Advancing Discovery with Trainable AI

Insightful and robust AI creation for pathology

Unlock Limitless Possibilities with HALO AI

HALO AI has you covered. Whether you’re looking to solve image analysis challenges in segmentation, classification, or phenotyping, the advanced deep learning networks of HALO AI are easily tuned for your specific application with the train-by-example interface.


Diverse and Powerful AI

Underpinned by modern deep learning networks, HALO AI provides a collection of segmentation, classification, and phenotyping tools for brightfield and fluorescence applications. Train a classifier to quantify tissue classes, to find rare events or cells in tissues, or to categorize cell populations into specific phenotypes.

AI-Powered Annotation

Use a point and click workflow to rapidly annotate images and develop training data.

Nuclear Segmentation

Choose between several trainable AI-based nuclear segmentation networks to optimize segmentation accuracy when nuclear morphologies vary.

Membrane Segmentation

Train AI-powered membrane segmentation networks to accurately delineate cell membranes for your specific assay.

Tissue Segmentation

Train AI-based MiniNet and DenseNet networks to classify tissue into user-defined classes.

Cell & Object Phenotyping

Quickly train a phenotyper to quantify cell types or objects of interest.

Slide Quality Control

Leverage a trainable AI-powered quality control network to detect common artifacts in H&E and IHC images.

Simple & Intuitive Workflow

HALO AI is fully integrated with the intuitive, easy-to-use HALO® and HALO Link platforms and employs a simple three-step workflow. After defining tissue classes or cell phenotypes, you train the neural network by drawing annotations – no computer programming or AI knowledge required. 

Define Classes

Train Network

Apply Classifier

Choose to integrate your trained HALO AI networks in multiple locations within HALO modules to perform functions including tissue classification, nuclear and membrane segmentation, and phenotyping. Alternatively, apply HALO AI directly to an entire study of whole slide images or to select regions of interest – the choice is yours.

Manage Variability

Extreme variability runs rife in image analysis but is no match for HALO AI. Common sources of variability include diverse morphologies, alterations of morphologies from staining protocols, differences in tissue quality, uneven staining, and more.

HALO AI can be easily trained to accommodate variability to deliver accurate segmentation and classification results across large studies. HALO AI can even be trained to work across vastly different stains such as PAMS, Trichrome, H&E, and IHC.

Exceptional Identification

Pre-trained nuclear and membrane segmentors available in HALO are advanced tools for nuclear and membrane segmentation, but when you need to optimize a network for a bespoke application, you need HALO AI.

As with the tissue classification and segmentation networks, you can quickly add training data with the AI-based annotation tool and train the network to optimally segment nuclei or cell membranes. And once trained, any HALO AI network can be incorporated into HALO modules for maximum utility.

Powerful Tissue Classification

The degree of variability encountered in pathology often demands an approach above and beyond a simple random forest tissue classifier.

HALO AI has two advanced pre-trained networks capable of creating high-resolution classifiers in brightfield and fluorescence, depending on how much training data and time are available for training and optimization.

Curious to see HALO AI in action on your images? 

Advanced Phenotyping

Cell phenotyping can be used on any of the myriad supported image types to automatically assign cells into user defined phenotypes. Simply select a nuclear or membrane segmentation network, provide a few quick training examples, and train the network to identify phenotypes of interest.

Phenotyping is a highly effective and flexible tool to quantify cells based on morphology when biomarker information is absent and can assist with quantification and characterization of cells with complex morphologies, such as neuronal cells.

Top Notch Training and Support

At Indica Labs, we take pride in providing top-quality support. Unlimited IT support, scientific support, training, and access to our Learning Portal are included with every license at no additional cost. With support staff in the US, UK, EU, China, and Japan, you are supported in any time zone.

Wondering how easy developing your own algorithm can be without programming experience? Reach out to us to find out how intuitive HALO AI’s train-by-example interface really is.

What Our Customers Have to Say

Read independent reviews on our HALO AI Deep Learning Classifier Add-on and learn how these customers are using Indica Labs’ solutions to streamline their workflows.

Wonderful products, it is far beyond our expectations.

"Our lab both have the HALO Image Analysis System and the HALO AI, honestly speaking, the HALO system is beyond our expectations, especially the analysis process, annotation tools, the adjustable parameters, the texture and morphological recognition of nuclear and specific structure, and the after-sale care I strongly recommend you guys to choose it. it really worth the money."

lanlan Li


Great results, can't live without this instrument!

"Scientists can achieve lots of high-quality analysis data about figures by HALO software. Many functions of HALO are very helpful to study images, such as TMA, spatial analysis, high-Plex FL, and especially HALO AI. It is easy to operate. Its interface is simple, friendly, and convenient. Once we have problems, a Halo technician is able to help us to solve them in time."

