Table 1.

Features of EXACT regarding applications, data set annotation and machine learning.

Features	Description
Application
Online	EXACT is a Django-based server application with a browser client
Cross-platform	The server can be installed on Windows, Linux, Mac and all other systems with Docker support
Multi-center	The multi-center support allows sharing data across multiple institutes with appropriate data privacy management
REST-API	The REST-API supports language independent create, read, update and delete (CRUD) operations on all database fields including image upload and download
Language	The web-server is written in Python while the web-client is HTML and JavaScript based
Plugins	Allow the frontend and backend integration of domain-specific features and analyses
Image-set administration	Combining images to a folder like structure with team access rights and shared annotations scheme
File formats	Images: .tif, .png, bmp, .jpeg, .dcm (partially), .webp
	WSI scanner: .svs, .vms, .vmu, .ndpi, .scn, .mrxs, .iSyntax, .svslide, .bif, .czi, .tif
	Time-Series: .avi, .mkt, .tif
User management	Individual user or group rights for CRUD operations on the database
Annotation
Types	Box, polygon, line, circle and per image (classification) annotations
Templates	Define a framework for general annotation properties like type (box, polygon etc.), colour and default size
Single click	Single click annotations with background knowledge provied by the templates
Guided screening	A persistent screening mode in a user defined resolution which saves the progress
ML
Version control	Enables the versioning of image sets with the corresponding image list and annotations
Inference	Performing inference on client side via browser, on server side via Python or over the REST-API