DLgram cloud service for deep-learning analysis of microscopy images

Area of Science:

• * Science and technology image analysis
• * Quantitative image analysis
• * Deep learning applications

Background:

• * Manual image analysis for object counting and parameter determination is time-consuming and labor-intensive.
• * Existing methods may be limited by subjective factors and require specialized programming skills.
• * There is a need for a user-friendly and efficient tool for quantitative image analysis.

Purpose of the Study:

• * To introduce DLgram, a universal, user-friendly cloud service for automated image analysis.
• * To demonstrate the application of deep learning for object recognition and parameter determination.
• * To reduce the time and subjectivity involved in quantitative image analysis.

Main Methods:

• * Development of a cloud-based service (DLgram) utilizing deep learning technologies.
• * User-friendly interface requiring minimal input (labeling objects in images).
• * Automated neural network training for object recognition and parameter extraction.

Main Results:

• * DLgram successfully automates object recognition, counting, and parameter determination in images.
• * The service allows for user correction of results, including error removal and addition of missing objects.
• * Mathematical processing of extracted data provides information on object sizes, areas, and coordinates.

Conclusions:

• * DLgram significantly reduces time spent on quantitative image analysis.
• * The service minimizes the influence of subjective factors, increasing analysis accuracy.
• * DLgram is a powerful, accessible tool for diverse scientific and technological image analysis needs.