AI-Powered Image Analysis

AI-Powered Microscopy Image Analysis Software

Automated cell counting, colony counting, particle sizing, and image quantification — no coding required. ConductVision Image replaces manual workflows with AI segmentation, locked protocols, and publication-ready exports that work with any microscope.

ConductVision Image — Nuclei Detection Protocol
Nuclei Fluorescence Analysis
0cells detected
Analysis Running
0%
0+
Active Labs
0M+
Images Analyzed
0%
Detection Precision
0%
Below Legacy Pricing

Core capabilities

No macros. No scripting. No fragile spreadsheets.

Five capabilities that replace scattered manual workflows with one standardized, auditable platform.

SegmentMeasureThreshold

Standardize

Protocol Builder

Save fixed analysis workflows every team member runs identically. Lock parameters, prevent drift, and standardize methods across multi-site studies.

n=5

Detect

AI Segmentation

Automatic cell and feature identification without manual thresholding. Visual overlays show exactly what was detected, and why.

AUDIT TRAIL09:41min_prob0.4209:41max_flow0.9009:42scale_um0.28

Defend

Audit Trail

Complete transparency on every setting and adjustment. Reviewers see the exact pipeline that produced your data.

384 / 500

Scale

Batch Processing

Apply your locked protocol to thousands of images at once. High-throughput analysis without manual intervention or crashes.

CSVPNG.docx

Report

One-Click Reporting

Export annotated images, summary data, and a methods paragraph together. Publication-ready figures and tables in a single click.

See it in action

Automated cell tracking, start to finish

Watch ConductVision detect, segment, and track cells across a live imaging sequence — no thresholding by hand, no scripting.

Colony plate image prepared for ConductVision colony counting
ConductVision Colony Counting
CFU
counts
Batch
folders
Overlay
review

Featured application

ConductVision Colony Counting Software

A dedicated colony-counting workflow for agar plates: upload a plate image or image folder, set plate-aware gates, and export counts with colony locations, overlays, and distribution analysis. It is software for existing image files, not a separate device.

Typical workflow

Upload, gate, count

Use one agar plate image when you only need the final colony count or a count at a specific time point.

Time points

Compare existing image series

Analyze images collected at scheduled time points for growth curves without changing the counting workflow.

Batch mode

Plate folders and dilution series

Apply the same locked colony-detection settings across plate folders, dilution series, or repeated experiments.

AI output

Counts, locations, distribution

Export colony segmentation, total counts, colony coordinates, and spatial distribution summaries for one image or a folder.

Applications

19 built-in analysis workflows spanning life sciences and materials science, with full support for custom protocols.

Pathology and HistologyLife Science

Pathology and Histology

Slide QC, H&E nuclei, DAB quantification, morphology, and instrument workflow handoffs

Cell Count (Fluorescence)Life Science

Cell Count (Fluorescence)

Precise cell counting using fluorescent markers

Cell ConfluenceLife Science

Cell Confluence

Monitor culture coverage in real time

Nuclei (Fluorescence)Life Science

Nuclei (Fluorescence)

Fluorescence-based nuclei detection and quantification

Hematoxylin-Stained NucleiLife Science

Hematoxylin-Stained Nuclei

Detect nuclei in H&E histology sections

DAB Stained Tissue AnalysisLife Science

DAB Stained Tissue Analysis

Quantify DAB staining intensity in IHC samples

Live/Dead Cell ViabilityLife Science

Live/Dead Cell Viability

Dual-fluorescence viability assessment

Object Count (General)Life Science

Object Count (General)

Versatile object detection for any sample type

Colony Counting SoftwareLife Science

Colony Counting Software

AI colony counts from uploaded agar plate images

Percent Area (General)Life Science

Percent Area (General)

Coverage and area percentage calculations

Particle Size AnalysisMaterials Science

Particle Size Analysis

Image-based particle size distribution

Grain Size (ASTM E112)Materials Science

Grain Size (ASTM E112)

Standardized grain size per ASTM E112

Phase & Microstructure QuantificationMaterials Science

Phase & Microstructure Quantification

Measure area fractions of phases and structures

Defect Detection & CountingMaterials Science

Defect Detection & Counting

Automated defect identification and counting

Coating & Thin Film CoverageMaterials Science

Coating & Thin Film Coverage

Assess thickness and uniformity of coatings

Coating Thickness & PorosityMaterials Science

Coating Thickness & Porosity

Measure coating layers, black-area porosity, and Vickers indentation dimensions

Surface Coverage & PorosityMaterials Science

Surface Coverage & Porosity

Quantify porosity and surface coverage

Fiber & Inclusion AnalysisMaterials Science

Fiber & Inclusion Analysis

Measure fiber length, orientation, and density

Custom Inspection WorkflowsMaterials Science

Custom Inspection Workflows

Build repeatable routines for your own criteria

Microsphere AnalysisMedical Device & Pharma QC

Microsphere Analysis

Count, diameter & size distribution for drug-delivery and calibration microspheres

Validated Performance

99% precision. Validated on 23,000+ cells.

