"I’ve failed over and over and over again in my life. And that is why I succeed."
(Scroll to the bottom for the answer.)
Every four years, Olympic athletes compress years of training, setbacks, and incremental improvements into a few days on the world stage. For most, the path there is not a straight line. It’s missed marks, injuries, strategy changes, and constant refinement, repeated until the fundamentals hold under pressure.
My graduate and early career years were in protein crystallography, where failure is basically the default setting and success shows up on its own schedule. The job is simple to describe and stubborn to execute. You try a huge number of slightly different “recipes” to convince a protein to form a crystal. Most do nothing. Some make odd-looking plates or clear, birefringent chunks that gets your hopes up for about five seconds. And every once in a while, you get a tiny crystal that looks promising, then you iterate, wait, and hope it turns into a crystal that gives you beautiful data.
That loop, train-test-adjust-repeat, is shared across disciplines. We in science spend a lot of time earning small improvements that only become obvious in hindsight. The payoff is that our tools keep getting better at turning iteration into understanding. X-ray CT is a great example. It gives us a way to see and quantify internal structure without cutting parts open, so we can interrogate how things are built, how they change, and what to do next to make the next design, material, or process iteration better.
In this newsletter, we’ll present particle analysis results from IPSDK, a software package for 3D image analysis. We also introduce a new video highlighting our benchtop CT system, and give CT some tips. I hope you enjoy it.
- Angela
Packed granular media and filter beds show up everywhere, from industrial cartridges to engineered materials. The performance often hinges on details you cannot judge from the outside, including how uniformly particles are packed, where voids concentrate, and whether particle size and shape drift over time. CT makes these internal structures visible, but the real value comes when you can turn that visibility into repeatable, quantitative particle statistics and defensible distributions (not just a pretty rendering).
That’s where IPSDK Explorer comes in. It’s a mature platform that’s accumulated some impressive capabilities over the years, especially for moving from visualization to measurement. Below we show particle size analysis results generated in IPSDK. Here, particles were segmented using IPSDK’s AI tools, then a watershed transform was applied to separate touching particles before computing size statistics and distributions.
IPSDK Exploreris an image processing and analysis software package for 2D and 3D data. It uses parallelized algorithms to take advantage of modern computing hardware.
The software is designed to handle large datasets, including volumes that may exceed available system memory, by supporting workflows that process data without requiring the full dataset to be resident in RAM.
IPSDK Explorer includes tools for common analysis tasks such as denoising, automated segmentation, and morphological classification. These capabilities are used across applications, including porosity quantification, fiber characterization, and cultural heritage imaging, in academic, R&D, and industrial settings.
To support repeatable analysis, IPSDK Explorer provides plugins that guide users through step-by-step workflows. It also supports time-sequence data management for processing and visualizing dynamic 2D+t and 3D+t datasets.
The platform supports workflow automation to help standardize and scale analysis, including for high-resolution X-ray tomography datasets.
If you want a quick, practical look at the CT Lab HX, Ted walks through the workflow in a new 5-minute product video. It follows the full arc from setting the sample and running a scan to reconstruction and results review, so you get a feel for the day-to-day experience and the decisions that happen along the way.
The video also includes a brief look at automation, including the optional automatic sample changer for running multiple samples with less hands-on time.
To be efficient, inspired, and informed.
Here are a few practical considerations when mounting your samples for CT data collection.
1. Prevent sample motion
Sample stability is the top priority in CT. Any motion during acquisition shows up as blur or streaking in the reconstructed volume. Use rigid, repeatable constraints first (3D-printed sample holders, tubes, clamps), then add small foam wedges or shims to eliminate play. For long, high-resolution scans, check for creep up front by acquiring a projection, waiting 5 to 10 minutes, then re-acquiring and subtracting projection images. If the subtraction shows edge shifts, the mount is not stable enough and you should wait longer or resecure.
2. Choose mount materials intentionally
Select fixture materials that minimize artifacts and simplify downstream analysis. Keep low-Z, low-density materials close to the sample. Reliable options include foam, polystyrene, cardboard, and common plastics. When the mount is substantially less attenuating than the sample, it reduces beam-hardening risk and makes segmentation or surface determination faster and more robust.
3. Optimize geometry and orientation
Place the sample as close to the rotation axis as possible to minimize the required field of view (FOV) and maximize effective resolution for a given magnification. For larger parts, orient the longest dimension parallel to the rotation axis whenever possible to reduce the maximum X-ray path length through the sample. If the sample height exceeds your FOV, move to helical scanning or multi-scan stitching rather than orienting the sample with the thickest dimension parallel to the X-ray beam path.
4. Stabilize and protect hydrated or fragile samples.
For delicate or hydrated specimens, isolate the sample in thin plastic (film wrap, small tubes, or containers) to prevent dehydration-driven drift and handling damage. For biological samples, wrap in plastic and include a small amount of storage medium to maintain hydration during the scan. If the sample is especially fragile, embedding in a low-melting point agarose gel can provide uniform support while preserving hydration. Keep a small reservoir of medium in the container so the sample environment stays stable through the acquisition.
Real Scientists, Not Actors
A collection of priceless and embarrassing moments curated by Sam Robles.
Michael Jordan
Michael Jordan is a retired American basketball player and six-time NBA champion, best known for his time with the Chicago Bulls. His career is a reminder that elite performance is built through repetition, setbacks, and relentless iteration, long before the moments that end up in the spotlight.
"I’ve failed over and over and over again in my life. And that is why I succeed."
That's a wrap. Please let us know how we can help you learn more about X-ray CT. We love to hear from you!
.png?upscale=true&width=320&upscale=true&name=Untitled%20design%20(4).png)
