ANTSAPPLYTRANSFORMSTOPOINTS(1)

🦋 ANTSAPPLYTRANSFORMSTOPOINTS(1) 🦋

Okay bestie, this is the manual for antsApplyTransformsToPoints 💖 It lives in the User Commands section, and it's basically the tool for moving points around with transforms. Let's get into it 🫶

✨ NAME ✨

antsApplyTransformsToPoints - part of ANTS registration suite 🌈

💅 DESCRIPTION 💅

🎀 COMMAND 🎀

antsApplyTransformsToPoints

okay so antsApplyTransformsToPoints takes an input image and transforms it using a reference image and a transform (or a whole set of transforms) 🔥 It reads in a CSV file where the first D columns define the spatial location, and that location is in physical coordinates. Your CSV needs a header row, no cap 📄 Here's an example, iykyk 🧚

cat chicken-3.csv

x,y,z,t,label,comment
82.5,116.5,0,0,1,this is the breast
137.5,35.5,0,0,2,this is the beak

antsApplyTransformsToPoints -d 2 -i chicken-3.csv -o test.csv -t [chicken3to4.mat ,1 ]

cat test.csv

x,y,z,t,label,comment
10.8945447481644,162.082675013049,0,0,1,nan
7.5367085472988,52.099713111629,0,0,2,nan

💀 the nan shows up in the last column until the ITK CSV I/O can handle mixed numeric / string types. ngl that's the only tea here. if your input is fully numeric, all is well 🫶

We now can also read / write .mha files 🦋 It's a simple binary format (Meta format, go look it up, it's giving research energy 🤓) that is way faster to read/write than CSV format. Just a heads up: to write an .mha file, you must also pass an .mha file as input 😭

🎀 OPTIONS 🎀

• -d, --dimensionality 2/3 — this one forces the points to be treated as whatever dimensionality you specify 👑
• -p, --precision 0/1 — turns on double precision, 0/1 💜
• -f, --forantsr 0/1 — set it to true for ANTsR IO 🌸
• -i, --input inputFileName — the only input supported right now is a CSV file with columns including x,y,z,t (all 4) as column headers, bestie 📄 If you don't have 4D data, you still supply 4D and fill the extra spots with zero. The points should be defined in physical space. Now, the main tea ☕ Points are transformed in the OPPOSITE direction of images, so you should pass the inverse of what you'd need to warp the images. Like, if the image is warped by Affine.mat, you pass the inverse of Affine.mat to transform points defined in the same space as the image 🙈 If you're confused about converting coordinates from your files to the space antsApplyTransformsToPoints needs, try creating or drawing a simple label volume with only one voxel set to 1 and everything else set to 0. Write down the voxel coordinates. Then use ImageMaths LabelStats to find out what coordinates this voxel is expecting from antsApplyTransformsToPoints ✨ ITK uses an LPS coordinate system, so keep that in your back pocket 🧠 See http://sourceforge.net/p/advants/discussion/840261/thread/2a1e9307/ for more. ***Or you can pass in a 2D mha (meta format) binary image file 🦋
• -o, --output warpedOutputFileName — you can output the warped points to a CSV file 🎀
• -t, --transform transformFileName [transformFileName,useInverse] — okay this one is a whole vibe 🌈 Several transform options are supported, including all the ones defined in the ITK library, plus a deformation field transform. The ordering of the transformations follows the ordering you specify on the command line. An identity transform gets pushed onto the transformation stack first. Each new transform you hit on the command line also gets pushed onto that stack. Then, to warp the input object, each point in it gets warped by the last transform pushed, then the second-to-last, and so on, until you reach the last transform encountered, which is the identity transform 💅 Also, the inverse transform can be accommodated, with the usual caveat that such an inverse must be defined by the specified transform class ⚠️
• -h — prints the help menu (short version) 🌸 <VALUES>: 0
• --help — prints the help menu 🧚 <VALUES>: 1, 0

---

Okay that's the whole tea, bestie 💖 Tap in and go have fun with antsApplyTransformsToPoints ✨

antsApplyTransformsToPoints 2..., December 2025, ANTSAPPLYTRANSFORMSTOPOINTS(1) 🦋💜