Session S07: Physics in Medicine: Sensors, Models, and Data Analysis

Live

Parameter estimation for biexponential models is of importance throughout the biomedical sciences, where simultaneously occurring exponential processes yield a joint signal. One example of particular interest is magnetic resonance relaxometry, in which biexponential analysis is used to characterize tissue samples. Much work has been performed to improve parameter estimates for this notoriously ill-posed problem, including neural network (NN) approaches. Typically, these perform on par with nonlinear least-squares (NLLS) methods; this is not surprising, since the application of a NN does not change the problem's underlying condition number. We present a new method for NN-based biexponential analysis, in which regularized versions of an experimental decay are presented to the NN. This is fundamentally different from conventional error function regularization at the output layer. With our method, the NN's input is a collection of signals, which we call a regularization trajectory, defined by parameter recovery via nonregularized NLLS and Tikhonov regularized NLLS. We demonstrate that training on regularization trajectories results in a uniform and substantial improvement in both the accuracy and precision of parameter estimates across a wide range of biexponential models.   

Live

Neuromuscular hand strength evaluation is challenging with the growing importance of telemedicine. Objective documentation of the hand function requires a simple procedure. This study tests a method of measuring finger pinch-strength without a clinical visit to determine if a patient with a work-related injury meets criteria to be eligible for medical insurance benefits. The procedure involves pinching a piece of paper that is folded multiple times and wrapped around a bucket handle. Diminished pinch strength will not support a specified weight of water in the bucket without slipping. An individual who meets criteria using this method may qualify for compensation. Volunteers in this pilot study provide calibration for this procedure, each doing 50 repeated trials for each hand. The weight at which the apparatus began to slip was measured. The coefficient of friction of skin on paper was calculated based on the weight of the apparatus and compared with a standard measurement using a clinical pinch dynamometer. These measurements and the calculated coefficient are presented in this paper.   

Live

Transport of cells in fluid flow plays a critical role in cancer metastasis. Recent development of microfluidic devices has enabled the understanding of cellular dynamics in vitro conditions such as tubes and channels. However, it is challenging to obtain precise characteristics of cellular dynamics using experimental method alone in complex situation such as human bone. This challenge motivates our new development of computational methods to provide hemodynamic data of cellular flows in complex geometries. Past computational works have utilized either particle-based methods (Disspative Particle Dynamics, Lattice-Boltzmann Method) or continuum approach (Finite Element Method). Since there exists a large disparity in spatial and temporal scales in this problem, it is highly desirable to develop a hybrid continuum-particle approach to resolve local dynamics of cells while providing large-scale flow patterns in the vicnity of bone and the vascular vessel. In this work, a new Fluid-Structure Interaction formulation is proposed based on the use of DPD model for the cellular membrance and the use of the Immersed Boundary Method for the fluid plasma. This hybrid approach allows accurate descriptions of cell deformation as well as resolving flow structures in details.   

Live

This talk continues the presentation Mathematical Models for Living Forms in Medical Physics Submodel 1: The information processing from teeth to Nerves from the American Physical Society April Meeting 2020 Conference. In the Submodel 1 the information processing from teeth to the nerves is modeled. The information is passed via p-waves through the tooth layers enamel and dentin. Odontoblasts located in the liquid in the tubules of the tooth dentin layer perform finally the transformation into electrical information (an electrical signal) that passes along nerves. The Submodel 2 of the project is dedicated to the information coding of the information from an entity hitting/touching a tooth and to the information processing of the coded unit through the nerves. Emphasized are the information representation as an electrical code and the coded information flow in the living system.   

On Demand

Measuring biological interaction affinities, specifically protein-protein interaction (PPI), is fundamental to biochemistry, yet many critical interactions are unmeasurable due to a scarcity of reagents and limitations in the affinity ranges that can be measured. Here we present a novel technique which utilizes rolling magnetic probes (RMP) to measure PPI interaction affinities by leveraging the fundamental physical concept of friction induced by binding interactions. RMP measures the translational displacement of protein coated particles on a protein functionalized substrate induced by the increase in friction due to the strength and density of the biological interactions. The translational displacement scales with the effective friction induced by the biological interaction, producing a mechanical signal to indicate binding event. RMP can measure interactions across a wide range of affinities, $10^{-3}-10^{-15}$M, has high resolution, and measures $\Delta \Delta$G differences of approximately 0.4$\frac{kcal}{mol}$, and utilizes a small amount of reagents, 20 pmol. Here RMP provides quantitative insights into the interplay between epigentic modifications.