A VR-based assessment system for testing human perceptual sensitivity is proposed in this paper. The main goal of the system is to detect mild traumatic brain injuries. It is based on measuring human perceptual sensitivity to dynamic erroneous motion in an open virtual environment. Dynamic erroneous motion, in a computer-simulated environment, can be defined as a situation in which the physics of a given entity do not follow the physical laws of its environment. More precisely, an erroneous dynamic motion is a non-conforming behavior in a virtual reality scene. For example, a non-conforming behavior can be a flag that is waving in the opposite direction of the wind.

The assessment tool proposed by the authors is based on an outdoor scene with three test parameters: (1) flag movement direction, (2) flag movement intensity, and (3) eccentricity. The details of the parameters and how they are measured can be found in section 3 of the paper.

It is very important to note that the virtual scenes proposed to users are physically based, mainly due to the first and second parameters where the flag has to interact with the wind (that is, simulation of cloth).

Besides being physically based, the authors add control to the physical entities. In this manner, they can wave the flag in an incorrect direction. Adding control implies applying forces, and sometimes this can be very challenging.

In general, the paper is well organized and clearly explains the details of the testing parameters and how they apply forces to control the inherent physics. However, there are different aspects that are not described in the paper and are important to know in order to understand the degree of user immersion.

The authors only mention the use of the game engine Unreal development kit (UDK), and nothing is said about the type of computer, or virtual reality cave, or the type virtual reality glasses used (if any). In this sense, the paper only includes a PC screenshot.

There is a big issue here: a cave or glasses that can isolate the user are necessary to create deep immersion. Besides, current VR systems usually have real-time problems with simulations including soft bodies, cloth, fluids, and so on, due to the high level of mathematical complexity. Since the simulation seems to be executed on a PC, I wonder if a VR cave or glasses (Oculus Rift for example) could lead to this kind of real-time issue or if the low immersion of the system can lead to false results. Therefore, it seems that this is not a full VR experience, but a computer-simulated environment for perception assessment.

Nevertheless, the results presented in section 4 are very promising. The authors have shown, based on their experiments, that perceptual sensitivity of physically conforming or erroneous motion clearly can be an assessment tool for the diagnosis of mild traumatic brain injuries. I hope that future work can translate the system to real VR systems or more isolated environments.