By the end of this activity, you will be able to:

• MO#1. Translate Clinical Reasoning into Computational Models
  Students will learn to convert theoretical models of clinical decision-making into computational frameworks using finite-difference methods.
• MO #2 Integrate Generative AI Outputs into MATLAB Workflows
  Students will incorporate AI-generated clinical data into MATLAB scripts to enhance decision tree modeling and outcome prediction.
• MO # 3. Simulate Clinical Scenarios Using Generative AI and MATLAB
  Students will use generative AI tools to generate realistic patient scenarios and simulate diagnostic pathways using MATLAB-based models.

This activity is part of the laboratory component of a lower-level biology course designed for students with no prerequisites. The lab is structured around three integrated activities that introduce students to key concepts in physiology, data analysis, and computational modeling. The first activity focuses on oxygen physiology, where students explore how oxygen is transported and utilized in biological systems. In the second activity, students learn to process and visualize physiological data using MATLAB, building foundational skills in data literacy. The final activity brings together these concepts by having students simulate clinical scenarios using generative AI and MATLAB, allowing them to model decision-making processes and explore the role of artificial intelligence in healthcare. This progression ensures that students, regardless of prior experience, can engage with modern tools and interdisciplinary thinking in a meaningful and accessible way.

Convert a theoretical model of physiological oxygen transport and utilization into a finite-difference computational framework.

Online or in-person Lab Activity:

Case Study: Read the paper by Okpodu and Okpodu-Pyuzza (Training the Next Generation: A Generative AI-enhanced Framework for Teaching Bioinformatics to Life Science Students with Relevance to Cardiovascular Innovation, https://www.clinsurggroup.com/articles/GJMCCR-12-314.php )

Examine how oxygen saturation (SpO₂) levels change in response to physiological stressors such as mild exertion or altitude. The paper provides a conceptual model of oxygen delivery and utilization that can be use as a foundation. After reading the paper run the code finitedifferences.m (Matlab File 845bytes Oct27 25) and then answer the following question:

Discussion Question
Question: How can a finite difference methods enhance our understanding of oxygen transport dynamics in biological systems, and what are the limitations of using these numerical approaches to simulate SpO₂ changes under varying physiological conditions?

Now modify the code to change the diffusion rate to simulate stress or recovery after exercise.  

The paper outlines a model of oxygen delivery and utilization that integrates bioinformatics and computational tools to simulate cardiovascular responses. By applying FDM, students can approximate differential equations governing oxygen transport, enabling them to visualize how SpO₂ levels fluctuate in response to changes in cardiac output, hemoglobin concentration, and tissue oxygen demand. This method supports experiential learning by bridging theoretical concepts with real-world physiological data.

However, there are limitations to using finite-difference methods in simulating SpO₂ changes:

1. Simplification of Complex Systems: FDM often requires assumptions that simplify the biological system, such as linearity or homogeneity in tissue oxygen demand, which may not reflect actual physiological variability.
2. Numerical Stability and Accuracy: The accuracy of FDM depends on the size of the time and spatial steps. Too large a step can lead to instability or inaccurate results, while too small a step increases computational load.
3. Boundary Conditions and Initial Values: The method's reliability hinges on well-defined boundary and initial conditions, which can be difficult to obtain or estimate in biological systems.
4. Limited Scope for Nonlinear Dynamics: Biological systems often exhibit nonlinear behavior, especially under stress. FDM may struggle to capture these dynamics without advanced modifications or hybrid modeling approaches.

In summary, finite-difference methods provide a powerful tool for simulating and understanding oxygen transport, especially in educational settings. Yet, their limitations highlight the need for careful model design and validation against empirical data to ensure physiological relevance.

Rubric for Grading Responses (Microsoft Word 2007 (.docx) 17kB Sep1 25) 
Training the Next Generation (Acrobat (PDF) 375kB Sep1 25) 

Answer: Finite-difference methods (FDM) enhance our understanding of oxygen transport dynamics by allowing us to discretize continuous physiological processes—such as changes in oxygen saturation (SpO₂)—into manageable numerical steps. This approach is particularly useful in modeling time-dependent changes in SpO₂ under varying physiological stressors like mild exertion or altitude, as discussed in the conceptual framework presented in the paper.

The paper outlines a model of oxygen delivery and utilization that integrates bioinformatics and computational tools to simulate cardiovascular responses. By applying FDM, students can approximate differential equations governing oxygen transport, enabling them to visualize how SpO₂ levels fluctuate in response to changes in cardiac output, hemoglobin concentration, and tissue oxygen demand. This method supports experiential learning by bridging theoretical concepts with real-world physiological data.

However, there are limitations to using finite-difference methods in simulating SpO₂ changes:

1. Simplification of Complex Systems: FDM often requires assumptions that simplify the biological system, such as linearity or homogeneity in tissue oxygen demand, which may not reflect actual physiological variability.
2. Numerical Stability and Accuracy: The accuracy of FDM depends on the size of the time and spatial steps. Too large a step can lead to instability or inaccurate results, while too small a step increases computational load.
3. Boundary Conditions and Initial Values: The method's reliability hinges on well-defined boundary and initial conditions, which can be difficult to obtain or estimate in biological systems.
4. Limited Scope for Nonlinear Dynamics: Biological systems often exhibit nonlinear behavior, especially under stress. FDM may struggle to capture these dynamics without advanced modifications or hybrid modeling approaches.

In summary, finite-difference methods provide a powerful tool for simulating and understanding oxygen transport, especially in educational settings. Yet, their limitations highlight the need for careful model design and validation against empirical data to ensure physiological relevance.

finite differences for matlab script teachersv2.docx (Microsoft Word 2007 (.docx) 18kB Oct27 25)

Tips

1. Understanding of Clinical Reasoning — Encourage students to walk through a full diagnostic pathway using the AI-generated patient scenario. Ask them to identify symptoms, propose differential diagnoses, and explain how MATLAB decision trees help simulate clinical logic.
Look for examples like chest pain, shortness of breath, or fever, and how these are processed in a decision tree.

2. MATLab Coding Skills — Encourage students to use MATLAB Copilot to help them debug and refine their code as they calculate diffusion rates. Ask them to describe how they structured their script—for example, how they imported data, defined variables, and calculated rate changes over time. Look for logical flow, clear comments, and evidence that they used Copilot suggestions to improve functionality. You may also ask students to submit screenshots or code snippets to verify clarity and correctness.

3.  Data Interpretation — Guide students to analyze model outputs (e.g., prediction accuracy, confusion matrix, or ROC curves). Ask them to explain what the results mean in a clinical context.   Encourage them to reflect on how feature selection (e.g., lab values, vitals) influenced predictions.

4.  Ethical Reasoning — Look for thoughtful engagement with issues like data validity, privacy, and responsible AI use. Prompt students to consider questions like:

• Is synthetic data representative of real populations?
• What are the risks of bias in AI-generated scenarios?
• How do we ensure transparency and fairness in model training?

5. Participation & Communication — Foster peer-to-peer discussion. Consider using breakout groups or online forums where students can share insights and respond to each other. Reward clarity, engagement, and respectful dialogue. Encourage students to reference course materials or external sources to support their points.

The rubric is attached and explains how to grade the response to the discussion question and measure the student's understanding of physiology, data analysis, and computational skills. Biology_Lab_Assessment_Rubric.docx (Microsoft Word 2007 (.docx) 17kB Oct27 25)