reading of PETSIRD list-mode function by KrisThielemans · Pull Request #1604 · UCL/STIR

KrisThielemans · 2025-07-16T13:22:37Z

Adding support for PETSIRD.

Warning: this changes SAFIR reading (although we think it should be safe), as we lifted a lot of its code to a new class CListModeDataBasedOnCoordinateMap

To be able to test with existing PETSIRD files, @NikEfth also added support for even number of TOF bins. So...
Fixes #1333

NikEfth · 2025-12-10T20:17:37Z

Made some progress today:

Automatically detect BlocksOnCylindrical geometry and reconstruct with OSEM25 (1 full iteration):

NikEfth · 2025-12-17T17:30:06Z

PETSIRD requires additional dependencies (yardl, pinned xtensor) that are not part of STIR’s standard CI environment.
The following PR therefore adds an optional, isolated CI workflow that validates PETSIRD builds without impacting existing CI jobs.

NikEfth · 2025-12-17T17:34:55Z

There is a problem with the TOF tests. Very likely, this happened because in this PR I allowed the use of an even number of TOF bins (because my PETSIRD files had an even number). I may have overlooked something regarding the odd number (default in the tests).

Copilot

Pull request overview

This PR introduces initial support for reading PETSIRD (PET Standard for Image Reconstruction and Data) list-mode files into STIR. The work includes scanner geometry inference, coordinate mapping, and the necessary infrastructure to handle PETSIRD data formats (both HDF5 and binary).

Adds PETSIRD submodule integration with CMake configuration
Implements PETSIRD list-mode data reader with automatic scanner geometry detection
Refactors SAFIR list-mode classes to share common base functionality
Includes TOF handling improvements for even-numbered TOF bins

Reviewed changes

Copilot reviewed 35 out of 36 changed files in this pull request and generated 16 comments.

Show a summary per file

File	Description
CMakeLists.txt	Adds PETSIRD as optional dependency with submodule management
src/buildblock/PETSIRDInfo.cxx	Core geometry inference logic for PETSIRD scanners
src/listmode_buildblock/CListModeDataPETSIRD.cxx	Main PETSIRD list-mode reader implementation
src/include/stir/listmode/CListModeDataBasedOnCoordinateMap.h	New base class for coordinate-mapped list-mode data
src/buildblock/ProjDataInfo.cxx	Refactored TOF mash factor logic with even bin support
src/include/stir/ProjDataInfoCylindricalNoArcCorr.inl	Fixed TOF bin mapping for even TOF bins
src/IO/PETSIRDCListmodeInputFileFormat.cxx	File format detection for PETSIRD files
.github/workflows/build-test.yml	CI updates to disable PETSIRD in existing tests

Comments suppressed due to low confidence (4)

src/recon_buildblock/find_basic_vs_nums_in_subset.cxx:1

The commented-out loop and loose braces suggest incomplete refactoring. Either remove the commented code or complete the intended changes to maintain code clarity.
src/recon_buildblock/find_basic_vs_nums_in_subset.cxx:1
Debug print statements should be removed before merging. These commented-out std::cout calls are leftover debugging code that clutters the source.
src/include/stir/listmode/CListModeDataBasedOnCoordinateMap.h:1
File header comment references wrong class. This header is for CListModeDataBasedOnCoordinateMap, not CListModeDataSAFIR.

/* CListModeDataSAFIR.h

CMakeLists.txt:1

All optional dependencies have been changed from OFF to ON (disabled by default). This appears to be a temporary development change that should not be merged. These defaults would break existing functionality for other users.

# This file is part of STIR.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-12-19T07:22:33Z

@NikEfth I've opened a new pull request, #1671, to work on those changes. Once the pull request is ready, I'll request review from you.

Copilot · 2025-12-19T08:38:48Z

@NikEfth I've opened a new pull request, #1672, to work on those changes. Once the pull request is ready, I'll request review from you.

NikEfth · 2025-12-20T14:32:05Z

Introduce PETSIRDInfo as a dataset-level integration object for PETSIRD
listmode support and refactor PETSIRD listmode classes to align with
STIR’s CListRecord / CListEvent design.

PETSIRDInfo encapsulates all PETSIRD-specific static information needed
by STIR, including:

PETSIRD ScannerInformation and optional Header metadata
construction of the corresponding STIR Scanner geometry
mapping from PETSIRD ExpandedDetectionBins to STIR DetectionPositions
optional forced geometry selection (cylindrical vs block)

PETSIRDInfo is immutable after construction and shared via
shared_ptr by all PETSIRD-related components
(CListModeDataPETSIRD, CListEventPETSIRD, normalisation, etc.), providing
a single source of truth for PETSIRD → STIR interpretation.