Tested against expert manual counting — the gold standard in cell biology.

96%
Overall Accuracy

Combined precision and detection rate vs. expert counts

99%
Precision

Virtually no false detections

93%
Detection Rate

Finds the vast majority of real cells

Cell counting validation — predicted instance segmentation, overlay, and original input
Representative output from the evaluation set. Left: predicted instance segmentation with unique colors per cell. Center: segmentation contours overlaid on the original image. Right: original fluorescence microscopy input. Evaluated on 200 images containing 23,615 manually annotated cells at IoU = 0.5.

Trust your counts

99% precision means nearly every detected cell is real. No phantom counts inflating your data or triggering unnecessary review cycles.

Miss less

93% detection rate across diverse cell types. The model finds cells that manual counting overlooks, especially in dense or overlapping regions.

Standardize results

Same algorithm, same parameters, every time. Eliminate inter-operator variability and produce counts that hold up across reviewers and publications.

Full Evaluation Data

IoU ThresholdTrue Pos.False Pos.False Neg.PrecisionRecallF1 Score
0.3 (lenient)21,9401291,67599.4%92.9%96.1%
0.5 (standard)21,8911781,72499.2%92.7%95.8%
0.7 (strict)21,5934762,02297.8%91.4%94.5%
0.9 (very strict)17,6524,4175,96380.0%74.7%77.3%

Live-Cell Tracking

Beyond counting: track every cell through time

Point ConductVision Image at a time-lapse video and it follows each cell frame by frame — reconstructing how cells move, divide, and disappear. You get trajectories, full lineage trees, and population dynamics, not just a single-frame count.

Tracking Result — trajectories & lineage

Cell trajectories

Division / lineage tree

Dividing cell Leaf (no division)Root → daughter over frames

Motion

Per-cell motion

  • Total path length
  • Net displacement
  • Directionality
  • Mean squared displacement (MSD)

Lineage

Division & lineage

  • Parent → daughter tree
  • Depth & descendants
  • Division events
  • Censored / dead / orphan flags

Temporal

Population over time

  • Average speed
  • Proliferation rate
  • Survival rate
  • Frame-by-frame curves

Population

Forest-level metrics

  • Tree & branching rate
  • Max / mean depth
  • Leaf fraction
  • Mean / max lifespan
Exports high-resolution:Cell trajectory chartsCell division / lineage treesSpatial activity heat mapsTemporal charts (speed, proliferation, survival)

Morphology

Sixteen shape and size measurements per object — automatically

Every segmented cell, nucleus, particle, or grain is measured for its full morphology in the same pass that counts it. Results report in pixels and convert to microns the moment you set a scale.

Used across cell biology, pathology, microbiology, materials science, and neuroscience — wherever object shape and texture carry the signal.

Core measurements

AreaContour areaPerimeterEquivalent diameterCircularityMajor axisMinor axisAspect ratioElongationOrientationSolidityConvexityCentroidMax FeretMin FeretLabel ID

Advanced descriptors

Hu MomentsZernike MomentsShape descriptorsGLCM textureNearest-neighbor distance

Particle & Microsphere Sizing

Full size distributions, not just counts

Microsphere and particle protocols return the distribution statistics that QC and formulation teams actually report — percentiles, spread, and a fitted normal curve over the diameter histogram.

D10 / D50 / D90
Size percentiles
Mean ± SD
Diameter distribution
Median
Central tendency
Min / Max
Full range

Scientific rigor

Built for numbers you can defend

The engine and controls behind every count, measurement, and distribution.

Flow-field AI segmentation

Every pixel gets a cell-probability and a flow vector toward its cell center, so touching and overlapping objects separate cleanly — no manual thresholding.

MicroscopyPlateMicrosphereAI MODEL

Selectable AI models

Match the model to your image. Choose the microscopy, plate, or microsphere image type and the system routes to the AI model tuned for it.

max flowmin probmin size128cells

Interactive recount

Adjust max flow, minimum cell probability, or minimum size and re-run the count on the spot — tune to your ground truth without reprocessing from scratch.

1 px = 0.28 µmarea · diameter · Feret → µm

Pixel-to-micron calibration

Define the pixel-to-micron ratio once and every measurement — area, diameter, Feret, distances — converts to real-world units automatically.

