroscience↗

bioRxiv2020

Mapping the human praxis network: an investigation of white matter disconnection in limb apraxia

Left hemispheric cerebral stroke can cause apraxia, a motor-cognitive disorder characterised by deficits of higher-order motor skills such as the failure to accurately produce meaningful gestures. This disorder provides unique insights into the anatomical and cognitive architecture of the human praxis system. The present study aimed to map the structural brain network that is damaged in apraxia. We assessed the ability to perform meaningful gestures with the hand in 101 patients with chronic left hemisphere stroke. Structural white matter fibre damage was directly assessed by diffusion tensor imaging and fractional anisotropy mapping. We used multivariate topographical inference on tract-based fractional anisotropy topographies to identify white matter disconnection associated with apraxia. We found relevant pathological white matter alterations in a densely connected fronto-temporo-parietal network of short and long association fibres. Hence, the findings suggest that heterogeneous topographical results in previous lesion mapping studies might not only result from differences in study design, but also from the general methodological limitations of univariate topographical mapping in uncovering the structural praxis network. A striking role of middle and superior temporal lobe disconnection, including temporo-temporal short association fibres, was found, suggesting strong involvement of the temporal lobe in the praxis network. Further, the results stressed the importance of subcortical disconnections for the emergence of apractic symptoms. Our study provides a fine-grain view into the structural connectivity of the human praxis network and suggests a potential value of disconnection measures in the clinical prediction of behavioural post-stroke outcome.

neuroscience↗

bioRxiv2020

The Neuroanatomy of Speech Processing: A Large-Scale Lesion Study

The neural basis of language has been studied for centuries, yet the networks critically involved in simply identifying or understanding a spoken word remain elusive. Several functional-anatomical models of critical neural substrates of receptive speech have been proposed, including (1) auditory-related regions in the left mid-posterior superior temporal lobe, (2) motor-related regions in the left frontal lobe (in normal and/or noisy conditions), the left anterior superior temporal lobe, or (4) bilateral mid-posterior superior temporal areas. One difficulty in comparing these models is that they often focus on different aspects of the sound-to-meaning pathway and are supported by different types of stimuli and tasks. Two auditory tasks that are typically used in separate studies--syllable discrimination and word comprehension--often yield different conclusions. We assessed syllable discrimination (words and nonwords) and word comprehension (clear speech and with a noise masker) in 158 individuals with focal brain damage: left (n=113) or right (n=19) hemisphere stroke, left (n=18) or right (n=8) anterior temporal lobectomy, and 26 neurologically-intact controls. Discrimination and comprehension tasks are doubly dissociable both behaviorally and neurologically. In support of a bilateral model, clear speech comprehension was near ceiling in 95% of left stroke cases and right temporal damage impaired syllable discrimination. Lesion-symptom mapping analyses for the syllable discrimination and noisy word comprehension tasks each implicated most of the left superior temporal gyrus (STG). Comprehension but not discrimination tasks also implicated the left pMTG, while discrimination but not comprehension tasks also implicated more dorsal sensorimotor regions in posterior perisylvian cortex.

neuroscience↗

bioRxiv2020

Agrammatism and paragrammatism: a cortical double dissociation revealed by lesion-symptom mapping

The fundamental distinction of grammatical deficits in aphasia, agrammatism and paragrammatism, was made over a century ago. However, the extent to which the agrammatism/paragrammatism distinction exists independently of differences in speech fluency has not clearly been investigated. Despite much research on agrammatism, the lesion correlates of paragrammatism are essentially unknown. Lesion-symptom mapping was used to investigate the degree to which the lesion correlates of agrammatism and paragrammatism overlap or dissociate. Four expert raters assessed videos of 53 right-handed patients with aphasia following chronic left hemisphere stroke retelling the Cinderella story. Consensus discussion determined each subjects classification with respect to grammatical deficits as Agrammatic, Paragrammatic, Both, or No Grammatical Deficit. Each subjects lesion was manually drawn on a high-resolution MRI and warped to standard space for group analyses. Lesion-symptom mapping analyses were performed in NiiStat including lesion volume as a covariate. Secondary analyses included speech rate (words per minute) as an additional covariate. Region of interest analyses identified a double dissociation between these syndromes: damage to Brocas area was significantly associated with agrammatism, p = 0.001 (but not paragrammatism, p = 0.930), while damage to the left posterior superior and middle temporal gyri was significantly associated with paragrammatism, p < 0.001 (but not agrammatism, p = 0.873). The same results obtained when regressing out the effect of speech rate, and non-overlapping lesion distributions between the syndromes were confirmed by uncorrected whole brain analyses. Our results support a fundamental distinction between agrammatism and paragrammatism.

neuroscience↗

bioRxiv2019

Neural organization of speech production: A lesion-based study of error patterns in connected speech

While numerous studies have explored single-word naming, few have evaluated the behavioral and neural correlates of more naturalistic language, like connected speech, which we produce every day. Here, in a retrospective analysis of 120 participants at least six months following left hemisphere stroke, we evaluated the distribution of word errors (paraphasias) and associated brain damage during connected speech (picture description) and object naming. While paraphasias in connected speech and naming shared underlying neural substrates, analysis of the distribution of paraphasias suggested that lexical-semantic load is likely reduced during connected speech. Using voxelwise lesion-symptom mapping (VLSM), we demonstrated that verbal (real word: semantically related and unrelated) and sound (phonemic and neologistic) paraphasias during both connected speech and naming loaded onto the left hemisphere ventral and dorsal streams of language, respectively. Furthermore, for the first time using both connected speech and naming data, we localized semantically related paraphasias to more anterior left hemisphere temporal cortex and unrelated paraphasias to more posterior left temporal and temporoparietal cortex. The connected speech results, in particular, highlight a gradient of specificity as one translates visual recognition from left temporo-occipital cortex to posterior and subsequently anterior temporal cortex. The robustness of VLSM results for sound paraphasias derived during connected speech was notable, in that analyses performed on sound paraphasias from the connected speech task, and not the naming task, demonstrated significant results following removal of lesion volume variance and related apraxia of speech variance. Therefore, connected speech may be a particularly sensitive task on which to evaluate further lexical-phonological processing in the brain. The results presented here demonstrate the related, though different, distribution of paraphasias during connected speech, confirm that paraphasias arising in connected speech and single-word naming likely share neural origins, and endorse the need for continued evaluation of the neural substrates of connected speech processes.

neuroscience↗