The PETSIRD listmode implementation was updated accordingly:

CListModeDataPETSIRD now owns PETSIRDInfo and exposes it read-only
CListRecordPETSIRD remains a storage container and embeds
CListEventPETSIRD, following existing STIR patterns
CListEventPETSIRD stores raw PETSIRD coincidence data and performs
PETSIRD expansion and geometry mapping on demand using PETSIRDInfo
PETSIRD expansion and STIR bin mapping are kept out of the record layer

These changes preserve STIR’s separation of concerns, avoid PETSIRD
logic leaking into core reconstruction code, and allow PETSIRD geometry
and metadata to be reused consistently by listmode, IO, and
normalisation components.

KrisThielemans

This is mostly clean. thanks! A lot of my comments are housekeeping, but there are some important ones. Major thing is to keep this PR separate from other things (I won't accept this PR with them...), for example:

there are some changes here for set_proj_data_info which probably came from another PR. Maybe they'll disappear when you merge master here
there are changes here to cope with even number of TOF bins. They might be good, but it doesn't have anything to do with PETSIRD.
there's some changes in find_basic_vs_nums... which don't belong here
I see no reason to make InputFormat::can_read non-const. If this is really required for PETSIRD, please explain. We should then work around it.

I guess the BinNormalisation isn't finished yet. I would suggest we currently call error and do it in a separate PR.

KrisThielemans · 2026-03-04T11:00:36Z

You van have mutable members Kris

________________________________ From: Nikos Efthimiou ***@***.***> Sent: Wednesday, March 4, 2026 10:02:46 AM To: UCL/STIR ***@***.***> Cc: Kris Thielemans ***@***.***>; Author ***@***.***> Subject: Re: [UCL/STIR] reading of PETSIRD list-mode function (PR #1604) ⚠ Caution: External sender @NikEfth commented on this pull request.

________________________________ In src/include/stir/IO/InputFileFormat.h<#1604 (comment)>:

@@ -46,7 +46,7 @@ class InputFileFormat

{ return this->actual_can_read(signature, input); } - virtual bool can_read(const FileSignature& signature, const std::string& filename) const + virtual bool can_read(const FileSignature& signature, const std::string& filename) This was necessary to support both the HDF5 and binary versions of PETSIRD. I could not think another way arround it. — Reply to this email directly, view it on GitHub<#1604 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABQRIHMFCYAB5WGTRVX5QVT4O75MNAVCNFSM6AAAAACBU47CHKVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTQOBYGM4DKOJQHE>. You are receiving this because you authored the thread.Message ID: ***@***.***>

NikEfth

[x] there are some changes here for set_proj_data_info which probably came from another PR. Maybe they'll disappear when you merge master here

Merged with master is fine now.

[ x] there are changes here to cope with even number of TOF bins. They might be good, but it doesn't have anything to do with PETSIRD.

I know. But I had to make them as the data shared by people in the hackathon had even number of bins. If you you want to compare with that we need these changes here. I can document more.

[x] there's some changes in find_basic_vs_nums... which don't belong here

Yes that code doesn't do anything atm.

[x] I see no reason to make InputFormat::can_read non-const. If this is really required for PETSIRD, please explain. We should then work around it.

Done

[x] I guess the BinNormalisation isn't finished yet. I would suggest we currently call error and do it in a separate PR.

No it is finished. Unless you have comments.

NikEfth · 2026-03-20T10:22:04Z

+  - Mat3: std::array<std::array<float,3>,3> for compact fixed-size storage.
+  - Vec3: std::array<float,3> for simple 3D vectors.


OK. I started a bit of this. From the test_ . It won't be hard.

KrisThielemans

some other clean-up suggestions

NikEfth

all done

NikEfth · 2026-03-20T17:22:41Z

Do you want the PETSIRDInfo_helpers with stir::Arrays?

codacy-production · 2026-05-22T09:48:48Z

Not up to standards ⛔

🔴 Issues 1 minor

Alerts:
⚠ 1 issue (≤ 0 issues of at least minor severity)

Results:
1 new issue

Category Results

BestPractice 1 minor

View in Codacy

🟢 Metrics 225 complexity · 8 duplication

Metric Results

Complexity 225

Duplication 8

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer}
_{TIP This summary will be updated as you push new changes.}

NikEfth · 2026-05-28T16:01:04Z

@casperdcl This is the PR

KrisThielemans

Looks like CI is working! Great. (The MacOS job fails because of too much stuff in .h/.inl, see previous review).

