arXiv ScienceSearch

arXiv subjects

Yujing Shen

Publications and source records attributed to Yujing Shen.

2 recordsLinked to original sources

Surf_2_Volume: a workflow for converting CIFTI parcellations to NIfTI volume space

Many neuroimaging programs require volumetric NIfTI files and cannot directly use parcellations stored in CIFTI format. We developed Surf_2_Volume, a workflow that uses Connectome Workbench, FreeSurfer, AFNI/SUMA, neuromaps, and Python to convert categorical CIFTI parcellations into NIfTI volumes. Cortical labels are transferred through fsaverage to a surface representation of the MNI152 template and assigned to voxels within a smoothed cortical ribbon mask. A threshold controls how much of this mask is included. We tested twelve thresholds with the Schaefer2018 atlas with 100 parcels and 7 networks, which is available as both an fsLR representation and a published FSL MNI152 1 mm volume. We compared each converted volume with the published reference and with standard Workbench ribbon and nearest-vertex mappings. At threshold 0.05, Surf_2_Volume had higher scores on the two measures of parcel overlap, Dice and Jaccard, than the best Workbench setting (0.729 vs. 0.709 and 0.580 vs. 0.556, respectively). It also left fewer reference voxels unlabeled (11.88% vs. 20.57%) and assigned fewer labels outside the reference (9.43% vs. 11.22%). Among the tested Surf_2_Volume thresholds, 0.05 also produced the lowest average parcel volume distortion. Surf_2_Volume provides a tunable and reproducible workflow for generating volumetric atlases from CIFTI parcellations when volume-based analyses are required.

q-bio.QM

NavigationNet: A Large-scale Interactive Indoor Navigation Dataset

Indoor navigation aims at performing navigation within buildings. In scenes like home and factory, most intelligent mobile devices require an functionality of routing to guide itself precisely through indoor scenes to complete various tasks in order to serve human. In most scenarios, we expected an intelligent device capable of navigating itself in unseen environment. Although several solutions have been proposed to deal with this issue, they usually require pre-installed beacons or a map pre-built with SLAM, which means that they are not capable of working in novel environments. To address this, we proposed NavigationNet, a computer vision dataset and benchmark to allow the utilization of deep reinforcement learning on scene-understanding-based indoor navigation. We also proposed and formalized several typical indoor routing problems that are suitable for deep reinforcement learning.

cs.CV