Kyriaki Mikellidou

Neuroscience is advancing standardization and tool development to support rigor and transparency. Consequently, data pipeline complexity has increased, hindering FAIR (findable, accessible, interoperable and reusable) access. brainlife.io was developed to democratize neuroimaging research. The platform provides data standardization, management, visualization and processing and automatically tracks the provenance history of thousands of data objects. Here, brainlife.io is described and evaluated for validity, reliability, reproducibility, replicability and scientific utility using four data modalities and 3,200 participants.

Our brain is skilled with the ability to perceive and process multimodal stimuli. This process known as crossmodal perceptual integration, has been in the research spotlight for a long time, providing evidence for the integration of information coming from different modalities. Prior research mostly utilized pictures and focused on the semantic content of a single sound or word. The present study aims to investigate crossmodal perceptual integration in realistic conditions using short movieclips (1500ms) and auditory meaningful three-word sentences to evaluate target detection in terms of accuracy and response times. Two experimental tasks were developed using PsychoPy, where participants had to indicate whether a target (noun for Experiment 1, verb for Experiment 2) was present or absent. In trials without a target, target-related information was always present, either through one of the two senses (vision or audition; incongruent condition) or through both senses (congruent condition). We observed superior performance when the target was absent generally (M _exp1 =93.9%, SD _exp1 =0.04, M _exp2 =84.2%, SD _exp2 =0.182) compared to when it was present (M _exp1 =82.3%, SD _exp1 =0.143, M _exp2 =73.7%, SD _exp2 =0.184). Moreover, superior performance was noted in incongruent target-related movieclips, which significantly decreased during congruent target-related movieclips. In Experiment 1, we observed that in the audio condition when the target-related word was a noun, participant performance was superior compared to when it was a verb (M=99.4% vs. M=86.7%; t _{incVerb vs. incNoun} =-8.428, p=.001). In Experiment 2, the judgment scores were similarly high in incongruent movieclips and significantly lower in congruent ones regardless of whether the target-related information presented was a verb or noun. The present results provide evidence regarding the role of complexity of semantics, and especially the diverse role verbs and nouns could play in crossmodal perceptual integration in more realistic situations. Our findings can enrich the content of learning techniques, as well as the design of AI models, by taking advantage of the supporting role of semantic audiovisual information, while taking into consideration the potential confusion that the complexity of semantic information can induce to perception experience.

We investigate the notion that basic visual information is acting as a building block for more complex cognitive processes in humans. Specifically, we measured individual visual orientation discrimination thresholds to report significant correlations against the total standardised intelligence quotient (IQ), verbal-IQ and non-verbal IQ scores evaluated using the Wechsler Abbreviated Scale of Intelligence Second Edition (WASI-II) test battery comprising Verbal Reasoning, Block Design, Similarities and Matrix Reasoning subtests (N = 92). A multiple linear regression analysis showed that participants’ performance in our visual discrimination task, could be explained only by individual scores in Verbal Reasoning which quantifies the ability to comprehend and describe words and Matrix Reasoning, which evaluates general visual processing skills including abstract and spatial perception. Our results demonstrate that low-level visual abilities and high-level cognitive processes are more tightly interwoven together than previously thought and this result could pave the way for further research on how cognition can be defined by basic sensory processes.

Serial dependence effects have been observed using a variety of stimuli and tasks, revealing that the recent past can bias current percepts, leading to increased similarity between two. The aim of this study is to determine whether this temporal integration occurs in egocentric or allocentric coordinates. We asked participants to perform an orientation reproduction task using grating stimuli while the head was kept at a fixed position, or after a 40° yaw rotation between trials, from left (−20°) to right (+20°), putting the egocentric and allocentric cues in conflict. Under these conditions, allocentric cues prevailed.

We investigate the notion that basic visual information is acting as a building block for more complex cognitive processes in humans. Specifically, we measured individual visual orientation discrimination thresholds to report significant correlations against the total standardised intelligence quotient (IQ), verbal-IQ and non-verbal IQ scores evaluated using the Wechsler Abbreviated Scale of Intelligence Second Edition (WASI-II) test battery comprising Verbal Reasoning, Block Design, Similarities and Matrix Reasoning subtests (N=92). A multiple linear regression analysis showed that participants’ performance in our visual discrimination task, could be explained only by individual scores in Verbal Reasoning which quantifies the ability to comprehend and describe words and Matrix Reasoning, which evaluates general visual processing skills including abstract and spatial perception. Our results demonstrate that low-level visual abilities and high-level cognitive processes are more tightly interwoven together than previously thought and this result could pave the way for further research on how cognition can be defined by basic sensory processes.

Serial dependence effects have been observed using a variety of stimuli and tasks, revealing that the recent past can bias current percepts, leading to increased similarity between two. The aim of this study is to determine whether this temporal integration occurs in egocentric or allocentric coordinates. We ask participants to perform an orientation reproduction task using grating stimuli while the head is kept at a fixed position throughout the whole session or while alternating position from one trial to the next, from left (−20°) to right (+20°), putting the egocentric and allocentric cues in conflict. Under these conditions, allocentric cues prevail.

The human visual system is capable of processing visual information from fovea to the far peripheral visual field. Recent fMRI studies have shown a full and detailed retinotopic map in area prostriata, located ventro-dorsally and anterior to the calcarine sulcus along the parieto-occipital sulcus with strong preference for peripheral and wide-field stimulation. Here, we report the anatomical pattern of white matter connections between area prostriata and the thalamus encompassing the lateral geniculate nucleus (LGN). To this end, we developed and utilized an automated pipeline comprising a series of Apps that run openly on the cloud computing platform brainlife.io to analyse 139 subjects of the Human Connectome Project (HCP). We observe a continuous and extended bundle of white matter fibers from which two subcomponents can be extracted: one passing ventrally parallel to the optic radiations (OR) and another passing dorsally circumventing the lateral ventricle. Interestingly, the loop travelling dorsally connects the thalamus with the central visual field representation of prostriata located anteriorly, while the other loop travelling more ventrally connects the LGN with the more peripheral visual field representation located posteriorly. We then analyse an additional cohort of 10 HCP subjects using a manual plane extraction method outside brainlife.io to study the relationship between the two extracted white matter subcomponents and eccentricity, myelin and cortical thickness gradients within prostriata. Our results are consistent with a retinotopic segregation recently demonstrated in the OR, connecting the LGN and V1 in humans and reveal for the first time a retinotopic segregation regarding the trajectory of a fiber bundle between the thalamus and an associative visual area.

The human visual system is capable of processing visual information from fovea to the far peripheral visual field. Recent fMRI studies have shown a full and detailed retinotopic map in area prostriata, located ventro-dorsally and anterior to the calcarine sulcus along the parietooccipital sulcus with strong preference for peripheral and wide-field stimulation. Here, we report the anatomical pattern of white-matter connections between area prostriata and the thalamus encompassing the lateral geniculate nucleus (LGN). We observe a continuous and extended bundle of white matter fibers from which two subcomponents can be extracted: one passing ventrally parallel to the optic radiations (OR) and another passing dorsally circumventing the lateral ventricle. Interestingly, the loop travelling dorsally connects the thalamus with the central visual field representation of prostriata, while the other loop travelling more ventrally connects the LGN with the more peripheral visual field representation. This is consistent with a retinotopic segregation recently demonstrated in the OR, connecting the LGN and V1 in humans. Our results demonstrate for the first time a retinotopic segregation regarding the trajectory of a fiber bundle between the thalamus and an associative visual area.