Jayakumar, J. | Search bioRxiv

bioRxiv2018

Diversity of feature selectivity in macaque visual cortex arising from limited number of broadly-tuned input channels

Spikes (action potential) responses of most primary visual cortical cells in the macaque are sharply tuned for the orientation of a line or an edge and neurons preferring similar orientations are clustered together in cortical columns. The preferred stimulus orientation of these columns span the full range of orientations, as observed in recordings of spikes, which represent the outputs of cortical neurons. However, when we imaged also the thalamic input to these cells that occur on a larger spatial scale, we found that the orientation domain map of the primary visual cortex did not show the diversity of orientations exhibited by signals representing outputs of the cells. This map was dominated by just the one orientation that is most commonly represented in subcortical responses. This supports cortical feature selectivity and columnar architecture being built upon feed-forward signals transmitted from the thalamus in a very limited number of broadly-tuned input channels.

neuroscience↗

bioRxiv2018

High precision automated detection of labelednuclei in terabyte-scale whole-brain volumetricimage data of mouse

There is a need in modern neuroscience for accurate and automated image processing techniques for analyzing the large volume of neuroanatomical data. For e.g., the use of light microscopy to image whole mouse brains in a mesoscopic scale produces individual neuroanatomical data volumes in the TerraByte range.. A fundamental task involves the detection and quantification of objects of a given type, e.g. neuronal nuclei or somata, in whole mouse brains. Traditionally this quantification is performed by human visual inspection with high accuracy, that is not scalable.. When state-of-the-art CNN and SVM-based methods are used to solve this classification problem, they achieve accuracy levels between 85 - 92%. However, higher rates of precision and recall, close to that of humans are necessary. In this paper, we describe an unsupervised, iterative algorithm, which provides a high close to human performance for a specific problem of broad interest, i.e. detection of Green Fluorescent Protein labeled nuclei in whole mouse brains. The algorithm judiciously combines classical computer vision (CV) techniques and is focused on the complex problem of decomposing strong overlapped objects (nuclei). Our proposed iterative method uses features detected on ridge lines over distance transformation and an arc based iterative spatial-filling method to solve the problem. We demonstrate our results on two whole mouse brain data sets of Gigabyte resolution and compare it with manual annotation of the brains. Our results show that an aptly designed CV algorithm with classical feature extractors when tailored to this problem of interest achieves near-ideal human-like performance. Quantitative analysis, when compared with the manually annotated ground truth, reveals that our approach performs better on whole mouse brain scans than general purpose machine learning (including deep CNN) methods.

neuroscience↗