Scaffolding in AI-Empowered Cognitive Training Improves Outcomes in Cognitive Impairment

NCT07788209 | Active Not Recruiting

• 1 other identifier: 2026229
• Study Type: observational
• Target: 516
• Locations: 1 country
• Sites: 1

Brief Summary

This retrospective study aimed to validate and compare two individualized, closed-loop, AI-driven computerized cognitive training programs with a conventional non-AI training program in a real-world clinical setting. Using objective tablet-recorded data from patients' prior treatment, participants were categorized into three groups according to the training recommendation strategy they actually received. Changes in global cognitive function and specific cognitive domains after 8 weeks of training were compared across groups to evaluate the association between different training strategies and cognitive outcomes.

Conditions

Cognitive Impairment

Keywords

Computerized cognitive training; Artificial intelligence; Scaffolding theory; Cognitive impairment; Personalization

Outcome Measures

Primary Outcomes (1)

• Change in general cognitive ability from baseline (week 1) to week 8

  The general cognitive ability was assessed by the Cognitive Index, calculated from task performance scores. The CCT platforms recorded the Cognitive Index in real time on a daily basis. Weekly Cognitive Index values were computed as the mean of the daily scores recorded over each 7-day period. The change in Cognitive Index from baseline to week 8 is reported.

  From baseline (week 1) to week 8

Secondary Outcomes (1)

• Changes in domain-specific cognitive ability from baseline (week 1) to week 8 across seven cognitive subdomains

  From baseline (week 1) to week 8

Other Outcomes (16)

• Maximum attainable improvement in general cognitive ability

  From baseline (week 1) to week 8

• Rate of improvement in general cognitive ability

  From baseline (week 1) to week 8

• Maximum attainable improvement in perception

  From baseline (week 1) to week 8

Study Arms (3)

Most-Improvement oriented CCT (AIMI)

An AI-based CCT program that based on compensatory principles and prioritized tasks expected to produce the greatest improvement relative to prior performance. Training data indicated that participants self-administered the CCT at home using a tablet or computer over an 8-week period, with sessions lasting at least 20 minutes and occurring approximately three times per week on average.

Behavioral: AI-Based Adaptive CCT Optimized for Cognitive Improvement

Best-Performance oriented CCT (AIBP)

An AI-based program that based on the scaffolding theory and prioritized tasks associated with the highest task performance score. Training data indicated that participants self-administered the CCT at home using a tablet or computer over an 8-week period, with sessions lasting at least 20 minutes and occurring approximately three times per week on average.

Behavioral: AI-Based Adaptive CCT Optimized for Cognitive Performance

Non-AI randomly CCT

A conventional non-AI program in which training tasks were randomly selected within the cognitive domains corresponding to participants' symptomatic impairments. Training data indicated that participants self-administered the CCT at home using a tablet or computer over an 8-week period, with sessions lasting at least 20 minutes and occurring approximately three times per week on average.

Behavioral: Conventional Non-AI CCT

Interventions

AI-Based Adaptive CCT Optimized for Cognitive Improvement BEHAVIORAL

Participants received a multi-domain computerized cognitive training program delivered on a tablet. The program used an AI-based adaptive, closed-loop recommendation strategy to individualize training tasks and difficulty levels, with the goal of maximizing cognitive improvement over the training period.

Most-Improvement oriented CCT (AIMI)

AI-Based Adaptive CCT Optimized for Cognitive Performance BEHAVIORAL

Participants received a multi-domain computerized cognitive training program delivered on a tablet. The program used an AI-based adaptive, closed-loop recommendation strategy to individualize training tasks and difficulty levels, with the goal of maximizing cognitive task performance during training.

Best-Performance oriented CCT (AIBP)

Conventional Non-AI CCT BEHAVIORAL

Participants received a conventional computerized cognitive training program delivered on a tablet. Training tasks were randomly selected within the cognitive domains corresponding to participants' symptomatic impairments and difficulty levels were fixed.

Non-AI randomly CCT

Eligibility Criteria

• Age: 45 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

The study population included patients with cognitive impairment who had received an 8-week tablet-based computerized cognitive training program at Xuanwu Hospital between October 2021 and May 2024. Participants were categorized into three groups according to the training recommendation strategy they actually received.

You may qualify if:

• age ≥ 45 years;
• a Montreal Cognitive Assessment (MoCA) score between 10 and 25;
• were able to communicate in Mandarin.

You may not qualify if:

• no completion of at least 8 weeks of training;
• no engagement in training for ≥ 60 minutes per week;
• diagnosis of depression or other psychiatric disorders;
• diagnosis of systemic diseases, such as hepatic or renal dysfunction, endocrine disorders, or vitamin deficiencies.

Sponsors & Collaborators

• Xuanwu Hospital, Beijing: lead

Study Sites (1)

Xuanwu Hospital, Capital Medical University

Beijing, China

Location

Related Publications (3)

• Liu L, Wang H, Xing Y, Zhang Z, Zhang Q, Dong M, Ma Z, Cai L, Wang X, Tang Y. Dose-response relationship between computerized cognitive training and cognitive improvement. NPJ Digit Med. 2024 Aug 15;7(1):214. doi: 10.1038/s41746-024-01210-9.

  PMID: 39147783 BACKGROUND

• Liu LY, Xing Y, Zhang ZH, Zhang QG, Dong M, Wang H, Cai L, Wang X, Tang Y. Validation of a Computerized Cognitive Training Tool to Assess Cognitive Impairment and Enable Differentiation Between Mild Cognitive Impairment and Dementia. J Alzheimers Dis. 2023;96(1):93-101. doi: 10.3233/JAD-230416.

  PMID: 37742644 BACKGROUND

• Tang Y, Xing Y, Zhu Z, He Y, Li F, Yang J, Liu Q, Li F, Teipel SJ, Zhao G, Jia J. The effects of 7-week cognitive training in patients with vascular cognitive impairment, no dementia (the Cog-VACCINE study): A randomized controlled trial. Alzheimers Dement. 2019 May;15(5):605-614. doi: 10.1016/j.jalz.2019.01.009. Epub 2019 Mar 17.

  PMID: 30894299 BACKGROUND

MeSH Terms

Conditions

Cognitive Dysfunction

Condition Hierarchy (Ancestors)

Cognition Disorders; Neurocognitive Disorders; Mental Disorders

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: RETROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: August 18, 2026
• First Posted: August 26, 2026
• Study Start: August 15, 2026
• Primary Completion (Estimated): August 25, 2027
• Study Completion (Estimated): August 25, 2027
• Last Updated: August 26, 2026

Record last verified: 2026-08

Data Sharing

• IPD Sharing: Will not share

Locations

CN (1)