Discovery

Visiopharm's Discovery is an AI-powered image analysis software designed for researchers working with tissue-based images. It is the first entry-level image analysis package to include AI deep learning as standard, enabling scientists of all experience levels — from beginners to advanced users — to tackle both simple and complex tissue datasets and generate reliable, quantitative results without requiring coding or prior image analysis expertise.

Discovery is built around the principle of turning expert biological knowledge into reproducible outcomes. By combining intuitive annotation tools, pretrained quickstart APPs, and deep learning capabilities, the platform allows any researcher with an understanding of tissue morphology to build, train, and apply analysis workflows across large datasets with confidence.

Supported Image Formats and Data Import

• Supports all major whole slide image formats, including Aperio (SVS, AFI), Nikon (ND2), 3DHistech (MRXS), Olympus/Evident (VSI), Hamamatsu (NDPI, NDPIS), Leica (SCN), Zeiss (CZI), Ventana (BIF), Philips (iSyntax, i2Syntax), JPG, TIF, BMP, OME.TIFF, and many more.
• Import annotations from other platforms to continue existing analysis pipelines where they leave off.
• Transfer and reuse existing annotations, and build classifiers by training on result masks.
• Enables easy transition of projects from other image analysis systems.

Defining Analysis with Quickstart APPs

• Pretrained quickstart APPs allow researchers to get started immediately, with the ability to combine, modify, and tailor APPs using expert knowledge.
• Add annotations or layers to customise APPs for specific research needs, and configure specific data exports.
• Available for all image modalities, including support for Bruker's GeoMx DSP to define areas of illumination.
• Example APPs include: GeoMx DSP Cell Detection, Tissue Detection (Brightfield), GeoMx DSP panCK Detection, Nuclei DAB Intensity Quantification, Nuclei Detection AI (Brightfield), IHC Glomerulus Detection, IHC Germinal Center Detection, and more.
• AI deep learning is standard across all research applications — no coding, no time-consuming setup, and no prior image analysis experience required.

AI Deep Learning Workflow: See It, Train It, Find It

1. See it: Apply tissue expertise by annotating the structures to be identified using an intuitive drawing tool that enables quick and seamless annotation.
2. Train it: The software automatically characterises the best features to robustly detect structures, even across heterogeneous tissues, without any manual coding.
3. Find it: Once trained, APPs can be applied to large datasets using batch processing for walk-away analysis. APPs are designed to be enriched and reused — re-annotating regions allows the APP to grow alongside the research.

Image Analysis Capabilities

• Batch mode: Run batch processing across entire studies with any number of research APPs, queuing analysis tasks to run in the background while working on other tasks simultaneously.
• Parallel processing: Discovery makes optimal use of available computational resources for parallelised, high-speed processing of image analysis jobs, improving overall productivity.

Results Exploration and Export

• Review cells and results in interactive plots and images simultaneously to quality-control data analysis and verify outliers or rare events.
• Clicking on a data point in a graph or plot opens the corresponding image or TMA and highlights the relevant cell in context, along with a thumbnail gallery view.
• Filter plots by specific regions, features, or split views between treatment groups or tissue data for deeper data exploration.
• Export all data, images, and plots in standard formats suitable for reports or publications.
• Perform subsequent analysis in external pipelines using .tsv files containing all detailed exports as saved to the database.

Additional Modules

• Tissuearray: A comprehensive automated solution for TMA (tissue microarray) analysis that manages all metadata of individual cores for data correlation. Provides customisable grid layers for an easy and quick workflow, with the ability to import grid layouts and block designs from TMA spotters.
• Tissuealign: Enables precise alignment of multiple serial sections of any image modality, combining different sections or stain cycles into one flexible analysis. Supports alignment of brightfield IHC or ISH with H&E, as well as IF or IMC images.

Training, Support, and Updates

• Easy onboarding through Visiopharm Academy courses and on-demand training session opportunities.
• Comprehensive training materials and documentation available for self-learning and problem solving.
• Continuous improvements delivered through regular software updates and maintenance via the Software Maintenance Service (SMS).
• Responsive expert support team recognised by customers for quick and efficient resolution of queries and issues.

Discovery is suitable for a wide range of research environments, from academic institutions to contract research organisations and pharmaceutical companies. Its flexible APP design allows analysis approaches to be tailored to study requirements rather than constraining study design to software limitations, making it a versatile platform for advancing tissue-based scientific discovery.