Research Projects

Rationale: Motility of sperm cells is a strong indicator of their ability to fertilize the egg in the female reproductive system. However, sperm cells exhibit diverse movement characteristics from moving progressively and travelling large distances, to either moving in circles (non-progressive) or being simply immotile. Hence, quantifying their motility is crucial to the diagnosis of infertility. WHO standards sperm analysis computes a set of kinematic parameters, from the observed trajectory of a sperm cell, to identify its motility-class. This method fails to consider: (1) the intrinsic activity of the sperm cell (nature of swimming) and, (2) the effect of interactions in the inference problem, making it inaccurate and not suitable for samples with large number of sperm cells. (Additionally, mainly Caucasian sperm samples are used for analysis.) Scientific objectives: Characterize the intrinsic motility of sperm cells from data. Self-propelling particles are modeled based on their specific movement characteristics—Active-Brownian/Levy-walk, run and tumble, burst and coast, etc. We do not yet have a class of models that capture the unique and diverse intrinsic movement of sperm cells. We will derive these models directly from data. Disentangle interaction-effects from intrinsic activity to classify them. Collisions and hydrodynamic interactions with other cells and debris in the semen sample render equation discovery challenging. A new framework will be developed to disentangle interaction-effects from intrinsic motility to classify sperm cells accurately. Hypothesis: Simple closed-form expressions for the propulsion and turning rates of the motile cells can be derived from data. Sperm cells belonging to different classes (progressive, non-progressive, hyperactive) will have qualitatively different activity models. Intrinsic motility of sperm cells can be accurately inferred from movement data if interactions are considered allowing us to distinguish for instance, a non-progressive sperm from a motile one trapped between other cells. Study planned: Methods to detect and track the movement of sperm cells from video data. Data-driven methods to identify models for sperm movement from tracked trajectories. Incorporating interaction-effects into the equation discovery procedure to get robust inferences. End-to-end systems from tracking sperm movement to estimating its intrinsic motility. Significance. Fundamental: Uncover a new class of self-propelled particle models that describe the diverse sperm movement seen. New methods to infer intrinsic properties in the presence of strong interactions with the neighborhood. Application: End-to-end toolbox to characterize the motility of sperm cells. In this pipeline, movement information will be extracted from microscopy videos, intrinsic motilities will be characterized while accounting for interactions. The toolbox will facilitate effective mapping of the fertility landscape of Indian males.