Disable / Decrease Sensitivity of Controller Rotation

[bastianilso]

In stroke rehabilitation patients typically have a healthy side ( typically right) and a paralyzed "neglected side" (typically left). For the purpose of stroke rehabilitation, patients will need to practice moving their healthy arm into their "neglected side." Currently, Whack-A-Mole VR can be played without moving your arm at all - you can simply rotate the controller with your hand.

To encourage patients to use their whole arm we should decrease the sensitivity of the controller or completely disable the rotational aspect of the controller - instead patients should point on the moles by moving their arm.

[QuentinDaveau]

@bastianilso I think this would need careful design to work. Reducing the Pointer sensivity would feel unintuitive, and disabling the pointer rotation would mean greatly reducing the wall size so the participant can hit every Mole (this may be in conflict with the idea that the whole Participant's field of view is covered by the wall). On the practical aspect I don't think this would really work, I can quickly implement it so we can test them, but I think we may want to try to find a better solution.

Could you give a bit more context on what the end goal is so we can to discuss the implementation more easily ?

[hendrikknoche]

yes, indeed this is a tricky one. But for upper limb rehabilitation and specifically the problem that some patients have a reduced motor space (in which they can move the controller) we need to be able to:

1. establish their motor control region (in 3D space).
2. change the control-display gain accordingly such that we don't have to reduce the size of the display field which we need for diagnostics of e.g. neglect. the CD gain should be such that they can easily reach all targets. Either in a pointing task including and excluding rotation (relying only on x,y translation movements and in some cases z axis, i.e. away from the body)
3. establish what are the smallest targets they can reliably and with ordinary effort acquire this could both apply to rotation and translation of the controller (but we should be able to control that independently). In case anyone is interested here's a student paper excerpt that looked into manipulating the CD gain (to increase of motion for people with neck pain) and specifically details previous work on the topic of fiddling with CD gain (in this case on head rotation (not the controllers) but we could assume that results would be similar).

Control-Display (C-D) Gain in 3D, a concept by Poupyrev et al.[10] is the ratio between a person’s head rotation and the virtual rotation on the display. Unity (1.0 or 0%) Gain is when the virtual movement is in sync with the head movement. Positive Gain means the virtual movement is increased compared to the head movement, while Negative gain means the virtual rotation is decreased. For a 10% or 1.1 gain, a 30◦ head rotation would mean a 33◦ virtual rotation. A -10% or 0.9 gain on the same 30◦ head rotation would mean a 27◦ virtual rotation. Since the mapping is relative, one point in 3D space will always be the same as the one in real life regardless of C-D gain, this point is when the user is looking in the "forward" direction set up in the program, with their heads level with the ground. Multiple experiments have been done regarding C-D gain, especially about finding the Just Noticeable Difference (JND). JND is the C-D gain where a participant cannot tell the difference between the virtual, and real movements. Chen et al. [2] and Harvie et al. [6] both investigated JND, and came to similar conclusions. The primary measure in both experiments was gain detection, while C-D gain was manipulated between 0.6 and 1.5. The method for getting the JND was different, as Chen et al. used Steinicke et al.’s [12] research to use a Detection Threshold (DT) at a 25%, meaning that the C-D gain where 25% of the participants notice the difference between unity and the altered C-D gain is the JND gain. Each participant had to go through C-D gains from 0.7 to 1.5 with a 0.1 step and perform 4 neck rotations at 55◦ left and right. Harvie et al. on the other hand used C-D gains from 0.6 to 1.3 with a step of 0.02, their 9 participants had to perform neck rotations, however it was not mentioned how many per C-D gain step, or the degree of rotations. Instead of using a fixed DT, they chose the JND gains as the two extreme values where participants identified the unity and altered C-D gains as identical. The positive JND gain was similar in both experiments, Chen et al.’s being 1.159 and Harvie et al.’s 1.18, however on the negative C-D gain, it was 0.903 and 0.72 respectively. A possible reason for this can be Harvie et al.’s method for identifying the JND. The two extremes were detected as identical in about 46% of the time, meaning that on these two C-D gains, 54% found it different - an arbitary DT, which is more than double than the DT used by Chen et al. Both Chen et al. [3] and Harvie et al. [6, 7] revisited their results in later experiments. Both looked into the application of C-D gain on neck pain patients, with different approaches. Harvie et al. [6] used their findings on JND as upper and lower boundaries in their experiment on 24 neck pain patients whether C-D gain has an effect on their Pain Free Range of Motion (PfROM), the angles of neck rotation when acute pain sets in. With the C-D gains 0.8, 1.0 and 1.2, they asked the participants to rotate their heads six times left and right until they feel pain. They found that at 0.8 gain the PfROM increased by 6% and at 1.2 it decreased by 7% compared to unity gain, and that the C-D gain had no effect on their average pain. Chen et al.[3] used a similar method to their previous experiment, this time with 9 neck pain patients, with a C-D gain ranging from 0.8 to 1.0 with a step of 0.05. The participants did not have to go over C-D gains of 1.0 as they would not rotate their heads over their demonstrated limits, which is in accordance to the findings of Harvey et al. [6]. The average JND Gain of the neck pain patients was at 0.95 C-D gain compared to the 0.903 of healthy people, suggesting that neck pain patients notice the difference in C-D gains before healthy people, which contradicts the findings of Harvie et al. [7], who, in a similar setup, with gains ranging from 0.6 to 1.4 was testing whether healthy or neck pain patients noticed altered visual feedback. They measured gain detection and scored it based on whether the answer was correct, and found that neck pain patients scored lower than healthy people. In their final experiment, Chen et al. [3] used the individual JND for each of their participants, and calculated a "nudged" gain, which was 0.5 less than their personal JND gain. They then asked the participants to perform the same looking exercise. The measures in the evaluation task were verbal, and patients at nudge Gain said that the neck movement was "feasible and as expected". As they only used negative C-D gains for this experiment, the results correlate again with Harvie et al.’s [6] findings. Chen et al. theorize that for sitting experiments a higher than 25% DT is possible, such as the one determined by Harvie et al. [6], and they note that the JND Gains are highly individual, and may not be applicable for other movements.

