Computerised Working Memory Training in Acquired Brain Injury

Participants will be recruited through Northampton Healthcare NHS Foundation Trust Acquired Brain Injury clinics. They will complete 5 weeks of training at home. During the first two weeks they will be visited in their home by a researcher who will administered the current stimulation together with the working memory training program. The remaining 3 weeks the patient will complete the training program at home, while receiving a motivational / catch up call every week. Weekends will be exempt from testing. The working memory training software program is accessed via a password-protected website on a secure server at Dalhousie University. Training data will be saved on the Dalhousie University server and will be downloaded in Birmingham for data analyses.

RECRUITMENT As the aim of this research is to improve spatial working memory, only participants with a working memory impairment will be included in the study. To identify a working memory impairment, cognitive tests from the Wechsler Adult Intelligence Test (WAIS-IV) and the Wechsler Memory Scale (WMS-IV), will be administered in paper and pencil versions or using iPads. These tests are administered to every patient, as part of the routine care protocol.

The investigators will consider the following scores:

• A Full-Scale Intelligent Quotient (FSIQ) obtained from the WAIS-IV higher than 70: this score identifies patients able to follow instructions. The FSIQ will also be used as a covariate in the analysis to control for low average overall ability rather than a specific working memory issue.
• A Working Memory Index (WMI), obtained as a combination of the digit span score and the arithmetic score in the WAIS-IV, smaller than 85: this score identifies a general working memory impairment;
• A Visual Working Memory Index (VWMI), obtained as a combination of the symbol span and the spatial addition score in the WMS-IV, smaller than 85: this score identifies impairment specific to spatial working memory.

These tests are co-normed on a large sample (mean = 100; SD = 15). In order to be included in the study, a patient should have a FSIQ >70 and either a WMI or a SWMI or both < 85. This means that the investigators use a criterion score of one standard deviation below the mean. This score doesn't reflect significant impairment, but it allows us to identify patients with an impairment significant enough to potentially improve with training. See below for a detailed description of the cognitive tasks used. Participants who don't present a working memory impairment, as defined above, will be excluded from further participation. Patients identified as potential participants in the study will be approached with a letter.

SCREENING PROCEDURE:

Following the informed consent procedure, each participant will be randomly assigned to one of the two training conditions (active or sham stimulation) and interviewed with a Screening Questionnaire, which includes questions about demographic information and health history that will evaluate the inclusion/exclusion criteria. The investigators will also ask about current medications and information about time since injury, handedness and colour blindness will be recorded. At this point, participants who don't meet the eligibility criteria will be excluded from further participation. Finally, participants will be made aware of possible side effects of brain stimulation and they will ask to confirm their willingness to participate in the study.

TESTING AND TRAINING:

The protocol is detailed in the protocol flowchart (Annexe 5). After the consent and screening procedures, eligible participants will begin the administration of the baseline measures (T0), and outcome measures (T1). The baseline measures involve completing a series of questionnaires (see below for a list and a brief explanation of each questionnaire), since these variables may impact performance on the tasks. The baseline data will be included in our data analyses as potential modifiers of performance. In total, the investigators expect the entirety of the first in-lab session (informed consent, screening, and baseline measures (T0)) to take about 1 hour. In addition, participants will be asked to refrain from excessive alcohol or coffee drinking during the intervention and to maintain good sleeping habits, where possible. On the following day participants will perform a series of pre-training outcome tasks (see above) to measure their working memory capacity baseline (T1) and to assess, during and at the end of the training intervention, the efficacy of the training and the stimulation regime. Participants will then be offered a break before familiarizing them with the stimulation procedure (sham or active according to the group they belong to) and the training game. The trainer will answer any questions about the study. This second session will last for about 1 h. On T1, the cognitive screening measures (WAIS-IV and WMS-IV) will not be repeated, as already administered as part of the routine care.

On the first training day after the T1 session, depending on the group, participants will receive active or sham brain stimulation (see below for stimulation parameters). At the same time, both groups will complete one session of the NIGMA game, e.g., one session of their training routine (20 minutes). Before each training session, participants will also be asked to answer short questions (level of alertness, engagement, etc.). Once the training session and the stimulation are completed, the participant will fill in a feedback form on the experienced side effects of brain stimulation, if any.

Participants will complete 10 additional consecutive training sessions (2 weeks, excluding weekends). Each session should take about 45 minutes (~10 minutes setting up of tDCS and ~20 minutes of N-IGMA training). This phase will be completed at home, with the researcher visiting the participant at a convenient time.

When the first training phase is complete, the participant will undergo time 2 (T2) assessment. Like before, this involves completing a series of computerised cognitive tasks to measures training gains and transfer (see "outcome measures" section, including also WAIS-IV and WMS-IV).

Participants will then start the second training phase, involving 3 weeks of training only (no brain stimulation). During this phase, patients will stay at home and access the training program via internet. Manualised phone calls will be used to educate the patients about attention and how they could apply the training to their daily life, to promote generalisation. When the second training phase is completed, participants will undergo time 3 (T3) assessment, during which they will repeat the same series of computerised cognitive tasks as in T2.

A last follow-up assessment, identical to the one at T2, and T3, will be carried out at T4, a month after the completion of the intervention, to assess maintenance of working memory improvements and transfer.

BRAIN STIMULATION PROTOCOL The brain stimulation targets the right dorsolateral prefrontal cortex (DLPFC). A bipolar setup will be used. The bipolar setup includes two Ag/AgCl electrodes filled with conductive EEG gel and placed on F4 (active electrode) and Fp1(return electrode). Reference electrodes will be attached to the earlobe and impedances will be measured throughout the stimulation. If impedances exceed 20kOhm at any time, the stimulation will automatically stop and will not resume until impedances are restored. The investigators will use a total current intensity of 2mA for 20 minutes, preceded by 30 seconds ramping up and followed by 30 seconds ramping down (total stimulation time = 21s). With these parameters and Ag/AgCl electrodes (area 3.14 cm2) a current density of approximately 0.6 mA/cm^2 is obtained, slightly higher than the one obtained with larger electrodes, but still well below the threshold for tissue damage (Antal et al., 2017; Bikson et al., 2016; Liebetanz et al., 2009)). During sham stimulation the investigators will use the same setup as in the active condition but after ramping up, the current will be brought back to zero and the process repeated 30 seconds before the end of the 21 minutes time interval (total sham stimulation time = 21s).

DATA ANALYSIS The investigators will analyse the data using parametric statistics if appropriate. These will include mixed and repeated measures ANOVAs, with factors such as group and test variables. Significant main effects and interactions (p < .05) will be followed by post-hoc tests corrected for multiple comparisons. Dependent variables on the computerized tasks will include mean reaction time (RT) and accuracy (% correct).