Datasets
Standard Dataset
IEEE-TMI paper, A 2D synthesized image improves the 3D search for foveated visual systems
- Citation Author(s):
- Submitted by:
- Devi Klein
- Last updated:
- Thu, 01/26/2023 - 18:56
- DOI:
- 10.21227/f8vk-aj29
- Data Format:
- License:
101 Views- Categories:
- Keywords:
Abstract
We provide the abstract from the paper below:
Current medical imaging increasingly relies on 3D volumetric data making it difficult to thoroughly search all regions of the volume. In some applications (e.g., Digital Breast Tomosynthesis), the volumetric data is typically paired with a synthesized 2D image (2D-S) generated from the corresponding 3D volume. We investigate how this image pairing affects the search for spatially large and small signals. Observers searched for these signals in 3D volumes, 2D-S images, and while viewing both. We hypothesize that lower spatial acuity in the observers’ visual periphery hinders the search for the small signals in the 3D images. However, the inclusion of the 2D-S guides eye movements to suspicious locations, improving the observer’s ability to find the signals in 3D. Behavioral results show that the 2D-S, used as an adjunct to the volumetric data, improves the localization and detection of the small (but not large) signal compared to 3D alone. There is a concomitant reduction in search errors as well. To understand this process at a computational level, we implement a foveated search model (FSM) that executes human eye movements and then processes points in the image with varying spatial detail based on their eccentricity from fixations. The FSM predicts human performance for both signals and captures the reduction in search errors when the 2D-S supplements the 3D search. Our experimental and modeling results delineate the utility of 2D synthesized images in 3D search—limit the reliance on low-resolution peripheral processing by guiding attention to regions of interest, effectively reducing errors.
There are two seperate files in this data repository.
The first is titled, "MRMC_analysis.csv". This .csv file can be directly read into the R software program so that one can perform an MRMC analysis on the dataset using the software package, "MRMCaov." (https://brian-j-smith.github.io/MRMCaov/using.html)
The second .csv file is called, "human FSM data.csv." It contains all the data used to generate the figures in the paper titled, "A 2D synthesized image improves the 3D search for foveated visual systems." Most of the column headers are self explanatory. However, we describe the column headers that are less intuitive.
1) "foveate target" - This variable states whether the participant in the psychophyiscs experiment stared at the target (i.e., an eye tracker recorded a fixation within 2 degrees visual angle from the center of the target profile). On target-absent trials the value is set to -1. Otherwise, it is True or False.
2) "suspicious click target" - This variable variable pertains to the 2D-S + 3D condition only. It states whether the participant localized the target (i.e., clicked on it on the monitor screen) in the 2D-S image before performing the 3D search. Please see paper for me details.
3) "signal loc r", "signal loc c" and "signal loc s" refer to row, column, and z-coordinate indices of the target position in the 3D volume, respectively. This is based on a numpy array (0 index start). For target-absent trials, these values are set to -1.
4) "subject clk loc r", "subject clk loc c" and "subject clk loc s" refer to the row, column, and z-coordinate indices of the subjects click in the 3D volume, respectively. If the subject did not click anywhere, then these values are set to -1.
5) "FSM response mass" and "FSM response micro" represent the Foveated Search Model's output at the end of the trial for both of the signals the participants were instructed to search for while performing the task. See the paper for more details on the FSM implementation.
Dataset Files
- human FSM data.csv (333.50 kB)
- MRMC_analysis.csv (72.07 kB)
