iCLOTS
Lead data scientist and senior software developer
iCLOTS (https://www.iclots.org) is an open-source software designed to automate image analysis for microfluidic-based microscopy experiments. It addresses the lack of easy-to-use, adaptable analysis tools for experiments involving fluid flow, cell adhesion, and dynamic cell tracking in a wide range of biomedical assays. iCLOTS combines traditional image analysis with machine learning algorithms to extract detailed quantitative metrics from complex datasets, helping researchers overcome the limitations of manual data analysis.
The software supports four primary assay categories
Some of the key challenges I addressed as a part of this project include:
- Lack of specialized tools: Existing image analysis software (e.g., ImageJ, Ilastik), while excellent, left a specific gap in tracking capabilities. iCLOTS was designed as an image analysis software tailored for time-dependent and fluid flow experiments, eliminating the reliance on manual data analysis and reducing errors.
- Efficient data management: Methods were developed to handle large, multi-dimensional datasets from microfluidic systems, enabling the detection of small yet significant changes in cell behavior, particularly in diseases like SCD and sepsis.
- Interpretability: Each cell has individual behaviors that may contribute to important subpopulations. Additionally, computational methods can be a black box where interpretability is limited. By providing an index and single-cell metrics for every cell detected, users could feel confident they understood their results.
- Increased accessibility:iCLOTS was made to be user-friendly and accessible to researchers without programming expertise, allowing advanced computational tools to be used without requiring coding skills.
- Enhanced reproducibility:The design of iCLOTS minimizes bias and manual errors, significantly improving the reproducibility of results compared to traditional analysis methods.
- Broad applicability:iCLOTS was engineered to be adaptable across various biospecimens and microfluidic devices, accommodating different research needs without the need for significant modifications.
Skills developed include: advanced image processing, computer vision, in vitro experiments, microfluidic experiments, hematology experiments, standalone software development, open-source work, machine learning, AI