Brain-Computer Interface (BCI)-Based Feedback for Chronic Pain Management

Feasibility of a Brain-Computer Interface (BCI)-Based Adaptive Sensing and Feedback for Chronic Pain Management

The feasibility study proposed here will primarily examine the sensitivity of an electroencephalogram (EEG)-based Brain Computer Interface (BCI) in detecting significant differences in brain signals in patients with chronic low back pain (N=10), lower limb pain (N=10) and healthy controls (N=10) through perceived movements via a video and during actual movements. The BCI device has been approved for use in previous trials (e.g. NNI-IRB/07/001, DSRB Domain D/09/608, DSRB Domain D/10/072) and the safety and effectiveness of this non-invasive EEG-based BCI device validated through these trials. However, the validation has not been specific to its use in pain. Related to the primary objective of the study, we will develop and validate an adaptive and participant-specific pain detection and analysis program by exploring and identifying discriminative and robust patterns in spontaneous EEG from our study sample. For the secondary objective, we will develop and validate a BCI and computer based pain and attention diversion training system with interactive audio-visual feedbacks for Phase 2 of the study. These feedbacks will inform the user about the current brain activation level and attention level, and guide the user in learning to modulate the EEG characteristics and develop skills to manage attention to alleviate perceived fear-related pains. The BCI system captures EEG signals and decodes the underlying brain states in relation to cognition and fear-related pain perception. Such decoded brain states are then presented to the participant in visual or other form to guide the participant to learn to regulate the brain states towards better pain management. For example, the participant may over a few sessions learn to focus on the visual feedback while inhibiting the brain function activity in relation to fear-related pain perception. With practice, the user is encouraged to achieve brain activity modulation without external feedback so fear-related pain can be reduced in realistic situations.

