An Ecological Momentary Assessment Based Program to Support Self-Management for Post-stroke Patients

Combining Machine Learning and Ecological Momentary Assessment of Depression, Cognitive Function and Daily Activity Behavior to Support Self-Management for Post-stroke Patients: A Feasibility Study

The goal of this feasibility study is to test whether a new approach that combines real-time symptom tracking (Ecological Momentary Assessment) with machine learning can help people recovering from a stroke to better manage depression, thinking difficulties, and daily functioning. The main questions it aims to answer are:

Is this combined approach practical and acceptable for post-stroke survivors? Does the program improve mood, cognitive function, or functional ability to carry out daily activities?

Participants will:

Use a smartphone app called RehabCare Companion for a period of time Answer brief daily surveys about their mood, thoughts, and activities for one week Receive personalized self-care suggestions generated by machine learning based on their responses Complete assessments of depression, cognition, and functioning before and after the program Take part in a group interview to share their experiences

Inclusion Criteria:

• adults aged 18 years or older
• having a confirmed diagnosis of stroke
• having sufficient proficiency in Cantonese to understand and respond to the assessments and prompts
• having a Hong Kong version of the Montreal Cognitive Assessment (HK-MoCA) score of 18 or higher
• having a smartphone.

Exclusion Criteria:

• they have moderate to severe cognitive impairment (a HK-MoCA score of 17 or lower)
• they have active psychiatric disorders other than depression (e.g., schizophrenia, bipolar disorder)
• they have a terminal illness or prognosis that suggests a limited life expectancy
• they have conditions that compromise their self-care ability (i.e., bedbound).