COMPUTATION HUB FOR FASTER RESULTS

SERVER OR CLOUD

Eliminate computing limitations
Leverage underutilized resources
Get results faster

SCALABILITY

Run experiments and jobs in parallel
Burst through bottlenecks
Distribute computing resources and access them from anywhere

COST EFFICIENCY

Screen compounds more rapidly
Using imaging to better diagnose defects
Instantly share information produced from images

Scale your imaging Science to get results faster

VisionHub works for your image analysis experts. It schedules computations on appropriate hardware to maximize throughput and monitors progress. The Hub onboards and processes images already in storage or as they are made by your imaging systems. Spatially resolved results are viewable directly on raw datasets.

Ask for More!

icons-mockup-partner

arivis-VisionHub

A nucleus of computational activity and collaboration at scale. Vision Hub enables design and execution of large-scale experiments that produce results from images, whether datasets are already stored or are actively being produced. Mass quantities of Imagery are onboarded and processed by a virtual team of computational workers that come to life when you need them and smartly utilize processing cores in servers or workstations. When lots of images need to be processed, we approach the maximum speeds possible. Your researchers can make analysis pipelines in the Vision4D GUI and upload these to the Hub where they can be vetted and used by others. Also, your computing experts can leverage and parallelize their own innovative processing algorithms via convenient API’s.

Register and browse image collections, experiments, results, etc.
Manage access for users and groups
Import datasets into the arivis framework and archive originals
Create experiments that produce results at scale

Leverage computation onsite and offsite in servers or workstations
Scalable pricing model to support bursts of throughput
Implement quality checks into experiments for increased efficiency and validity

CREATIVITY
THROUGHPUT
RESULTS

Create tailored and optimized workflows that extract valuable information from your stores of images. Run at scale to produce results faster and spanning all levels of analysis. Examine and share the results or expose them to AI.

IMAGING SCIENCE WORKFLOW

Acquire Images

Scientific images have varying spatial dimensions (2D or 3D), timings (still or time-lapse), numbers of channels, and regions of interest (continuous tiles or disconnected views). It’s essential that ALL kinds of images at ALL size scales, tiny to massive, are securely stored and easily accessed for imaging science workflows. arivis applications can monitor image production and onboard images as they are created. Flexible importers enable custom structuring of image files inside the Scalable Image Storage (SIS) format.

Visualize

Inspection of images and results must occur at many times at many steps in imaging science workflows. Quality checks after acquisition and processing ensure that raw data are correct and ready for accurate analysis. At the end, direct interactive visualization of results on raw images provide an instant view of patterns in data. Viewers in arivis applications provide fast visual access to support immediate quality checks and discovery of results. Scalable Image Storage (SIS) files are always viewable, even during processing.

Process & Enhance

A host of algorithms improve and standardize background and noise characteristics of images or stitch together 3D regions of interest (for example, for whole organ imaging). Notably, new AI-based reconstruction algorithms can now make incredible image enhancements. arivis lets you quickly evaluate a multitude of processing and enhancement strategies to discover the optimal ones for your images. Processing is highly parallel and proceeds with speed and efficiency.

Extract Results

The process of producing spatially resolved objects on scientific images drives verified discoveries and conclusions. There are many different approaches to creating results objects whether they exist in a volume, over time, or represent a collection, or class, of associated objects. arivis applications enable you to leverage the right approach for your data and databases of objects for each image can contain useful computed features. Each image’s story is told in relevant objective measurements that can always be viewed on the raw or enhanced images.

Summarize

High quality images and sets of results must be distilled to provide snapshots of project progress and outcomes. Rather than directly converting results to external formats, it’s often more efficient to first summarize them in the framework used to view the images and create the results. This saves time and enables efficient feedback loops. Flexible charting in arivis affords custom summaries of measurements while high-res rendering gives information-rich views of each image or time-series. Users can export results (tables, surfaces, renderings, etc.) as they like.

Enable Meta Analysis

High-throughput and enterprise-level imaging science produce results across a huge number of images and experiments. Unfortunately, in many campuses and organizations these results are completely disconnected, which means powerful metanalyses are not possible or extremely costly. The arivis environment can be configured to expose spatially resolved results, summaries of results, and raw data contained by results objects to metanalysis algorithms - including AI.

ARCHITECTURE

The arivis VisionHub adds massively parallel computation to your image science. As your campuses produce and stockpile images, the Hub extracts value from the images - faster. Imaging experts leverage the latest image processing strategies and can connect relevant datasets to onsite or offsite computing servers.

Imaging systems - Revolutionary fluorescence, CT, EM, high-content, MRI, and histology systems digitize your experiments. Images are a heavy data burden but can contain incredible information. Traditionally, entire image files had to be transferred to view even just small parts of images. arivis applications take any images and convert them into Scalable Image Storage (SIS), enabling access to any part of any dataset, of any size, at any time, for visualization, computation, or examination of results.

Datasets - Scientific images are stored on computers, workstations, and servers throughout your organization. The machines often lack power for the heavy image processing required to produce results rapidly. We efficiently access files to enable experts to inspect data or design computational pipelines. Datasets from anywhere can be registered to the Hub to eliminate bottlenecks and data silos.

Experts - Your scientists have ideas about how to extract information from images. Currently, they may have to rely on special staff to help them connect datasets to computational servers and run jobs. The Hub makes it easier to submit jobs for computation and enables you to add computational capacity as you need it, whether onsite or in the cloud.

Processing Servers - Powerful computing is available on your campuses but may not be used to the fullest. When not used, servers take up space, consume energy, and depreciate, while information hides in image storages. Offsite resources can be leased for bursts of capacity.

The Hub collects processors into a computational resource and smartly divvies jobs to achieve maximum throughput and efficiency.

Architecture-VisionHub

Talk with an Expert

Scalability and Stability Based on Reliable Scientific Cloud Solution

Easy access of additional PC power
On premise server
Cloud / Web Any data size
Multi-dimensional data 2D, 3D, 4D,
Well plates
Experiments
Tiles

Small Scale Analysis on a Single Desktop for Single User

Segmentation via flexible Analysis Pipeline
Workflows for spheroids and organoids
Workflows to count, measure and track
Manual proofreading, semi-automatic analysis
Interactive work on statistics and 3D result representation

Large Scale Batch Analysis for Institutes and Core Facilities

Images are stored on a central location
Create pipeline for specific 3D segmentation task
Count labelled cells
Total volume of cells
Track cells in 3D
Measure intensity over time
Automated Segmentation via Analysis Pipeline
Batch processing of several Image files without moving data

Computing Infrastructure

arivis VisionHub is a highly scalable and customizable solution that can be adopted to a wide variety of existing data storage and processing infrastructure, from single department and core facility to campus-wide deployment and distributed data centers for multiple locations in pharma industry. However, there is a rising need for out-of-the-box solutions for image storage and processing particularly for small to mid-size businesses such as startups in biotech, CRO, or newly established academia spin-offs that require cost-efficient, modular solutions with minimal administrative hassles.

This is why arivis recommends the HIVE® modular server platform by ACQUIFER™ for rapid and efficient VisionHub deployment. Designed for the pragmatic, biology-focused scientist, the HIVE platform boosts productivity in microscopy facilities and screening labs. HIVE is a high speed centralized data repository that removes the need to constantly move or duplicate data sets in processing and analysis workflows. Its modular design integrates high speed processing, visualization, remote access, project management, flexibility, data security, scalability and ease of use in one unit.

It saves time, space, costs and administrator workload.

HIVE_r