Author: Jenghwa Chang, Marissa Joyce Vaccarelli 👨🔬
Affiliation: Northwell, Hofstra University Medical Physics Program 🌍
Purpose: AAPM Report 365 recommends medical physics graduate programs offer courses covering mathematical/statistical methods (section 3.1.7) as well as computational methods/medical informatics (section 3.1.8). In this paper, we present our efforts in developing a new teaching approach that addresses both recommendations in one course.
Methods: The major challenge of designing the curriculum for this course is insufficient semester hours designated for teaching two topics in one. To overcome this obstacle, we created a novel approach to homework assignments. That is, unlike the traditional approach which asks students to complete homework manually, the students in this class were asked to write computer programs to solve most homework questions. These carefully designed assessments not only enhance students’ understanding of the course materials but also require students to utilize appropriate computational methods. Basic Python programming instructions were provided along with each homework assignment, and the students were instructed to learn additional Python coding by self-guided learning.
Results: The final curriculum covered the following key mathematical concepts: signals and systems, Fourier transform, probability, statistical inference, image quality, optimization methods and introduction to machine learning. To solve the homework assignments, the students developed coding skills including plots, numerical integration, convolution, continuous/discrete-time/fast Fourier transform, random number generator, point estimation, confidence interval, hypothesis testing, linear models, DICOM, Rose model, conjugate-gradient descent, and iterative methods for solving systems of linear equations, and support vector machine. This course was offered in the spring 2024 semester and received mixed feedback. As a common theme, most students found the course of educational benefit but indicated a high level of difficulty from the coding assignments.
Conclusion: We successfully developed and implemented a curriculum covering both mathematical/statistical methods and computational methods//medical informatics recommended in AAPM Report 365. We plan to improve this new curriculum in the future by providing rigorous coding guidelines.