Junfeng Hao

Institute of Biophysics, Chinese Academy of Sciences

Great result. Nice tool for translational pathology research. 

"We are interested in applying HALO AI in RNAscope and IHC analysis. Thanks to Yongtian ZHAO to give us wonderful training and support. We are improving and confident to apply it in the RNAscope assay development and scoring."

Fei Yang

Johnson and Johnson

Currently the best image viewer and analysis software on the market that I've used.

Nathan Aleynick


Platform Compatibility

HALO AI is compatible with all of the file formats that can be used in HALO and HALO Link. Yours not on the list? Email us your requirements.

File Formats:

  • Non-proprietary (JPG, TIF, OME.TIFF)
  • Nikon (ND2)
  • 3DHistech (MRXS)
  • Akoya (QPTIFF, component TIFF)
  • Olympus / Evident (VSI)
  • Hamamatsu (NDPI, NDPIS)
  • Aperio (SVS, AFI)
  • Zeiss (CZI)
  • Leica (SCN, LIF)
  • Ventana (BIF)
  • Philips (iSyntax, i2Syntax)
  • DICOM (DCM*)
    *whole-slide images

Advancing Discovery with AI: Features of HALO AI


Quickly acquire training annotations with the new AI annotation tool in HALO AI. Simply point and click your object of interest to add an annotation.

Use real-time tuning in HALO AI to watch a network as it trains in real time. Toggle the mark up on and off to evaluate performance, choose to add training data, or change parameters on-the-fly.

Once a HALO AI model is trained, a probability map can be used as an alternative output to a traditional mask to evaluate performance. Use real-time tuning to select an appropriate probability cut-off for a given class and view the output in a heatmap where blue represents the lowest probability and red the highest.

Investigate results with interactive markup images where you can toggle on and off each population of interest. Interactive markups can be combined with probability thresholding and are especially valuable in exploring validation outputs.

Deploy your HALO AI models in the context of HALO modules for maximum image analysis utility. Choose to utilize one or more models such as nuclear segmentation, membrane segmentation, tissue classification, cell or object phenotyping, and/or SlideQC.

Collaborate on the development of AI using the HALO Link collaborative image management platform. Simply invite your colleagues to a study and have them add training data.

Create classifier pipelines by connecting multiple HALO AI classifiers into a single workflow. Choose to run within HALO AI or within a HALO module.

Model interoperability leveraging an open format ONNX allows you to move networks between machines, bring in outside ONNX models, or export your model in the ONNX format thanks to the flexible import and export options in HALO AI.


Leverage the three-part Train – Validate – Test workflow to address the question: How good is my HALO AI classifier? Quantitative metrics are output based on the type of network.


Speed up your HALO AI analysis by adapting your trained network for your specific GPU. Also, you can queue training jobs based on number of iterations or time and set it to run at the end of your workday to maximize productivity and evaluate multiple model variants.

Free Image Analysis

See HALO AI in action on up to three of your images with a free analysis.

Knowledge Center

Learn more about HALO AI by exploring the tabs below. For an introduction to the capabilities of HALO AI, we recommend you check out our HALO AI white paper!

Imaging Mass Cytometry Application Note

See in this app note how automated analysis of highly multiplexed IMC images using HALO® and HALO AI yields rich cellular and spatial data from a streamlined workflow.

Glomeruli Quantification and Characterization Application Note

Don’t let varied stain selections hold back your digital image analysis of kidney sections! With HALO AI you can develop a tissue classifier that accurately detects and segments glomeruli in sections across a range of stains including H&E, trichrome, PAMS, and DAB with hematoxylin.

Quantify Single and Multiplexed RNAscope® Probes Across Whole Slide Tissues

Using our ISH module for brightfield and our FISH module for fluorescence, we demonstrate how to quantify ISH signal from RNAscope® assays on a per cell or per unit area basis and output an overall expression score based on ACD Bio’s recommended scoring guidelines.

HiPlex RNAscope™ Application Note

Using our FISH module we demonstrate how to perform 12-plex RNAscope image analysis using ACD’s HiPlexv2 assay.

Identifying Islets with HALO AI

Learn how HALO AI can detect islets of variable morphology in H&E-stained pancreatic tissue

Camelyon 17 Application Note​

CAMELYON17 challenged teams to develop automated methods for identifying and staging breast cancer metastases; read how the performance of a HALO AI classifier stood out among commercial entries!

NCI Case Study

Download this case study to learn about considerations the National Cancer Institute made when planning to migrate its HALO, HALO AI, and HALO Link deployments to the cloud, and benefits the NCI experienced after migration, including 50% faster HALO image analysis.