OVERLAYORIGINAL

Synced source & result review

Compare the segmentation overlay against the original image side by side at matched zoom and pan, so you verify every detection before it reaches your data.

denoise · normalize · neurite maps

AI preprocessing

Microscope-agnostic denoising and normalization, plus neurite probability maps and multi-scale handling for thin structures like axons, run before segmentation.

How It Works

From raw images to publication in four steps

Step 01

Define your protocol once

Set segmentation parameters, thresholds, and measurement rules. Lock the protocol so every analyst runs the identical pipeline.

Parameter lockingTeam sharingVersion control
Define your protocol once

Step 02

AI detects, you verify

Automated segmentation identifies cells, particles, defects, or regions of interest. Visual overlays let you confirm what was found.

AI segmentationVisual overlaysConfidence scores
AI detects, you verify

Step 03

Batch hundreds of images

Apply your locked protocol to entire image sets in one click. Process overnight runs, multi-well plates, or time-lapse series without touching a single parameter.

High throughputQueue managementProgress tracking
Batch hundreds of images

Step 04

Export publication-ready data

Annotated images, summary statistics, and method descriptions export together. Drop directly into your manuscript, or pipe CSVs into R, Python, or Prism.

CSV / ExcelAnnotated imagesMethods text
Export publication-ready data

Outgrew ImageJ?

If you're spending hours writing macros, fighting plugin conflicts, or getting inconsistent results across observers — ConductVision eliminates all of that.

  • No codingreproducible analysis protocols with point-and-click setup
  • No plugins17 analysis modules built in (cell count, colony count, confluence, viability, particle size, grain size, and more)
  • No observer biasAI segmentation delivers consistent results every time, across every operator
  • Commercial supportreal humans who understand microscopy, not Stack Overflow threads

ImageJ → ConductVision

Setup time per analysisHoursMinutes
Inter-observer variability15-30%<2%
Batch capacityManual5,000+ images
Protocol reproducibilityFragileLocked
Audit trailNoneComplete

How it compares

ConductVision vs. the Alternatives

How ConductVision Image stacks up against the free and commercial microscopy tools most labs already run.

Feature RecommendedConductVision ImageImageJ / FijiImage-Pro (MediaCybernetics)
AI segmentationBuilt-inPlugin required(v11+)
Coding requiredNoYes (macros / scripts)No
Works with any microscope
Batch processing5,000+ imagesManual scripting
ASTM E112 complianceBuilt-in
Locked protocolsVersion-controlledNo lockingPartial
Audit trailFullPartial
Commercial supportDedicated teamCommunity only
PricingRequest quoteFree / open source~$2K–$5.5K/year

Reproducibility

Standardize Your Science

Define your method once. Lock it forever. Ensure every researcher measures with the exact same ruler, regardless of who runs the experiment.

Create locked standards for everyday tasks

Cell and nuclei countingArea and coverage measurementsSignal intensity and ratio protocolsColony counting standardsParticle analysisSize and shape distribution
Explore Example Protocols

Below legacy pricing. Above legacy capability.

All plans include 5 seats, unlimited analysis, and automatic updates. No per-image fees.

1-Year Plan

5 seats included

$1,750/year
  • Unlimited image analysis
  • All 17 built-in protocols
  • Custom protocol builder
  • Batch processing
  • Standard support
  • Automatic updates
Select Plan

3-Year Plan

$4,950 total · Save 6%

$1,650/year
  • Everything in 1-Year
  • Priority support
  • Early access to new protocols
  • Dedicated onboarding call
  • Price lock for 3 years
  • Multi-lab discount eligible
Select Plan

5-Year Plan

$6,250 total · Save 29%

$1,250/year
  • Everything in 3-Year
  • Custom protocol development
  • Annual training session
  • Price lock for 5 years
  • Lowest per-seat cost
  • Grant budget friendly
Select Plan

Enterprise & Core Facilities

On-premise deployment, unlimited users, LIMS integration, and a dedicated account manager.

Contact Sales

Deployment

Your data, your rules

Both options use the same interface and produce identical, standardized outputs.

Cloud

Browser-based access with automatic updates. No IT setup required — your team is analyzing images in minutes.

  • Zero installation
  • Automatic updates & new protocols
  • Access from any device
  • SOC 2 compliant infrastructure
  • 99.9% uptime SLA

On-Premise

Local server or private cloud for institutions with strict data residency or air-gap requirements.

  • Data never leaves your network
  • HIPAA & GxP compatible
  • Active Directory / LDAP integration
  • Admin control panel
  • Dedicated support engineer

Common questions

Ready to standardize your image analysis?

Request a demo and see ConductVision Image running on your own images within the week.