Optimal Treatment Strategies of AIoT Based Wearable Devices for Stroke Walking in the Rehabilitation Training of Ischemic Stroke Patients: A Real-World Cohort Study Using Propensity-Score Matched Analysis

To evaluate the efficacy of stroke walking wearable devices based on artificial intelligence Internet of Things (AIoT) technology in the rehabilitation training of patients with ischemic stroke (IS). A total of 777 patients with IS were recruited and followed up for 6 months. The participants were divided into control (671 cases) and AIoT group (106 cases) according to whether they received AIoT treatment or not. The primary outcomes were Holden walking function grading, lower limb modified Ashworth muscle tone grading, lower limb Brunnstrom grading, joint range of motion, and gait between two groups of patients within 3 days before treatment and 1 month after treatment. Propensity score matching (PSM) analysis was performed based on various factors such as gender, age, and course of illness at admission. There was no significant difference (P>0.05) in Holden walking function grading, lower limb modified Ashworth muscle tone grading, lower limb Brunnstrom grading, joint range of motion, and gait between the two groups before treatment. After one month of treatment, Holden walking function grading, lower limb modified Ashworth muscle tone grading, lower limb Brunnstrom grading, joint range of motion, and gait between the two groups improved compared to before treatment, and the AIoT group was better than the control group, with significance (P<0.05). Moreover, logistic regression analysis showed that AIoT based walking wearable devices was independent risk factor for the development of 90-day readmission in patients with IS after rehabilitation training. AIoT based walking wearable devices for stroke rehabilitation training is feasible and safe with satisfactory therapeutic effects. Moreover, further prospective multicenter trials are warranted before incorporating AIoT into routine rehabilitation training.