Graduate Trainee, University of Minnesota
1 active project
Machine Learning Model for Heart Attack Risk
Scientific Questions Being Studied
Our research project aims to design an machine learning model to estimate the risk of myocardial infarction (heart attack) between regular doctor visits. This model will incorporate a wide range of data, such as individual health history, demographic information, and actigraphy data (which captures rest/activity cycles). By utilizing these diverse data sources, we aim to identify patterns, correlations, and risk factors that might otherwise remain hidden in traditional analytical methodologies. Health history can provide insights into predisposing conditions, while demographic data can shed light on socio-economic and lifestyle factors. Simultaneously, actigraphy data will allow us to understand the influence of physical activity and sleep patterns on cardiovascular health. The ultimate goal of this project is to enhance the predictive accuracy of heart attack risk, providing healthcare professionals with a powerful tool for early intervention.
- Disease Focused Research (myocardial infarction)
- Social / Behavioral
We believe that medical history can reveal predispositions due to past health conditions; demographic data can highlight socio-economic, age, or ethnicity-related risk factors, while actigraphy data can offer insights into the influences of lifestyle, physical activity, and sleep patterns on heart health. We will examine various machine learning algorithms utilizing the sklearn platform in Python. This will include traditional models like decision trees, random forests, support vector machines, and more contemporary incremental learning methods like passive-aggressive algorithms. Simultaneously, we aim to leverage the TensorFlow library to explore deep learning architectures. Deep learning models, such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory networks (LSTMs), offer the potential to detect complex, non-linear relationships within our data, often unseen by traditional methods.
We hope to find valuable insights into the multifaceted influences of health history, demographic factors, and actigraphy data on myocardial infarction risk. By developing a machine learning model capable of accurately predicting heart attack risk between doctor visits, we aim to contribute a crucial tool to the arsenal of predictive medicine. This could lead to a significant shift in the management of cardiovascular health, moving towards personalized risk assessment, early interventions, and potentially improved patient outcomes. Our exploration of various machine learning and deep learning models also stands to enrich the field's understanding of these methodologies' applications in healthcare. The comparative analysis of traditional, deep learning and incremental learning methods could provide evidence-based guidance for researchers and practitioners on the strengths and limitations of these techniques in medical contexts.
Demographic Categories of Interest
This study will not center on underrepresented populations.
Data Set UsedRegistered Tier
- Trevor Winger - Graduate Trainee, University of Minnesota
You can request that the All of Us Resource Access Board (RAB) review a research purpose description if you have concerns that this research project may stigmatize All of Us participants or violate the Data User Code of Conduct in some other way. To request a review, you must fill in a form, which you can access by selecting ‘request a review’ below.