We hypothesize that, using appropriately designed system learning and computing techniques, a brain-computer interface (BCI) technology can detect specific brain wave patterns in patients with chronic low back pain or lower limb pain compared to healthy controls. Also, that BCI technology can detect specific brain wave patterns for fear avoidance behaviours in motor activities known to trigger pain in such chronic pain patients. This is a validation and feasibility study. We calculated sufficient funding to cover 10 participants each with low back pain, lower limb pain, healthy controls and BCI training respectively. Participants will be screened and undergo a comprehensive assessment by the Rehabilitation Physician (RP) on the study team. The aim of the comprehensive assessment is to establish the safety and suitability of participants to be study subjects according to inclusion and exclusion criteria.To ensure a high level of compliance given the high frequency of visits of the subject during the study, it is important that motivated volunteers are chosen. Consent will be taken in an outpatient consultation room at the pain management clinic(PMC) at Tan Tock Seng Hospital, with full provision for privacy. PMC is where the study will be conducted. Informed consent will be taken at the convenience of the potential participant. A full explanation of the participant information sheet will be given and any questions answered. The study is conducted over 3 phases. Total span of study participation: 11-months. Phase 1: A total of 2 visits to the study site over 2 consecutive weeks is required. Each visit will take approximately 1 hour 15 minutes. The 1st session involves a visit to the doctor's office for a screening assessment-history taking, physical examination, and electronic notes review no different from the current standard pain consultation. Those found to be suitable and decide to take part, will be asked at the 2nd session to: Complete a digital set of self-report measures and a 6-minute walking test. These measures are completed at the start and end of the second session. The walking test will be conducted by a trained physiotherapist familiar with this measure. Participants then participate in two sets of non-invasive experiments one after another. The BCI system set-up comprising of a laptop connected to a brain signal recorder and a headset to record brain signals will be used. Sensors to measure sweat and pulse rate will also be used in the experiments. The brain signal recorder together with the sensors will be secured in a backpack and worn by participants during the experiments. Exp 1-participants watch 15 videos of people doing daily activities. Each video frame is approximately 1 min long. EEG signal patterns are recorded for each video frame. Exp 2-participants go through a series of 15 physical movements as guided by the physiotherapist. Participants are required to complete 15 repetitions each, within 1 min. EEG signal patterns are recorded for each movement. Two buzzers (labelled "lesser pain" and "more pain") are available for participants to press accordingly should they experience 'lesser' or 'more' pain at any point during the experiment. If 'more' pain is experienced, the physiotherapist will assess the participant's condition and the participant will have a choice to either (1) continue the experiment at a lower intensity or (2) stop the experiment. Phase 2: Approximately 6 months after participants' participation in Phase 1 and span a period of 7 weeks. If selected to complete Phase 2, participants will undergo 18 sessions (3 sessions/week) of brain training over a period of 6 consecutive weeks. Sessions are conducted on alternate days in the week with each session lasting approximately 1 hour. The same system set-up used in Phase 1 is used in each training session. At the start of the first session in week 1, and immediately after completing the last training session in Week 6 participants will also complete the same set of self-report questionnaires and the 6-minute walking test (as assessed in Phase 1). At week 7 (1 week post training) participants will complete the same full set of measures as set out in the second session of Phase 1. Phase 3: A total of 2 visits to the study site is required. Each visit will take approximately 1 hour 15 minutes. These 2 visits will be scheduled approximately 8.5 mths and 10.5 mths after completing Phase 1. Both visit sessions involve a repeat of the full set of measures that participants completed at the 2nd session in Phase 1. Potential Risks The potential risks to participants are minimal as none of the measures and components of the system are invasive. There are currently no known safety issues with the headset and sensors. However participants will be kept informed of any new findings. Participants may experience temporary exertion related discomfort in the main muscle groups that are used in the physiotherapy guided movement experiment (Experiment 2) although the were designed to fit within participants' abilities. Participants will be advised to contact the Principal Investigator (PI) should the muscular discomfort persist for more than 24 hours. If participants are screened positive for depression at any point on the study, the PI who is a trained psychologist will screen them for active depression and where appropriate refer to the RP for further treatment recommendations. In this case, the participant will be excluded or withdrawn from the study. Due to the potential of the videos and the guided movements to elicit fear of pain, participants may experience some increase in anxiety and slight worsening of usual pains. Experiment 2 is performed under close supervision of the physiotherapist to prevent any possible strain caused by incorrect movements. If participants experience 'more' pain, the physiotherapist will assess their condition and participants will have a choice to either (1) continue the experiment at a lower intensity or (2) stop the experiment. In the event that participants experience anxiety from participating on any of the experiments, the PI will assess their symptoms of anxiety and be able to administer appropriate intervention onsite as necessary at no cost. If further treatment is required, participants stop the experiment, get re-assessed by the RP and appropriate referrals will be made for treatment follow-up. As no other medication except simple analgesics such as Paracetamol, Paracetamol/Orphenadrine, Non-steroidal anti-inflammatory drugs or Tramadol to treat pain can be taken while participating on the study, there is a possibility pain may be exacerbated and their condition may worsen. If this occurs, the RP will re-assess their condition and make appropriate treatment recommendations. If participants pain condition is exacerbated at any point during the course of the study for reasons not indicated in points 1-4 above, they will be referred to the RP on the study team for an assessment who will then make appropriate treatment recommendations. if participants develop significant symptoms because of study participation which warrants medical intervention, a subsidized medical consult and follow-up treatment will be at their own costs. The RP on the study team will facilitate a fast tracked appointment at PMC within 2 weeks. Study Withdrawal: (a) Participants may withdraw voluntarily from participation in the study at any time. Participants may also withdraw voluntarily from receiving the study intervention for any reason. If voluntary withdrawal occurs, the participant will not be asked to continue scheduled evaluations and their involvement in the study will cease. (b) Any participant will be withdrawn from further participation on the study if a serious adverse event classified as having a reasonable causal relationship to the study occurs. In this case the participant will be given appropriate medical care until the symptoms resolve or the condition becomes stable. The participant will not be asked to continue scheduled evaluations and their involvement in the study will cease. (c) All participants will be withdrawn if there is study closure due to Domain Specific Review Board (DSRB) review. Participants who drop out of the study will not be replaced. (d) Any participant can be withdrawn from further participation on the study if the study team deem it is the best interest of the participant. This can also happen when the participant does not follow instructions required to complete the study adequately, or develops other medical problems during the study period. Specific Restrictions / Requirements: (a) Standard over the counter pregnancy testing kits will be used for the purposes of determining if a female of childbearing age is pregnant. (b) All forms of contraception are allowed during the entire duration of the study (c) Participants are limited to only taking simple analgesics such as Paracetamol, Paracetamol/Orphenadrine, Non-steroidal anti-inflammatory drugs (NSAIDs) or Tramadol while participating in the study (d) All medications (prescription and over the counter), vitamin and mineral supplements, and / or herbs taken by the participant will be documented. Data analysis: Self-report outcome measures and data from the walking test will be analysed via the Statistical Package for the Social Sciences (SPSS) package utilizing t-tests to measure group differences and pearson bivariate correlations to measure potential relationships between outcome measures and responses made on Exp 1 (video condition) and Exp 2 (physiotherapy guided movement condition). EEG patterns will be analysed via feature analysis in association with clinical signals and user states/condition including physiological measures of skin conductance and pulse rate. Data Storage: 1. All hardcopy data, will be stored in designated locked cabinet(s) that are accessible to authorized study personnel only. 2. All electronic data, they will be stored on in a secured computer that is password-protected. The databases will not contain participant identifiers and the data linking participant identifiers and the participant identification codes will be stored separately including those using portable media (e.g. CD, USB drives etc.). Data Access: 1. All study team members will have access to the research data which has password protection. 2. Data will not be released and shared with individuals or entities outside the institution. 3. Data will be destroyed after it has been stored for 6 years or minimum duration of retention period as specific by our institutional policy, whichever that is longer. Data Monitoring: 1. Safety data will be monitored at every visit with participants asked to report adverse events since the last visit. 2.There are no anticipated adverse events or complications expected. However reported adverse events if any, will be fully recorded, graded in terms of severity, and classified as either not related to the study, or having a reasonable causal relationship to the study. 3. The PI will be responsible for reporting all adverse events to DSRB. If excess risk to the participant/s is observed, or a statistically significant benefit is observed, the study is stopped early and the patient informed of the results. Futility is not used as a stopping criteria as this is a feasibility trial design of which there is limited data, and of a small sample size. The final study results are needed to proof the study concept here. 4. The PI will be responsible for the dissemination of any data and safety information to the study site. Communication will be conducted weekly via secure e-mail to the study team during active phase of the study in Phase 1, 2 and 3. A monthly face-to-face meetings and where necessary ad hoc meetings during the period of the study.

