Model For The Prediction Of Motion Sickness Incidence, Peripheral Hcii Presentation

Editor's Notes

• #3: Motion sickness is a general term that describes a number of symptoms related to discomfort and associated emesis (vomiting) induced by real or apparent motion. Symptoms of motion sickness may be: Discomfort Headache Pallor Unwillingness to continue working Vomiting. A problem with nausea I that there exists a large inter-subject variability on the degree that a person may feel sick when sensing a provocative motion. Some persons will be severely sick, others will feel sick but not at the extent of stopping their task. Some vomit once and then they feel OK, others continue vomiting until the provocative motion ceases. Finally, almost 5% never feel motion sick. Unfortunately, motion sickness effects are evident in numerous provocative motion environments, such as ships, aircraft, automobiles, and air-cushioned vehicles. What is actually interesting about motion sickness is that the term is a misnomer; “Sickness” implies that it is a type of disease, when in fact it is a perfectly normal response of a healthy individual without any functional disorders. This response is relayed to the actual or implied motion that includes seasickness, airsickness, space sickness, and simulator sickness.
• #4: The most widely accepted theory is the conflict mismatch theory, sensory rearrangement theory [2] or neural mismatch theory. According to these theories, the cause of motion sickness is that the vestibular system provides information about self motion that does not match the sensations of motion generated by visual or kinaesthetic (proprioceptive) systems, or what is expected from previous experience [3]. So, in more general terms : Motion sickness is the outcome of the comparison between what we sense and the neural store (sensory input memory) The neural mismatch hypothesis is comprised of two basic components [4]: A neural storage unit that retains the informational characteristics of the previous sensory input, A comparator unit that matches the contents of the store with the informational characteristics of the prevailing sensory influx from the motion and position senses. It is obvious that the aforementioned comparison is a continuous and dynamic physiological process. Therefore, we al Vestibular system: Collective term for the three semicircular canals and the two vestibular sacs (utricle and saccule) within the labyrinth of the inner ear. The vestibular system is involved in the perception of spatial orientation. Proprioception: The perception of information about the position, orientation and movement of the body and its parts. (Involves the somatosensory system ) .
• #5: The consequences are tremendous as far as human performance is involved.
• #6: A convenient index of motion sickness severity is the Motion Sickness Incidence (MSI), which is the percentage of people who vomit when exposed to nauseogenic environment. Positive aspects of MSI metric Easily and objectively identifiable. Measures of the number and severity of symptoms during the progression of the syndrome are varialble and idiosyncratic, whereas emesis is an observable, behavioural marker. Negative aspects Does not take into account the numerous symptoms of motion sickness It is not related, in a straight-forward manner, to human performance
• #7: In a number of Human Factors Research (HFR) experiments (O’Hanlon & McCauley, 1974; McCauley, Royal, Wylie, O’Hanlon & Mackie, 1976) a regression model was proposed for MSI estimation. That model although useful, has two drawbacks: Is not etiologic, and It refers only to vertical oscillation (only real motion, not vection) It is assumed that the subject is seated inside a closed, lighted compartment (simulator), without being able to receive visual or auditory information, from the outside environment. Furthermore, the subject is assumed to be passive to externally induced motion.
