top of page
Aivia 8.5
Boost your productivity with AI
AI image restoration
AI deconvolution

Acquire images faster with less light exposure and let Aivia restore your data

Predictive neuron reconstruction

Boost your productivity using innovative and intelligent tracing tools

Teravoxel rendering and analysis

Interactively render teravoxel datasets and efficiently analyze multiple 3D ROIs

Live recipe apply

Import images and apply analysis at the same time while the data is being acquired

AI deconvolution

Enhanced by artificial intelligence

  • No need to measure or model the PSF
  • Minimizes phototoxicity and photo bleaching
  • Enables imaging at higher temporal resolution
  • Works for a wide range of light microscopy modalities
  • Easy to train and apply
  • Greatly improves image quality and signal to noise
  • Restoration quality on par with state of the art classical PSF deconvolution
  • Update and customize model with your own data
  • Works in Aivia and Aivia Cloud

The quality of images created by diffraction-limited light microscopy modalities (e.g. most widefield, confocal and light sheet) is deteriorated by the point spread function (PSF) of the system. Using the PSF, classical deconvolution methods, such as the Richardson–Lucy algorithm, can be applied to computationally restore images. The PSF can be obtained by 1) direct measurement from the images or 2) estimation. Neither solution is ideal as it requires either a laborious manual measuring process or expert level knowledge of many hardware components that affect the modeling function.


With deep learning (DL), the PSF is directly learned from easy to generate training images - no need to measure it or deterministically model it. Training of a DL model is an iterative process which can be efficiently done by using multiple GPUs, e.g. via Aivia Cloud. Once created, it can be used on new images in a single step. By contrast, with classical deconvolution the restoration process is iterative for every image one desires to process.

Aivia 8.5 expands our suite of AI-powered image processing models with deep learning deconvolution (AI deconvolution) for a wide range of light microscopy systems (e.g. widefield, confocal and light sheet). The models can be applied to 2D and 3D datasets to drastically improve their quality versus the originally acquired data. The AI deconvolution solution in Aivia 8.5 effectively increases the photon budget (important when phototoxicity is a concern), allows you to increase the temporal resolution or can reduce your total imaging time (important when photobleaching and / or hardware availability is limited).

The AI deconvolution models work both in Aivia and Aivia Cloud. You can create and / or augment the applicability of Aivia’s AI deconvolution models with your own images. No AI or coding knowledge required.

AI deconvolution output (drag slider to reveal)
Compare classical and AI deconvolution
Classical deconvolution
AI deconvolution
Need to get PSF

(Direct measurement or estimation)

PSF learned from data

Restoration is iterative

Requires training

(Training is iterative)

GPU acceleration

Possible, depending on software

Restoration quality improves with amount of data generated and processed

Efforts needed to reduce restoration-induced artifacts


(Adjust hardware or

algorithm, remeasure PSF)


(Train with more data)

Number of image pairs needed for training


5 to 10 pairs of images

(512 x 512 x 50)

Predictive neuron reconstruction

Predictive neuron reconstruction

Smart, interactive tracing

Contemporary sample preparation techniques coupled with state of the art imaging modalities make it possible to routinely image whole brain samples at cellular resolution. However, interactively displaying hundreds of GBs of data and efficiently reconstructing complex multi cellular neuron networks is a real challenge.


Aivia already offered fully automated and manual neuron analysis options, but we now have added two new predictive neuron reconstruction modes to the Neuron Composer - Aivia's neuron reconstructing tool set. Point-to-Point: you define the start and the end point, Aivia automatically creates a 3D trace between them. Segment Prediction: you define a start point and Aivia offers one or more predicted dendrite segments which you can validate with a single click.

  • Fast and accurate neuron reconstructions
  • Trace in large, thick, cell-dense samples
  • Efficiently reconstruct single neurons and large neuronal networks
  • Two predictive tracing modes
  • Reconstruct and analyze soma, dendrites and spines

The new tracing modes also benefit from adaptive options to intelligently hide parts of the raw data so you can stay focused on the region you are interested in tracing - great for efficiently tracing large and dense networks. To further boost your productivity dendrite connections are created automatically as you/Aivia creates them and the field of view will seamlessly follow your actions. Aivia's Neuron Composer sets a new standard for neuron reconstruction in large samples.

All neurons, soma, dendrites and spines created in Aivia can be used to create machine learning powered classification and phenotyping classifiers using Aivia's Object Classification.   

Media gallery

Teravoxel rendering and analysis

Teravoxel rendering and analysis

Bigger and better

Interactive rendering of teravoxel datasets is now a reality thanks to a new round of optimizations to Aivia's innovative volume rendering engine in Aivia 8.5. The latest developments make it possible to display huge 3D volumes (e.g. whole mouse organs or multi-mm sized prostate biopsies) within seconds of loading an Aivia TIFF file.


These super-large datasets can be time-consuming and computationally intensive to analyze in full. Region of interest (ROI) processing in Aivia 8.5 lets you apply on a sub-region of the image. You can define a cube-shaped ROI directly on the image or use a previously created mesh of arbitrary shape to define the region to apply a recipe to. Aivia applies the analysis to the ROIs only - a major time saver for large datasets. Additionally, you can apply the same recipe settings to multiple ROIs at the same time. Preview has been extended to 3D ROIs which enables you to preview your results in 3D.

  • Optimized volume rendering pipeline with support for up to 512 image blocks per image
  • Render teravoxel images of whole organs in seconds
  • Apply recipe to a small region of interest (ROI) within the image
  • Define an ROI using on-screen tools or import a mesh of any shape
  • Create multiple ROIs and apply the recipe to all at the same time
  • Preview results in 3D ROIs
Media gallery

Live recipe apply

Live recipe apply

Acquire and analyze simultaneously

It is possible to acquire 2D+time image sequences and analyze them at the same time with live recipe apply. Aivia 8.5 extends the functionality of Live Import (introduced in 2018 as part of Aivia 7.5) to enable applying Aivia's image analysis recipes to images as they are being acquired in real-time.

Start by defining the recipe parameters. Launch Live Import in Aivia to choose the acquisition folder and input image settings. Aivia polls the acquisition folder periodically for new files added to the folder and appends the image to the current image sequence. The recipe is applied to the newly acquired image at the same time. A real time saver!

  • Get analysis results while your experiment is ongoing
  • Aivia imports and analyzes your images as they are being acquired.
  • No image conversion needed - all images are automatically converted to Aivia TIFF

More features and improvements

  • 2D image stitching

  • Volume transformer - re-align 3D volumetric data

  • User defined GUI layout options

More improvements
Contact us

Please fill in your contact details and tell us more about your research and image analysis requirements

Thank you! We will contact you shortly.

Tel: 1-855-423-5577

15405 SE 37th St, Ste 100

Bellevue, WA 98006

bottom of page