The Trainable Weka Segmentation (TWS) tool is designed to address the challenge of manually annotating large image datasets, which is time-consuming and labor-intensive. TWS leverages machine learning to train classifiers using a limited number of manual annotations, allowing for automatic segmentation of the remaining data. The tool is interactive, enabling users to guide the training process by providing corrections to the classifier output. TWS supports both 2D and 3D images and can handle various image modalities, including magnetic resonance imaging, two-photon microscopy, and electron microscopy. The TWS tool integrates the image processing toolkit Fiji with the machine learning toolkit WEKA, providing a user-friendly interface for loading and segmenting images. The graphical user interface (GUI) allows users to add traces to classes, pan, zoom, and scroll through the image stack. The training panel includes buttons for training the classifier, toggling the result overlay, creating the result image, getting probability maps, and plotting the model performance chart. The options panel allows users to apply the classifier to other images, load and save classifiers and data, create new classes, and adjust settings such as training features, classifier options, and result overlay opacity. TWS supports macro language compatibility, allowing users to automate tasks using ImageJ macros. The library methods are implemented in a modular and transparent way, making it easy to integrate with other Fiji plugins and scripts. The document also provides detailed instructions on how to script the Trainable Segmentation, including initializing the segmentator, adding training samples, training the classifier, applying the classifier to images, and saving and loading classifiers. The image features used in TWS are categorized into edge detection, texture description, noise removal, and membrane detection. Users can select appropriate features based on the specific segmentation task and the characteristics of the image data. Feature selection is crucial for improving the accuracy of the segmentation results, and the document provides recommendations for choosing the right features and adjusting their parameters. Overall, TWS is a powerful tool for automating image segmentation tasks, offering a combination of machine learning capabilities and user-friendly interfaces, making it suitable for a wide range of imaging pipelines and applications.The Trainable Weka Segmentation (TWS) tool is designed to address the challenge of manually annotating large image datasets, which is time-consuming and labor-intensive. TWS leverages machine learning to train classifiers using a limited number of manual annotations, allowing for automatic segmentation of the remaining data. The tool is interactive, enabling users to guide the training process by providing corrections to the classifier output. TWS supports both 2D and 3D images and can handle various image modalities, including magnetic resonance imaging, two-photon microscopy, and electron microscopy. The TWS tool integrates the image processing toolkit Fiji with the machine learning toolkit WEKA, providing a user-friendly interface for loading and segmenting images. The graphical user interface (GUI) allows users to add traces to classes, pan, zoom, and scroll through the image stack. The training panel includes buttons for training the classifier, toggling the result overlay, creating the result image, getting probability maps, and plotting the model performance chart. The options panel allows users to apply the classifier to other images, load and save classifiers and data, create new classes, and adjust settings such as training features, classifier options, and result overlay opacity. TWS supports macro language compatibility, allowing users to automate tasks using ImageJ macros. The library methods are implemented in a modular and transparent way, making it easy to integrate with other Fiji plugins and scripts. The document also provides detailed instructions on how to script the Trainable Segmentation, including initializing the segmentator, adding training samples, training the classifier, applying the classifier to images, and saving and loading classifiers. The image features used in TWS are categorized into edge detection, texture description, noise removal, and membrane detection. Users can select appropriate features based on the specific segmentation task and the characteristics of the image data. Feature selection is crucial for improving the accuracy of the segmentation results, and the document provides recommendations for choosing the right features and adjusting their parameters. Overall, TWS is a powerful tool for automating image segmentation tasks, offering a combination of machine learning capabilities and user-friendly interfaces, making it suitable for a wide range of imaging pipelines and applications.