references [1] Benjamin Bolte, Marc de Lussanet, and Markus Lappe. 2016. Virtual reality system for the enhancement of mobility in patients with chronic back pain. International Journal of Child Health and Human Development 9, 3 (2016), 305. https://search. proquest.com/docview/1850780073?rfr_id=info%3Axri%2Fsid%3Aprimo [2] Karen B. Chen, Kevin Ponto, Mary E. Sesto, and Robert G. Radwin. 2014. Influence of altered visual feedback on neck movement for a virtual reality rehabilitative system. Proceedings of the Human Factors and Ergonomics Society Annual Meeting 58, 1 (Sept. 2014), 693–697. https://doi.org/10.1177/1541931214581162 [3] K. B. Chen, M. E. Sesto, K. Ponto, J. Leonard, A. Mason, G. Vanderheiden, J. Williams, and R. G. Radwin. 2017. Use of Virtual Reality Feedback for Patients with Chronic Neck Pain and Kinesiophobia. IEEE Transactions on Neural Systems and Rehabilitation Engineering 25, 8 (Aug. 2017), 1240–1248. https://doi.org/10. 1109/TNSRE.2016.2621886 [4] Seneth Gajasinghe, Mandika Wijayaratna, and Anuja Abayadeera. 2010. Correlation between numerical rating scale (NRS) and visual analogue scale (VAS) in assessment of pain in post operative patients. Sri Lankan Journal of Anaesthesiology 18, 2 (Nov. 2010), 81–83. https://doi.org/10.4038/slja.v18i2.2445 [5] Jan Hartvigsen, Mark J Hancock, Alice Kongsted, Quinette Louw, Manuela L Ferreira, StÃľphane Genevay, Damian Hoy, Jaro Karppinen, Glenn Pransky, Joachim Sieper, Rob J Smeets, Martin Underwood, Rachelle Buchbinder, Jan Hartvigsen, Dan Cherkin, Nadine E Foster, Chris G Maher, Martin Underwood, Maurits van Tulder, Johannes R Anema, Roger Chou, Stephen P Cohen, LucÃŋola Menezes Costa, Peter Croft, Manuela Ferreira, Paulo H Ferreira, Julie M Fritz, StÃľphane Genevay, Douglas P Gross, Mark J Hancock, Damian Hoy, Jaro Karppinen, Bart W Koes, Alice Kongsted, Quinette Louw, Birgitta ÃŰberg, Wilco C Peul, Glenn Pransky, Mark Schoene, Joachim Sieper, Rob J Smeets, Judith A Turner, and Anthony Woolf. 2018. What low back pain is and why we need to pay attention. The Lancet 391, 10137 (June 2018), 2356–2367. https://doi.org/10.1016/S0140-6736(18)30480-X [6] Daniel S. Harvie, Markus Broecker, Ross T. Smith, Ann Meulders, Victoria J. Madden, and G. Lorimer Moseley. 2015. Bogus Visual Feedback Alters Onset of Movement-Evoked Pain in People With Neck Pain. Psychological Science 26, 4 (April 2015), 385–392. https://doi.org/10.1177/0956797614563339 [7] Daniel S. Harvie, Susan Hillier, Victoria J. Madden, Ross T. Smith, Markus Broecker, Ann Meulders, and G. Lorimer Moseley. 2016. Neck Pain and Proprioception Revisited Using the Proprioception Incongruence Detection Test. Physical Therapy 96, 5 (May 2016), 671–678. https://doi.org/10.2522/ptj.20150210 [8] Marianne Jensen Hjermstad, Peter M. Fayers, Dagny F. Haugen, Augusto Caraceni, Geoffrey W. Hanks, Jon H. Loge, Robin Fainsinger, Nina Aass, Stein Kaasa, and European Palliative Care Research Collaborative (EPCRC). 2011. Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. Journal of Pain and Symptom Management 41, 6 (June 2011), 1073–1093. https://doi.org/10.1016/j.jpainsymman.2010.08.016 [9] Ted Jones, Todd Moore, and James Choo. 2016. The Impact of Virtual Reality on Chronic Pain. PLOS ONE 11, 12 (Dec. 2016), e0167523. https://doi.org/10.1371/ journal.pone.0167523 [10] Ivan Poupyrev, Suzanne Weghorst, Takahiro Otsuka, and Tadao Ichikawa. 1999. Amplifying Spatial Rotations in 3D Interfaces. In CHI ’99 Extended Abstracts on Human Factors in Computing Systems (CHI EA ’99). ACM, New York, NY, USA, 256–257. https://doi.org/10.1145/632716.632874 [11] K. M. Refshauge and C. G. Maher. 2006. Low back pain investigations and prognosis: a review. British Journal of Sports Medicine 40, 6 (June 2006), 494–498. https://doi.org/10.1136/bjsm.2004.016659 [12] Frank Steinicke, Gerd Bruder, Jason Jerald, Harald Frenz, and Markus Lappe.

