ScanGrow

ScanGrow analyzes scanned microplate images with deep learning to generate bacterial growth curves as an image-based alternative to spectrophotometric optical density measurements.

Key Features:

• Image-based growth curve generation: Uses deep learning to classify scanned images of bacterial broths in microplates and derive growth curves without direct spectrophotometric optical density measurements.
• Flatbed scanner integration: Processes time-series images captured by flatbed scanners of microplates to produce growth-curve time series.
• Data pre-processing: Performs image and measurement pre-processing to prepare inputs for model training and evaluation.

Scientific Applications:

• Routine bacterial growth monitoring: Monitoring of bacterial growth curves in microplate-format experiments.
• High-throughput time-series experiments: Generation of growth curves for large-scale or frequently sampled studies requiring automated image-based measurement.

Methodology:

Train a deep learning model on images of bacterial broths in microplates and integrate the model into an application that processes flatbed-scanner images to produce growth-curve representations.