• #8: The proposed model is conceptually based on the ideas from already existing theories and research The whole concept is based on Reason’s (1978) neural mismatch theory, The subjective discomfort model by Oman (1982), Merfeld et al. (1993) model, TNO motion sickness model by Bos et al. (2002), The system is structured in combination with observer theory concepts from control systems. MSI estimation is based on two sources of error: the intra-vestibular error derived from the estimation of gravity, and the vestibulo-ocular error derived from the residual optical flow Model input parameters True motion characteristics (Vertical acceleration frequency and amplitude) detected by the vestibular system Apparent motion characteristics (Vertical acceleration frequency and amplitude) detected by peripheral vision
• #13: The proposed model predicts MSI for two-hour exposure with less than ±5% difference, compared to the HFR descriptive model (McCauley, et al., 1976) and the corresponding experimental data, in the frequency range between 0.082 Hz and 0.30 Hz. It is interesting to observe that: The frequency region of low error, compared to HFR data, is significantly increased. The proposed model predicted error at the higher region of the HFR experimental frequencies is significantly decreased (from 30% to 15%)
• #14: The model’s predicted MSI in vection settings (lack of vestibular stimulus) is significantly increased at lower frequency regions when compared to the MSI observed in HFR experiments (vestibular stimulus combined with stationary visual surroundings). This prediction is consistent with experimental results (Parker, Duh & Furness, 2001; Duh, Parker, Phillips & Furness, 2004). Unfortunately, the model cannot be validated for apparent motion settings due to lack of adequate experimental data.
• #15: The modeled independent parameters of MSI prediction are implemented parametrically, thus the model can be easily extended to include multiple nausiogenic combinations of environmental conditions The model’s predicted MSI as presented in this work, is derived without adjustment to the experimental data. We chose acceptable values for the parameters leading to a simplified and stable model (e.g. +1 or -1). Thus, the predicted precision may be easily increased. The timeline of MSI and the adaptation process is adequately approximated and the corresponding results are following the ones given by the HFR experiments. The error region (+ or – 5%) is small compared to other models on this topic. It is etiologic. Takes into account main human physiology subsystems which are known to contribute to motion sickness incidence. Therefore, it The HFR model is a regression approximation It’s linear and time invariant, thus it is easily analyzed. Ofcourse, human physiology processes are non-linear, but in this work the linearity assumption has led to acceptable precision.
• #16: The usefulness of the proposed model can be seen only if we describe what the current state is and what the future state will be CURRENT STATE Motion is detected through the vestibular system and the peripheral vision. Seated subject No voluntary motions (the subject is passive to the externally induced motion) MSI is derived from the two errors: The error produced from the comparison between the input data from the vestibular system (which senses true motion induced at the head of the subject) and the peripheral vision system (which senses visual world motion; this motion is cognitively translated to self motion feeling) and motion already in the neural store (which is the average motion of a walking subject) The retinal slip (the error produced when tracking a visual target) Therefore, the combination of environmental settings refer to a ship’s CIC (combat information center) or other compartment without any visual contact with the external world. How can we be sure that motion sickness developed in other environmental settings will be of the same severity and characteristics (frequency and amplitude of induced motion)? Furthermore, what about settings where true motion is non-existent or negligible, and apparent motion is significant (vection – the self motion feeling developed mainly by the visual input). Such combinations exist in virtual environments and simulators. The same question can be asked about situations of combined true and apparent motion (being in a ship’s bridge in daylight). FUTURE STATE True and apparent motion detection. Apparent motion detection will be based on data from the vestibular system and from peripheral vision Proprioception Refinement of Neural Store model (more motion memories) Parametric input of implemented systems attributes (for example field of view) Therefore, the presented model is the first step in creating a truly useful tool for the system designer to evaluate the consequences of the implemented ideas, as far as motion sickness is concerned.
• #17: Include motion in 6 degrees of freedom Implementation of all physiological systems contributing to motion sickness development For example, proprioception CNS non-linear characteristics time delays Non-linear detection of motion amplitude implemented at the sensor
• #20: Matrix depicting the model state in detail. Current State Future State Inputs True motion True motion Visually detected motion Human systems involved Vestibular Vestibular Central Vision Central Vision Peripheral Vision Proprioception Neural Store One average motion Multiple motion characteristics Cue errors contributing Gravity vector estimation Gravity vector estimation to MSI Retinal Slip Retinal Slip Difference between true motion and vection
• #21: The last slide depicts where our efforts are focused on. The YELLOW color in the “FUTURE STATE” shows the latest development of the proposed model. Unfortunately, it wasn’t possible to include it in this presentation. The significant change is that we have implemented visually detected motion through Peripheral Vision, and the results are quite promising.