Medtronic, a supporter of the Student Postdoctoral Scholar Independent Research Scholarship, hosted scholarship recipient Perry “PJ” Johnson on August, 13, 2024, at their Memphis office.
Johnson presented his research project “Predictive Measures of Grand Reaction Forces Using Inertial Measurement Units and Machine Learning Technology” to guests, and a tour of the Medtronic facility followed.
Dean Donald “Don” Thomason asked PJ a few questions about his research scholarship.
Dean Thomason: What insight(s) has the independent research project given you about your research?
PJ Johnson: This IMU and Machine learning research project has provided a new perspective for how we can take this technology and apply it to the rehabilitation of treating tibial bone stress injuries (TBSI). Researchers are trying to develop a consistent and universal method for how clinicians and researchers can best treat TBSI. There is no explicitly detailed universal standard, all clinicians follow a general pattern of treatment but tailor the treatment to their own experience and patient(s). Generally, patients who are diagnosed with a TBSI go through a period of reduced loading, then so long as they have no pain in their injured limb while walking and during certain activities, they are able to progressively load their limb to return back to their sport.
For this machine learning project, we are aiming to utilize inertial measurement units (IMU) to gather acceleration data and input the acceleration data into a machine learning model to help estimate ground reaction force – time curve throughout the stance phase of running then be able to estimate the tibial compressive force. With this understanding, it might be possible to understand how a patient, who is rehabbing from a TBSI, is loading their tibia and help them to know if they can continue to progressively load their tibia as the bone is healing or if they need to back off and provide more time to rest.
Dean Thomason: Is there something that you have discovered that surprised you?
PJ Johnson: So far, what has surprised me is the various machine learning models. Depending on the task, will dictate the type of learning methods you will utilize for your model.
Machine learning is a field that focuses on the learning aspects of Artificial Intelligence (AI) by developing algorithms that best represent a set of data. It uses subsets of data to generate an algorithm that may use novel or different combinations of features and weights.
Dean Thomason: Have you developed a skill, as a result of the project, that you think would be valuable to you and an employer?
PJ Johnson: The skill I am developing through this project is proficiency in Python computer programming language for developing a machine learning model. I’m currently familiar with MATLAB computer programming so learning another computer programming language can help me to become more versatile in the future when applying to jobs.
Johnson’s scholarship was funded in part by a grant from Medtronic. If you’d like to support student research, please make your online gift for the program of your preference: http://giving.uthsc.edu/CGHSStudentResearch.