Related Experiment Video
Updated: Jul 16, 2025
07:29
Reconstruction of Single-Cell Innate Fluorescence Signatures by Confocal Microscopy
Published on: May 27, 2020
2.8K
Using Recurrent Neural Networks to Reconstruct Temperatures from Simulated Fluorescent Data for use in
Jacob Kullberg1, Derek Sanchez2, Brendan Mitchell2
1Computer Science Department, Brigham Young University, 3361 TMCB, Provo, 84602, UT, USA.
Research Square
|September 18, 2023
Summary
A new neural network, the Multi-Directional Fluorescent Temperature Long Short-Term Memory Network (MFTLSTM), offers high-accuracy, pixel-level temperature mapping for microfluidic systems. This advancement enables precise temperature monitoring in biological studies.
Area of Science:
- Biophysics
- Microfluidics
- Machine Learning
Background:
- Biological systems often require precise temperature control within narrow operating ranges.
- Existing temperature sensors struggle to provide both high accuracy and spatial resolution in microfluidic devices.
- Accurate temperature monitoring is crucial for studying isolated biological systems.
Approach:
- Introduced the Multi-Directional Fluorescent Temperature Long Short-Term Memory Network (MFTLSTM), a novel neural network.
- Developed MFTLSTM to calculate temperature at every pixel in fluorescent images, improving upon standard fitting methods.
- Leveraged heat diffusion principles within images for enhanced temperature calculation.
Key Points:
- MFTLSTM achieved an accuracy of ±0.0199 K RMSE with simulated data (298K-308K).
- Applied to experimental data from 3D printed microfluidics, it reached ±0.0684 K RMSE (290K-380K).
- The network effectively relates fluorescent data to temperature with high precision.
Conclusions:
- MFTLSTM provides superior temperature resolution compared to existing methods in microfluidic applications.
- This technology has the potential to significantly advance research in temperature-sensitive biological systems.
- Enables high-accuracy, spatially distributed temperature measurements critical for microfluidic biological studies.