NikEfth · 2026-06-03T15:27:12Z

Hi, I wanted to replace the std::arrays with STIR Arrays but I don't have the time to do it now. I will come back to this later, but you can merge if you are happy with the rest.

NikEfth · 2026-06-17T09:39:15Z

I tried the interactive rebase, but it introduces a lot of old code and mege conflicts

KrisThielemans · 2026-06-17T09:40:03Z

please don't push that!

NikEfth · 2026-06-17T09:44:50Z

which one? The interactive rebase? Sure :) any other advice?

KrisThielemans · 2026-06-17T09:48:09Z

The trick would be to interactive rebase on the commit that we branched off from (not on master).

jonc125 · 2026-06-17T15:15:34Z

I want to flag for you that the macOS builds here are (since 15 May) consuming the free Actions budget we have available in the UCL organisation, because even though this is a public repo the use of larger runners is always charged for. The budget for June has now been used up. Could you look into doing fewer Mac builds or reducing the build time?

jonc125 · 2026-06-18T12:05:10Z

I want to flag for you that the macOS builds here are (since 15 May) consuming the free Actions budget we have available in the UCL organisation, because even though this is a public repo the use of larger runners is always charged for. The budget for June has now been used up. Could you look into doing fewer Mac builds or reducing the build time?

It appears this was a false alarm: despite GitHub's web UI reporting this repository as the biggest consumer of our actions budget, actually the runners used are free and don't count towards the limit. It was actually a couple of other private repositories that don't even make the top 10 overall users that were consuming the limited budget! But you have to look at the raw data rather than any of the dashboards to be able to tell that...

KrisThielemans · 2026-06-18T13:03:05Z

Thanks @jonc125!

In any case, there's a good message here that we should be careful in pushing to Github. Small commits use computing resources that should be used elsewhere.

codacy-production · 2026-06-19T08:42:36Z

Not up to standards ⛔

🔴 Issues 1 minor

Alerts:
⚠ 1 issue (≤ 0 issues of at least minor severity)

Results:
1 new issue

Category Results

BestPractice 1 minor

View in Codacy

🟢 Metrics 232 complexity · 8 duplication

Metric Results

Complexity 232

Duplication 8

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer}
_{TIP This summary will be updated as you push new changes.}

NikEfth · 2026-06-19T09:43:22Z

@KrisThielemans I don't think I can clean up the history more. At 220bec9e4b242741a668d726707036d18829cec6 I merged with the master and every time I try to rebase beyond that point I get many conficts.
Any advice?

mainly a copy at this stage

KrisThielemans · 2026-06-27T21:15:14Z

-      if (num_tof_bins % 2 == 0)
-        error("ProjDataInfo: Number of TOF bins should be an odd number. Abort.");
+  const int num_mashed_bins = max_timing_poss / tof_mash_factor;
+  num_tof_bins = num_mashed_bins;


did you fix this?

KrisThielemans · 2026-06-27T21:26:19Z

+BinNormalisationFromPETSIRD::read_norm_data(const string& filename)
+{
+  petsird::Header header;
+  petsird_data_sptr.reset(new petsird::binary::PETSIRDReader(filename));


I guess cope with HDF5

Using PETSIRD 0.9.1

KrisThielemans · 2026-06-27T22:28:37Z

Looks like we're green, and only 2 comments left! However, i haven't actually run this.

It'd be great to add a test like

run STIR2PETSIRD on mMR or ROOT data
read PETSIRD file with STIR
compare with original

However, this is probably non-trivial due to the possibility of different coordinates/order. @NikEfth what do you think?

KrisThielemans · 2026-06-27T22:36:45Z

well, at least I hope it's green now...