2008. Analyses of human sensitivity to redirected walking. In Proceedings of the 2008 ACM symposium on Virtual reality software and technology - VRST ’08. ACM Press, Bordeaux, France, 149. https://doi.org/10.1145/1450579.1450611 [13] E. Taal, M. M. Veehof, and M. A. van de Laar. 2013. THU0482-HPR Comparison of agreement and responsiveness to treatment effects of 100 mm visual analog scale and 11 point numerical rating scale to assess wrist pain in patients with rheumatoid arthritis. Annals of the Rheumatic Diseases 71, Suppl 3 (June 2013), 741–741. https://doi.org/10.1136/annrheumdis-2012-eular.2447 [14] Gul Deniz Yilmaz Yelvar, Yasemin ÃĞÄśrak, Murat DalkÄślÄśnÃğ, Yasemin Parlak Demir, Zeynep Guner, and Ayŧenur Boydak. 2017. Is physiotherapy integrated virtual walking effective on pain, function, and kinesiophobia in patients with non-specific low-back pain? Randomised controlled trial. European Spine Journal 26, 2 (Feb. 2017), 538–545. https://doi.org/10.1007/s00586-016-4892-7

[bastianilso]

@QuentinDaveau I see what you mean with the unintuitiveness part, very good point. One way this could be resolved is if we add a different "explanation source" for the laser. For instance a shield-like circular object which follows the controllers location, but is unaffected by the controller's rotation. I think this way we get the best of both worlds - 1) people have 1-1 mapping between what they hold and what they see, and 2) we also get a nice explanation for where this laser is coming from, and get the correct "clinical behavior". 👨‍⚕