All participants complete two experiments one after another-EEG patterns, skin conductance and pulse rate are recorded. The EEG cap is mounted on the participant's head, the GSR sensor secured on the fingers and the PPG sensor secured on the wrist using a Velcro strap. Exp 1-participants watch a series of 15, 1 min videos of people doing daily activities. Exp 2-participants do 15 movements (15 reps each in 1 min) as guided by a physiotherapist. 2 identical buzzers (labelled "lesser pain" and "more pain") are available to press accordingly should participants experience 'lesser' or 'more' pain any time during the experiment. If 'more' pain experienced, participants' condition will be assessed and they can choose to continue the experiment at a lower intensity or stop.

Scalp EEG signals that are acquired and recorded from experiment 1 and 2 are analysed for healthy controls, participants with low back pain and participants with lower limb pain. Comparisons of signal patterns are made between groups. Assuming unique signal patterns are obtained for each groups, these signals are further analysed and used to develop a feedback brain training system to better manage pain and ideally reduce fear-avoidant behaviors towards pain.

Measure of patients' perceptions about the consequences of harmful situations based on a series of photographs

Assessing change in Pain Interference Scale on the Brief Pain Inventory (BPI) (Cleeland & Ryan, 1994)

Inclusion Criteria: Only English speaking participants are recruited. Participants who are healthy controls need to meet the following inclusion criteria: Not suffering from any active pain condition Without any known disease Not on any long- term medication or health supplements AND Able to complete a digital self-report questionnaire in English without need for translation. Participants who have either low back pain or lower limb pain need to meet the following inclusion criteria: Diagnosed to have musculoskeletal low back pain OR lower limb pain lasting for > 3 months. Not currently on any active therapy (Example: physiotherapy, stabilisation exercise therapy) or medication EXCEPT simple analgesics (Paracetamol, Paracetamol/Orphenadrine, Non-steriodal anti-inflammatory drugs (NSAIDs), Tramadol). Able to complete a digital self-report questionnaire without need for translation. Exclusion Criteria: Participants are excluded from the study if they meet any of the following criteria: Pregnant or lactating. History of drug or alcohol abuse History of significant moderate to severe systemic disease including a confirmed diagnosis or current investigation for neoplasm, immunological condition involving musculo-skeletal system; active liver, heart and or renal failure; respiratory diseases needing supplementary oxygen or Bilevel Positive Airway Pressure (BiPAP) machine and any other conditions that make it improbable to engage in physical movements. History of psychiatric condition- in particular a psychiatric diagnosis meeting Diagnostic and Statistical Manual of Mental Disorders-V (DSM-V) criteria for major depression, anxiety, phobias or any other conditions that make it improbable to engage in physical movements. History of back surgery and surgery involving the joints and bones of the lower limbs, such as knee and hip replacement surgery. Spine or lower limb pain that has not been medically stabilised. History of previous injuries to the head and/or brain, including but not limited to epilepsy, skull defects or cranioplasty Ongoing compensation claims or litigation.

Jensen MP, Sherlin LH, Fregni F, Gianas A, Howe JD, Hakimian S. Baseline brain activity predicts response to neuromodulatory pain treatment. Pain Med. 2014 Dec;15(12):2055-63. doi: 10.1111/pme.12546. Epub 2014 Oct 7.

Jensen MP, Hakimian S, Sherlin LH, Fregni F. New insights into neuromodulatory approaches for the treatment of pain. J Pain. 2008 Mar;9(3):193-9. doi: 10.1016/j.jpain.2007.11.003. Epub 2007 Dec 21.

Pinheiro ES, de Queiros FC, Montoya P, Santos CL, do Nascimento MA, Ito CH, Silva M, Nunes Santos DB, Benevides S, Miranda JG, Sa KN, Baptista AF. Electroencephalographic Patterns in Chronic Pain: A Systematic Review of the Literature. PLoS One. 2016 Feb 25;11(2):e0149085. doi: 10.1371/journal.pone.0149085. eCollection 2016.