KrisThielemans self-assigned this Jul 16, 2025

NikEfth requested a review from Copilot December 18, 2025 17:34

Copilot AI reviewed Dec 18, 2025

View reviewed changes

Copilot AI mentioned this pull request Dec 19, 2025

Fix spelling errors in PETSIRDInfo.cxx geometry detection message #1671

Merged

Copilot AI mentioned this pull request Dec 19, 2025

[WIP] WIP address changes based on feedback for PETSIRD list-mode function #1672

Closed

7 tasks

NikEfth marked this pull request as ready for review December 22, 2025 14:34

NikEfth marked this pull request as draft December 23, 2025 08:38

NikEfth added the enhancement label Dec 23, 2025

KrisThielemans commented Mar 4, 2026

View reviewed changes

casperdcl added this to 2026-03 AIRBI Hackathon Mar 16, 2026

casperdcl moved this to In Progress in 2026-03 AIRBI Hackathon Mar 16, 2026

KrisThielemans commented Mar 19, 2026

View reviewed changes

Comment thread src/test/test_PETSIRDInfo_helpers.cxx Outdated

KrisThielemans mentioned this pull request Mar 19, 2026

Fix CMake export and include interfaces ETSInitiative/PETSIRD#181

Closed

1 task

NikEfth reviewed Mar 20, 2026

View reviewed changes

NikEfth marked this pull request as ready for review March 20, 2026 13:04

KrisThielemans commented Mar 20, 2026

View reviewed changes

NikEfth reviewed Mar 20, 2026

View reviewed changes

KrisThielemans commented May 29, 2026

View reviewed changes

Comment thread .github/workflows/build-test.yml

Comment thread .github/workflows/build-test.yml Outdated

Comment thread .github/workflows/build-test.yml Outdated

Comment thread .github/workflows/build-test.yml Outdated

Comment thread .github/workflows/build-test.yml Outdated

KrisThielemans commented Jun 16, 2026

View reviewed changes

Comment thread .github/workflows/build-test.yml Outdated

Comment thread src/include/stir/IO/PETSIRDCListmodeInputFileFormat.h Outdated

Comment thread src/include/stir/PETSIRDInfo.h

NikEfth force-pushed the PETSIRD branch from 8f11522 to 40db559 Compare June 19, 2026 08:39

NikEfth force-pushed the PETSIRD branch from 40db559 to cf9fa23 Compare June 19, 2026 08:57

NikEfth force-pushed the PETSIRD branch from cf9fa23 to 105adb2 Compare June 19, 2026 11:50

NikEfth and others added 8 commits June 27, 2026 21:46

Create CListModeDataBasedOnCoordinateMap from SAFIR

d53ac76

creating template classes for PETSIRD [WIP]

ff068cd

mainly a copy at this stage

initial attempt for CListModeDataPETSIRD [WIP]

d1e0d41

reading PETSIRD header [WIP]

4e98c2a

Updates on the PETSIRD interface and CMake [WIP]

feba1c3

PETSIRD: extracting position of each detector [WIP]

830e4f2

PETSIRD: orientations etc [WIP]

045274a

PETSIRD: fixes for detector map [WIP]

d1fbd7c

KrisThielemans force-pushed the PETSIRD branch from 105adb2 to 1b083af Compare June 27, 2026 21:11

KrisThielemans commented Jun 27, 2026

View reviewed changes

PETSIRD: first working version

35c5013

Using PETSIRD 0.9.1

KrisThielemans force-pushed the PETSIRD branch from 1b083af to e4192da Compare June 27, 2026 22:23

KrisThielemans force-pushed the PETSIRD branch from e4192da to 6b84c68 Compare June 27, 2026 22:35

PETSIRD: safe checking on type of scanner geometry

7c12a85

KrisThielemans force-pushed the PETSIRD branch from 6b84c68 to 7c12a85 Compare June 27, 2026 22:46

		- Mat3: std::array<std::array<float,3>,3> for compact fixed-size storage.
		- Vec3: std::array<float,3> for simple 3D vectors.

Uh oh!

Conversation

KrisThielemans commented Jul 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

NikEfth commented Dec 10, 2025

Uh oh!

NikEfth commented Dec 17, 2025

Uh oh!

NikEfth commented Dec 17, 2025

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI commented Dec 19, 2025

Uh oh!

Copilot AI commented Dec 19, 2025

Uh oh!

NikEfth commented Dec 20, 2025

Uh oh!

KrisThielemans left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

KrisThielemans commented Mar 4, 2026 via email

Uh oh!

Uh oh!

NikEfth left a comment

Choose a reason for hiding this comment

Uh oh!

NikEfth Mar 20, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

KrisThielemans left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

NikEfth left a comment

Choose a reason for hiding this comment

Uh oh!

NikEfth commented Mar 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codacy-production Bot commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Not up to standards ⛔

Uh oh!

NikEfth commented May 28, 2026

KrisThielemans commented Jul 16, 2025 •

edited

Loading

NikEfth commented Mar 20, 2026 •

edited

Loading

codacy-production Bot commented May 22, 2026 •

edited

Loading

codacy-production Bot commented Jun 19, 2026 •

edited

Loading