cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Choose Language Hide Translation Bar
Using JMP Functional Data Explorer in Predicting the Performance of LigaSure™ Vessel Sealing Devices (2019-US-45MP-212)

Level: Intermediate

 

Jim Pappas, Senior Principal Statistician, Medtronic

 

 LigaSure™ devices, powered by the Valleylab™ FT10 energy platform, typically complete a vessel seal in two to four seconds. With each seal, continuous electrical data, including voltage (V) and impedance (Z) signals, are recorded by the FT10 generator. Understanding key features within this time series data that are critical to device performance is important for developing safe and effective vessel sealing devices.

 

This presentation will demonstrate the use of the Functional Data Explorer (FDE) platform in JMP Pro to characterize time series data associated with vessel sealing. The use of dynamic time warping (DTW) for data preprocessing has proven to be a valuable tool for exploration and interpreting data features. Additionally, FDE has been used to obtain functional principal component (FPC) scores for use in subsequent analyses and predictive models. Graph Builder has been powerful for understanding time-warped data, data features and their relationships with FPC scores. A real-world example will demonstrate interactive JMP platforms, including Graph Builder, FDE, and the Multivariate and Predictive Modeling platforms.

 

 

Comments
DavidLeber

Really appreciated this discussion - helped me understand the realities of applying FDE.

 

One thought re: the R graph he displayed: that seems like the type of graph that you can distill into a single number; something like "Variance of Time Warped" or "Sum of Squared Time Warping", or even "Number of warp lines that cross". That number could then be used to select a better guess for the reference line? That would be a very helpful feature.