Nurse-in-the-Loop Artificial Intelligence for Precision Management of Type 2 Diabetes in a Clinical Trial Utilizing Transfer-Learned Predictive Digital Twin

Syed Hasib Akhter Faruqui,Adel Alaeddini,Yan Du,Shiyu Li,Kumar Sharma,Jing Wang
2024-01-05
Abstract:Background: Type 2 diabetes (T2D) is a prevalent chronic disease with a significant risk of serious health complications and negative impacts on the quality of life. Given the impact of individual characteristics and lifestyle on the treatment plan and patient outcomes, it is crucial to develop precise and personalized management strategies. Artificial intelligence (AI) provides great promise in combining patterns from various data sources with nurses' expertise to achieve optimal care. Methods: This is a 6-month ancillary study among T2D patients (n = 20, age = 57 +- 10). Participants were randomly assigned to an intervention (AI, n=10) group to receive daily AI-generated individualized feedback or a control group without receiving the daily feedback (non-AI, n=10) in the last three months. The study developed an online nurse-in-the-loop predictive control (ONLC) model that utilizes a predictive digital twin (PDT). The PDT was developed using a transfer-learning-based Artificial Neural Network. The PDT was trained on participants self-monitoring data (weight, food logs, physical activity, glucose) from the first three months, and the online control algorithm applied particle swarm optimization to identify impactful behavioral changes for maintaining the patient's glucose and weight levels for the next three months. The ONLC provided the intervention group with individualized feedback and recommendations via text messages. The PDT was re-trained weekly to improve its performance. Findings: The trained ONLC model achieved >=80% prediction accuracy across all patients while the model was tuned online. Participants in the intervention group exhibited a trend of improved daily steps and stable or improved total caloric and total carb intake as recommended.
Machine Learning,Artificial Intelligence
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: How to combine artificial intelligence (AI) and nurses' professional knowledge to provide personalized and precise self - management strategies for patients with type 2 diabetes (T2D) to improve their blood - glucose and weight control. ### Specific problems: 1. **Personalized management needs**: Since individual characteristics and lifestyles have a significant impact on treatment plans and patient outcomes, it is crucial to develop accurate and personalized management strategies. 2. **Limitations of existing methods**: Most existing studies mainly focus on managing blood - glucose levels by controlling insulin infusion devices, while paying less attention to the impact of diet and activity behaviors on blood - glucose and weight. Therefore, an efficient method is required to consider the real - time status of patients and generate supportive measures, such as motivational messages, diet suggestions, and exercise guidance, etc. 3. **Data sparsity and complexity**: Under the conditions of sparse or complex data, how to effectively train the model to adapt to the specific needs of different patients is also a challenge. ### Solutions: This research proposes a "nurse - in - the - loop" artificial intelligence feedback system based on Predictive Digital Twin (PDT), aiming to provide daily personalized feedback and suggestions for T2D patients. Specifically: - **Predictive Digital Twin model**: Utilize Transfer Learning and Artificial Neural Network (ANN) to train a predictive model based on patients' self - monitoring data in the previous three months (such as weight, food logs, physical activity, blood - glucose, etc.). - **Online control model**: Use Particle Swarm Optimization (PSO) algorithm to help patients maintain blood - glucose and weight levels according to the behavior - change suggestions generated by the predictive model. - **Nurse participation**: Adjust and optimize the output of the model through nurses' professional knowledge to ensure the rationality and feasibility of the suggestions. ### Main contributions: 1. Proposed a "nurse - in - the - loop" artificial intelligence feedback system. 2. Developed a Predictive Digital Twin model based on Transfer Learning. 3. Implemented an online control model based on evolutionary optimization for mHealth devices. Through these methods, the research hopes to improve patients' compliance with recommended healthy lifestyle behaviors, thereby better managing blood - glucose and weight.