Cloud Services and AI Diagnostics Case Study

Learn how the optimized, scalable infrastructure provided by Indica Labs Cloud Services enables the rapid development and robust collaboration of our AI Diagnostics group, powering the next generation of automated AI-based decision support tools.

Lunaphore COMET Hyperplex IF and HALO® Image Analysis eBook

Our collaborative eBook with Lunaphore shows how HALO and HALO AI, alongside the Lunaphore COMET instrument, enable a high-throughput workflow for phenotypic and spatial analysis of cells in the tumor microenvironment.

Sequential Same Slide mIF and H&E eBook

Reveal more data on a single slide with this workflow for sequential same slide multiplex IF and H&E staining and imaging using HALO and HALO AI.

mIF Co-Registration and Cellular Analysis eBook

Learn how HALO offers a streamlined workflow for co-registration and cellular analysis of immunofluorescence images from cyclically stained and imaged slides.

Qualitative and Quantitative Evaluation of the TME eBook

Check out this ebook to see how AI-based nuclear segmentation and tissue classification can streamline cell phenotyping across tissue compartments and samples.

Lunaphore COMET Hyperplex IF and HALO® Image Analysis of the TME

Download our collaborative poster with Lunaphore to see how HALO and HALO AI, alongside the Lunaphore COMET instrument, enable a high throughput workflow for phenotypic and spatial analysis of cells in the tumor microenvironment.

Clustering Workflow for Advanced Cell Phenotyping of Hyperplex IF Images

Download our collaborative poster with Lunaphore to learn about a clustering workflow that uses unsupervised learning for COMET™ images.

Spatial Characterization of the TME with Cell DIVE Multiplexed Imaging and HALO Image Analysis

Check out our collaborative poster with Cell Signaling Technology and Leica Microsystems to learn about a combined workflow for spatial analysis of the TME.

Quantitative Image Analysis of a Combined FISH-IF Assay

Learn how to detect protein and RNA in a single assay using ACD’s Co-Detection workflow and HALO image analysis.

Characterizing the TME Using IMC and HALO Image Analysis

Download this poster to learn how HALO and HALO AI can be used to analyze imaging mass cytometry data.

Automated Tumor Budding Quantification in Colorectal Carcinoma H&E Images

Learn about quantification of tumor budding in colorectal carcinoma using HALO and HALO AI.

Spatial Multiplex Profiling of Immune Markers with the RNAscope Hiplex v2 Assay

Download this poster to learn about combining ACD’s HiPlex v2 in situ hybridization assay with HALO image analysis.

Evaluation of the TME by High Resolution 17-plex IF and HALO Image Analysis

Learn how to perform highly multiplexed IF and image analysis using the Orion instrument combined with HALO and HALO AI image analysis.

Our customers are making vital discoveries in oncology, neuroscience, and diabetes research.  Check out some of the research our customers are publishing using the HALO AI Deep Learning Classifier Add-on.

Prognostic Value of CD8+ Lymphocytes in Hepatocellular Carcinoma and Perineoplastic Parenchyma Assessed by Interface Density Profiles in Liver Resection Samples

This study examined CD8+ cell distribution in hepatocellular carcinoma (HCC) and peritumoral liver tissue to investigate the overall survival (OS) and recurrence-free survival (RFS). It is well understood that CD8+ lymphocytes are involved in both the anti-tumor response and in...

Learn More
Polycomb deficiency drives a FOXP2-high aggressive state targetable by epigenetic inhibitors

EZH2 is the catalytic component of Polycomb Repressive Complex 2 (PRC2) and performs trimethylation of histone H3 at lysine 27 (H3K27me3) to silence chromatin. PRC2 is known to play different roles in different cancers and inhibitors of PRC2 histone methyltransferase...

Learn More
Multimodal integration of radiology, pathology and genomics for prediction of response to PD-(L)1 blockade in patients with non-small cell lung cancer

Predictive biomarkers to determine which non small cell lung cancer (NSCLC) patients will respond to immunotherapy are currently lacking. In this Nature Cancer publication, the authors combine computed tomography imaging, PD-L1 immunohistochemistry (IHC), and genomic data to predict responses to...

Learn More

HALO AI High Impact Publications

Review this selection of high-impact publications for examples of HALO AI applications in fields ranging from metabolism to immuno-oncology and myology.

HALO AI White Paper

Check out this white paper for an in-depth introduction to the deep learning-based add-on to HALO, including available networks, common functions, and the “train – validate – test” Validation workflow.

HALO Link Pharma Services White Paper

Download our white paper to learn how Indica Labs’ Pharma Services collaborates with customers and delivers HALO and HALO AI image analysis using HALO Link.

Quantitative RNAscope™ Image Analysis Guide

From experimental design considerations to optimized setup of HALO image analysis parameters, our guide will help take your quantitative RNAscope™ image analysis to the next level.