Image analysis, from raw files to quantified results
Turn microscopy into numbers you can trust.
Upload your microscopy and answer a couple of questions about what you are measuring. Conspecta assembles the pipeline from your answers, re-runs it live on one image as you change a setting, and runs it across the whole set: objects detected and counted, area coverage and marker intensity measured off the same run, all in the browser.
Every result stays attached to the image it came from and the sample behind it, so a count always points back to what you measured. Multi-channel, z-stack, and time-series viewers read LIF, TIFF, PNG, JPEG, and BMP, and any result can flow straight into a figure.
Answer a few questions, get a pipeline you can read
Say what your images show and what you are measuring, and the builder assembles the pipeline from your answers. Each step reads as an action, from evening out the background to finding objects by a brightness cutoff, and opening one shows settings named the same way: measure bright areas or dark, threshold automatic or fixed. Change a setting and the whole pipeline re-runs on the image in front of you, so every adjustment shows its result on screen before you commit the set to it.
Run only the slices that are in focus
A z-stack is counted by slice, so a 40-slice stack uses 40. Conspecta scores every plane's sharpness on your own machine and offers the band worth running, such as slices 5 to 17 of 40. Accept it, widen it, use every slice, or set one range across every stack in the set.
Cut the debris out of a count in one drag
A segmenter counts the dust and the debris along with the cells. A band under the image narrows what counts by size, confidence, brightness, or color, drawn as a histogram of the objects themselves with cut-outs along the axis so the number you land on has a picture behind it. The size cluster most objects share is highlighted, with a floor set under the specks below it. Filtering reads results you already have, so it lands instantly and costs no usage.
Every run writes its own data table
A finished run charts its count per image and writes the same numbers into a data table named after it, one row per image, the moment it finishes. The Chart view and the figure picker both read from that table, so a result is one click from a chart and two from a figure, and the number under a published panel still traces back to the field it was measured in.
More in the Image analysis workspace
Files and formats
TIFF, PNG, JPEG, BMP, and LIF
Up to 200 files at a time, 100 MB an image and 5 GB a LIF. Every series inside a LIF arrives as its own image with z-slices, channels, and timepoints intact.
Label images from their filenames
Fill the experiment design grid from a pattern such as {timepoint}_{condition}_r{replicate}. The line under the field counts how many of your filenames it fits and spells out what it reads from one of them.
Viewing and markup
The image viewer
Zoom past 4x to real pixels, pan, and step through a collection with the arrow keys. Details gives dimensions in pixels and, where the file carries its calibration, in micrometers.
Enhance and color LUTs
Presets for weak signal, publication, and high contrast, sliders from brightness to deconvolution, and seven lookup tables including colorblind-safe Viridis. Your stored file stays exactly as it was.
Channels, projections, and 3D
Merge or split fluorescence channels with per-channel gain, contrast, and flattening, project a stack by maximum, average, or sum, or turn the whole volume in 3D View.
Timelapse playback
Play, pause, and step a time series from 0.25x to 8x, with the current frame in a box you can type into and a loop back to the start.
Annotate and crop
Brush, rectangle, ellipse, polygon, arrow, and a scale bar, plus text labels and a click counter for hand-counting a field. Marks sit over the image, and undo walks back through every change you made this session.
Present a collection
A full-screen slideshow on a 2 to 30 second interval, region screenshots, and a download of the image as you have it set up, enhancements and all.
Results and sharing
Compare runs side by side
Put two or more runs in one data table grouped by run, condition, timepoint, or replicate. The preview gives each group's image count, mean, and standard deviation before anything is created.
Linked to samples and figures
Results stay linked to their images, the samples behind them, and the figures they end up in, so a number in a paper traces back to the field it came from.
Review and sign-off
Draft through Approved or Archived, with comments and sign-off. Collections and image analysis are each private to you or shared with the project, and changeable afterwards.
Learn more: Analyze images·Image tools·vs. ImageJ·vs. CellProfiler·vs. Imaris
Bring your first images
Drop in a LIF or TIFF and watch the count come back, linked to your samples. Free for you and one collaborator.