Eye-tracking metrics as a new tool to assess patient's adherence to robot-assisted gait training

The growing use of robotic rehabilitation in clinical practice has changed the dyadic interaction between physiotherapist and patient, introducing novel elements such as performance-based feedback (FB). The latter has a strong impact on patients’ motivation, participation and error-driver learning, which are crucial aspects to consider in order to optimize the rehabilitation session to achieve successful outcomes. In addition, the presence of FB has a relevant contribution in promoting the active participation of the patient during the execution of movement. Several studies have demonstrated, through the administration of psychological questionnaires, that patients appreciate the FB presence during the rehabilitation sessions, but none of the studies have quantified the patient’s engagement expressed in terms of visual attention. We approached this issue in a recent work, by measuring Eye-Tracking (ET) data in individuals with incomplete spinal cord injury during treadmill-based robotic assisted gait training, focusing on different types of FB and different levels of patient-therapist interaction. ET metrics, indeed, are quantitative measures that track eye behavior and provide valuable information related to cognitive and emotional processes. These play a key role in active gaze control affecting the observer’s attention allocation in complex scenes. Here, we aim to investigate the evolution of visual attention in 8 individuals with incomplete spinal cord injury along a Lokomat-based rehabilitation protocol of four weeks, in which we encouraged a high patient-therapist interaction. The ET data were collected during two rehabilitation sessions, one at the beginning (session 1) and one at the end (session 12) of the entire rehabilitation program. The results highlight how the patient’s eye-gazed focused more and longer on the screen showing the FB in the last session compared to the first, as evidenced by the ET metrics considered (i.e. number of fixations, total duration of fixations and average duration of fixations), underlining an improvement in adherence to the rehabilitation training in the majority of the patients. Clinical Relevance — This work demonstrates how the eyetracker can be valuable tool for assessing patient behaviour during the rehabilitation session and consequently understanding which aspects of the rehabilitation protocol can be optimized to achieve better clinical outcomes.