Foresighting Outcomes and Risk Evaluation With Computational Artificial Intelligence in Stroke Trials: The FORECAST...

Area of Science:

• Neurology
• Artificial Intelligence
• Public Health

Background:

• Stroke is a major cause of death and disability in Bangladesh.
• Existing risk models do not account for local socioeconomic, environmental, and mental health factors.
• There is a need for a tailored stroke prediction model for the Bangladeshi population.

Purpose of the Study:

• To develop an artificial intelligence (AI)-based risk prediction model for stroke occurrence and poor long-term outcomes in Bangladesh.
• To incorporate unique local risk factors, including rural-urban disparities and mental health comorbidities.
• To establish the first AI-based stroke prediction model for Bangladesh.

Main Methods:

• An ambispective case-control study (FORECAST) enrolling 4,000 participants (2,000 cases, 2,000 controls).
• Development of ensemble machine learning models using comprehensive phenotypic data.
• Application of explainable AI techniques and k-fold cross-validation for interpretability and validation.

Main Results:

• The study protocol outlines the development of an AI-based stroke prediction model.
• Expected accuracy exceeding 90% based on recent advances.
• The model will account for unique Bangladeshi risk factors.

Conclusions:

• The FORECAST study will provide the first AI-driven stroke prediction model for Bangladesh.
• This model will address critical gaps in stroke risk assessment in low-middle-income countries.
• Incorporating culturally relevant factors into AI frameworks enhances stroke prediction accuracy.