A regularized higher order Markov clustering algorithm for monitoring chronic health conditions

Nutalapati, Phani Teja
Journal Title
Journal ISSN
Volume Title

The average life expectancy of humans nowadays is predicted to be 65-70 years. There are many reasons for their death. The most natural way of death is because of their age factor, but this is no more resulting in most cases of illness or death. Out of all the reasons, chronic diseases are the leading cause of death and disability around the world and United States. It accounts for 70% of all deaths in U.S., which is 1.7 million each year. Data from the World Health Organization show that chronic disease is also the major cause of premature death around the world. The term chronic is usually applied when the course of the disease lasts for more than three months. For instance, a chronic headache for a long time may lead to brain hemorrhage and long term post-traumatic stress disorder leads to insufficient sleep, where insufficient sleep leads to diabetics, cardiovascular arrest. So, one kind of chronic disease is leading to multiple chronic conditions (MCC) which reduces the life expectancy. Technologies are growing fast to increase the health of humans through development in biomechanical devices. But the path the diseases are following has much more importance than development in the above fields. This study helps to identify major patterns of MCC transitions in patients with multiple chronic conditions (MCC). For this purpose, first, MCC transition matrix is calculated using historical data. Next, Markov Clustering Algorithm (MCL) is employed to partition the transition matrix into major MCC patterns. In addition, Markov Clustering Algorithm is regularized and extended to higher orders to increase the quality and accuracy of the results. The algorithm results is verified using a real dataset from a local medical center.

This item is available only to currently enrolled UTSA students, faculty or staff.
Mechanical Engineering