Investigating the Reliability of Population Receptive Field Size Estimates Using fMRI

Lage‐Castellanos, Agustín; Valente, Giancarlo; Senden, Mario; Martino, Federico De

doi:10.3389/fnins.2020.00825

Cited by 25 publications

(37 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…outside the region of stimulation). This is in line with the observation that linear encoding methods (such as ridge regression) fail to reliably estimate large receptive fields (Lage-Castellanos et al, 2020); or rather the receptive fields that partially lie beyond the field of view.…”

Section: Resultssupporting

confidence: 85%

Extremely Fast pRF Mapping for Real-Time Applications

Bhat

Lührs

Goebel

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Population receptive field (pRF) mapping is a popular tool in computational neuroimaging that allows for the investigation of receptive field properties, their topography and interrelations in health and disease. Furthermore, the possibility to invert population receptive fields provides a decoding model for constructing stimuli from observed cortical activation patterns. This has been suggested to pave the road towards pRF-based brain-computer interface (BCI) communication systems, which would be able to directly decode internally visualized letters from topographically organized brain activity. A major stumbling block for such an application is, however, that the pRF mapping procedure is computationally heavy and time consuming. To address this, we propose a novel and fast pRF mapping procedure that is suitable for real-time applications. The method is build upon hashed-Gaussian encoding of the stimulus, which significantly reduces computational resources. After the stimulus is encoded, mapping can be performed using either ridge regression for fast offline analyses or gradient descent for real-time applications. We validate our model-agnostic approach in silico, as well as on empirical fMRI data obtained from 3T and 7T MRI scanners. Our approach is capable of estimating receptive fields and their parameters for millions of voxels in mere seconds. This method thus facilitates real-time applications of population receptive field mapping.

show abstract

Section: Resultssupporting

confidence: 85%

Extremely Fast pRF Mapping for Real-Time Applications

Bhat

Lührs

Goebel

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…This value reduces asymptotically towards the aspect ratio of 1 for large pRF sizes (> 5 deg). Moreover, pRF size estimates are less robust than pRF center estimates and the absolute value depends strongly on the individual HRFs (Lage-Castellanos et al, 2020; Lerma-Usabiaga et al, 2020).…”

Section: Discussionmentioning

confidence: 99%

Population receptive field shapes in early visual cortex are nearly circular

Lerma-Usabiaga

Winawer

Wandell

2020

Preprint

View full text Add to dashboard Cite

The visual field region where a stimulus evokes a neural response is called the receptive field (RF). Analytical tools combined with functional MRI can estimate the receptive field of the population of neurons within a voxel. These population RF (pRF) methods accurately specify the central position of the pRF and provide some information about the spatial extent (size) of the receptive field. A number of investigators developed methods to further estimate the shape of the pRF, for example whether the shape is more circular or elliptical. This journal published a report that there are many pRFs with highly elliptical pRFs in early visual cortex (V1-V3; Silson et al., 2018). Such a large aspect ratio is difficult to reconcile with the spatial scale of orientation columns or visual field maps in early visual cortex. We started to replicate the experiments using their methods and found that the software used in the publication does not accurately estimate RF shape: it produces elliptical fits to circular ground-truth data. We analyzed an independent data set with a different software package that was validated over a specific range of measurement conditions. Our analysis of pRF elliptical fits show that in early visual cortex the aspect ratios are less than 2 and potentially consistent with nearly circular shapes. Furthermore, current empirical and theoretical methods do not have enough precision to discriminate ellipses with aspect ratios of 1.5 from circles (aspect ratio of 1). Through simulation we identify methods for improving sensitivity that may estimate ellipses with smaller aspect ratios. The results we present are quantitatively consistent with assessments using several other methodologies (Greene et al., 2014; Zeidman et al., 2018; Merkel et al., 2020).

show abstract

“…Second, they are likely to be insensitive to differences in voxel size. Third, pRF model predictions, given typical stimulation sequences, are sharply tuned to polar angle, enabling precise estimates of this parameter (Lage-Castellanos et al, 2020;Lerma-Usabiaga et al, 2020) The small but systematic differences in eccentricity and pRF size indicate that a measure of tolerance is required when generalizing from one dataset to another. For example, the anatomically defined retinotopic template from was derived using HCP data.…”

Section: Implications Of the Similarities And Differences Between Thementioning

confidence: 99%

Cross-dataset reproducibility of human retinotopic maps

Himmelberg

Kurzawski

Benson

et al. 2021

Preprint

View full text Add to dashboard Cite

Population receptive field (pRF) models fit to fMRI data are used to non-invasively measure retinotopic maps in human visual cortex, and these maps are a fundamental component of visual neuroscience experiments. We examined the reproducibility of retinotopic maps across two datasets: a newly acquired retinotopy dataset from New York University (NYU) (n=44) and a public dataset from the Human Connectome Project (HCP) (n=181). Our goal was to assess the degree to which pRF properties are similar across datasets, despite substantial differences in their experimental protocols. The two datasets differ in stimulus design, participant pool, fMRI protocol, MRI field strength, and preprocessing pipelines. We assessed cross-dataset reproducibility of the two datasets in terms of the similarity of vertex-wise pRF estimates and in terms of large-scale cortical magnification properties. Within V1, V2, V3, and hV4, the group-median NYU and HCP vertex-wise polar angle estimates were nearly identical. Both eccentricity and pRF size estimates were also strongly correlated between the two datasets, but with a slope different from 1; the eccentricity and pRF size estimates were systematically greater in the NYU data. Next, to compare large-scale map properties, we quantified two polar angle asymmetries in V1 cortical magnification previously identified in the HCP data. The prior work reported more cortical surface area representing the horizontal than vertical visual field meridian, and the lower than upper vertical visual field meridian. We confirm both of these results in the NYU dataset. Together, our findings show that the retinotopic properties of V1-hV4 can be reliably measured between two datasets, despite numerous differences in their experimental design. fMRI-derived retinotopic maps are reproducible because they rely on an explicit computational model that is grounded in physiological evidence of how visual receptive fields are organized, allowing one to quantitatively characterize the BOLD signal in terms of stimulus properties (i.e., location and size). The new NYU Retinotopy Dataset will serve as a useful benchmark for testing hypotheses about the organization of visual areas and for comparison to the HCP Retinotopy Dataset.

show abstract

Investigating the Reliability of Population Receptive Field Size Estimates Using fMRI

Cited by 25 publications

References 31 publications

Extremely Fast pRF Mapping for Real-Time Applications

Extremely Fast pRF Mapping for Real-Time Applications

Population receptive field shapes in early visual cortex are nearly circular

Cross-dataset reproducibility of human retinotopic maps

Contact Info

Product

Resources

About