Phantom Cameras Beat the Clock in Early Sepsis Detection
To successfully treat sepsis, every minute counts. Unfortunately, traditional lab-based sepsis tests have often been too slow or inaccurate to diagnose this potentially deadly condition in time to successfully treat patients. IntelliSep®, a new test system from Cytovale®, now promises to give more sepsis patients a fighting chance.
IntelliSep combines Phantom high-speed cameras, microfluidic techniques and machine learning algorithms to predict a patient’s likelihood of developing sepsis. The test takes less than ten minutes.
Laying the Groundwork With Microfluidic Cytometry
When developing this game-changing product, Cytovale researchers first carefully studied the physical and mechanical properties of large quantities of cells and gathered data to make statistical predictions. Obtaining this data via traditional methods like atomic force microscopy and micropipette aspiration would have required a lot of time and labor. Instead, the team created a cross-slot microfluidic channel as a stage to observe cell characteristics. They then recorded their experiments with a custom Phantom VEO 710 high-speed camera and inverted microscope.
While suspended in a solution, hundreds of thousands of cells flowed through the channel at 3 meters per second (m/s) to an observation zone measuring 45 by 120 micrometers (μm), where the cells were subjected to hydrodynamic stresses to induce deformations. Recording at 500,000 frames per second (fps) at 102 x 48 resolution, the VEO 710 camera captured roughly 2,000 cells per second.
Each run involved 50,000 single cell events and provided researchers with cell metrics like size, morphology and strain rate. Critical to the application’s success, the VEO 710 features a custom CMOS sensor that provided a 7 gigapixel per second (Gpx/s) image throughput and fast data offload speeds to ensure IntelliSep’s 10-minute sample turnaround time.
Machine Learning Enables Key Insights
To determine the likelihood of sepsis based on the physical characteristics of cells, the Cytovale team had to determine which properties were better indicators of sepsis. To this end, they trained support vector machines (SVM) that analyze data for classification and regression.
The SVMs allowed the researchers to classify cell types and weigh the importance of cell parameters. By iteratively training the SVMs with larger collections of cells and better-ranked parameters, the team refined the accuracy of IntelliSep’s predictive model. The VEO 710 camera was integral in this machine learning process, providing a large quantity of high-fidelity data.
Not only did the Phantom VEO 710 support the foundational research for Cytovale’s new test, it also facilitated IntelliSep’s production. The camera enabled Cytovale to scale its design to a commercially available tool, making it the first time an embedded Phantom high-speed camera was used within a medical device with FDA 501(k) clearance.
Learn more in the case study.
To successfully treat sepsis, every minute counts. Unfortunately, traditional lab-based sepsis tests have often been too slow or inaccurate to diagnose this potentially deadly condition in time to successfully treat patients. IntelliSep®, a new test system from Cytovale®, now promises to give more sepsis patients a fighting chance.
IntelliSep combines Phantom high-speed cameras, microfluidic techniques and machine learning algorithms to predict a patient’s likelihood of developing sepsis. The test takes less than ten minutes.
Laying the Groundwork With Microfluidic Cytometry
When developing this game-changing product, Cytovale researchers first carefully studied the physical and mechanical properties of large quantities of cells and gathered data to make statistical predictions. Obtaining this data via traditional methods like atomic force microscopy and micropipette aspiration would have required a lot of time and labor. Instead, the team created a cross-slot microfluidic channel as a stage to observe cell characteristics. They then recorded their experiments with a custom Phantom VEO 710 high-speed camera and inverted microscope.
While suspended in a solution, hundreds of thousands of cells flowed through the channel at 3 meters per second (m/s) to an observation zone measuring 45 by 120 micrometers (μm), where the cells were subjected to hydrodynamic stresses to induce deformations. Recording at 500,000 frames per second (fps) at 102 x 48 resolution, the VEO 710 camera captured roughly 2,000 cells per second.
Each run involved 50,000 single cell events and provided researchers with cell metrics like size, morphology and strain rate. Critical to the application’s success, the VEO 710 features a custom CMOS sensor that provided a 7 gigapixel per second (Gpx/s) image throughput and fast data offload speeds to ensure IntelliSep’s 10-minute sample turnaround time.
Machine Learning Enables Key Insights
To determine the likelihood of sepsis based on the physical characteristics of cells, the Cytovale team had to determine which properties were better indicators of sepsis. To this end, they trained support vector machines (SVM) that analyze data for classification and regression.
The SVMs allowed the researchers to classify cell types and weigh the importance of cell parameters. By iteratively training the SVMs with larger collections of cells and better-ranked parameters, the team refined the accuracy of IntelliSep’s predictive model. The VEO 710 camera was integral in this machine learning process, providing a large quantity of high-fidelity data.
Not only did the Phantom VEO 710 support the foundational research for Cytovale’s new test, it also facilitated IntelliSep’s production. The camera enabled Cytovale to scale its design to a commercially available tool, making it the first time an embedded Phantom high-speed camera was used within a medical device with FDA 501(k) clearance.
Learn more in the case study.