The primary challenge in analyzing colocalization in fluorescence microscopy images is ensuring the reliability of results, which is influenced by various technical factors, particularly the handling of background noise. Over the years, we have developed tools to address this issue.
Our latest update introduces Smart Background Correction, a feature that allows for corrections on a selected region of interest (ROI) within an image, instead of applying it to the entire image. However, the algorithm was designed to identify Gaussian noise, which is common in photography, making it less effective for fluorescence images that typically exhibit Poisson noise.
We are excited to present CoLocalizer Pro 6, now enhanced with Machine Learning (ML) for improved background correction! ML is an advanced technology that allows computers to learn from data without explicit programming, enabling them to function more like humans.
We developed an ML model using tens of thousands of real fluorescence microscopy images from the three most common types of fluorescence microscopy (wide-field, two-photon, and confocal). This model accurately recognizes different noise levels with over 90% prediction accuracy and is seamlessly integrated via Core ML, ensuring fast performance.
To use the ML background correction model, simply open the image, navigate to the Background Correction panel, and click the ML Correct button. The model will instantly analyze the image, predict its noise level, and apply the appropriate correction settings. Automagically. This process occurs directly on your device, and the entire trained ML model is packaged within the CoLocalizer Pro app.
We believe that the introduction of ML Correction will be highly beneficial to you. Given the promising results and the potential impact on research, CoLocalizer Pro 6 for Mac represents one of the most significant updates since its original release. We are eager to see how it enhances your studies.
We welcome your suggestions for additional ML-based functionalities you would like to see incorporated into CoLocalizer. In the meantime, you can download this new version and utilize ML technology to process your images, providing you with full confidence in your data!
Note: ML Background Correction requires macOS 10.14 Mojave or newer.
