Date of Graduation
5-2026
Document Type
Thesis
Degree Name
Bachelor of Science in Biomedical Engineering
Degree Level
Undergraduate
Department
Biomedical Engineering
Advisor/Mentor
Dr. Raj Rao
Abstract
Background: Chronic Obstructive Pulmonary Disease (COPD) is the fourth leading cause of death worldwide, disproportionately affecting low- and middle-income countries (LMICs). Access to therapeutic equipment in these settings remains severely constrained, and existing reliance on foreign device donation has proven unsustainable. Inspiratory Muscle Strength Training (IMST) offers a clinically supported rehabilitative therapy for COPD management, but the design of a locally manufacturable, low-cost IMST device constitutes a constrained, multi-objective optimization problem for which classical, trial-and-error engineering methods are poorly suited. As such, this work proposes Bayesian Optimization (BO) to be a superior, alternative paradigm due to speed and computational cost-effectiveness.
Methods: A two-stage Bayesian Optimization (BO) procedure was developed: Stage 1 identifies the device geometry that delivers a clinically relevant target pressure-drop at a representative inspiratory flow rate of 60 L/min, via a fluid mechanics model incorporating frictional and orifice-plate pressure losses; Stage 2 identifies the manufacturing strategy that yields the lowest per-unit cost for the Stage 1 geometry, via an Arkansas-based cost model comparing silicone 3D printing and injection molding.
Results: Stage 1 converged to an optimal geometry delivering a pressure-drop within 0.1% of the clinical target of 31.5 cmH₂O, in 43 seconds. Stage 2 identified injection molding as the dominant manufacturing route at a per-unit cost of $1.11, compared to $4.42 for 3D printing under the same conditions, with a combined optimization runtime of 78.8 seconds.
Conclusion: The proposed two-stage BO paradigm produced a clinically targeted, manufacturable IMST device design at a price point compatible with LMIC deployment. This paradigm is applicable to a broader class of frugal-engineering problems in which clinical performance must be reconciled with cost-constrained, locally appropriate manufacturing.
Keywords
frugal engineering; global health innovation; optimization; medical devices
Citation
Alserhan, S., & Rao, R. (2026). Bayesian Optimization for Frugal Engineering Design: A Case Study on Inspiratory Muscle Strength Training Devices for Low-Resource Settings. Biomedical Engineering Undergraduate Honors Theses Retrieved from https://scholarworks.uark.edu/bmeguht/177
Included in
Analytical, Diagnostic and Therapeutic Techniques and Equipment Commons, Biomechanics and Biotransport Commons, Biomedical Devices and Instrumentation Commons
