List mode data is the richest output a PET scanner produces, and most sites throw it away. It is the raw, time-stamped record of every detected event in a PET acquisition, stored before any image is formed. Because it preserves the underlying event stream rather than a single fixed image, it can be reconstructed repeatedly under different parameters. Standard clinical acquisition produces one reconstruction and frequently discards the list mode file afterward, which is why it has to be requested and preserved deliberately.
For theranostics programs, that loss is expensive. Establishing how much activity a patient actually needs means comparing one acquisition reconstructed at several count levels, and only list mode makes that comparison possible. Preserved, the dose question gets answered from scans already on the books. Discarded, the alternatives are more imaging visits, more exposure, and more schedule. Bracken built a centralized reconstruction workflow to keep that option open across every scanner in a study.
Why List Mode Rarely Fits Neatly Into a Clinical Trial
List mode reconstruction itself is only one part of the challenge. The larger difficulty is coordinating the sites, scanner models, data transfers, reconstruction parameters, and quality processes required to make the resulting datasets comparable.
List mode availability varies by site and scanner. Most clinical sites cannot export list mode data. Many scanners discard the event stream automatically once a standard reconstruction completes, so export capability has to be confirmed during site selection rather than discovered afterward.
Reconstruction is scanner-specific. Duplicating a site's reconstruction methodology depends on scanner-specific parameters that live with the site physicist, not in the protocol.
List mode datasets are large enough to defeat standard trial platforms. Several imaging and data capture systems will not accept them at all, and bandwidth limits stall the transfers that do begin.
Dose-finding requires several reconstructions of a single acquisition. Producing 100%, 75%, 50%, and 25% count levels consistently across multiple sites and scanner models demands one controlled pipeline, not site-by-site improvisation.
Quality control and site queries fragment across systems. When sites are asked to manage separate logins for upload and for query resolution, response times slow and the query record scatters.
A Centralized Reconstruction Workflow
Bracken established a six-stage workflow with clear ownership from study initiation through analysis handback. Centralizing the reconstruction is what buys comparability: every dataset in the study runs through one method, so results from different sites and scanner models can be read against each other.
Initiate and Align
Bracken opens physicist-to-physicist communication with each site, runs the kickoff, and collects the site list and scanner details needed to scope the build.
Configure Scanners
Bracken builds a scanner adapter for each model in the study, duplicating that site's reconstruction methodology so outputs from different models stay comparable. Timing varies per scanner model. The build is a one-time effort: any later study on the same model inherits the configuration.
Transfer Data
Sites send acquired data to the sponsor, who forwards it to Bracken. Transfer runs electronically through the study platform where supported, with physical media held as needed as a documented contingency for datasets that exceed platform limits. Rolling, within five days of image acquisition.
QC and Query
Bracken reviews each dataset for completeness and quality and issues queries where something is missing or out of specification. QC is conducted on a rolling basis as each dataset is received, with query management centralized so sites can work within a single system.
Reconstruction
Each validated dataset is reconstructed to selected simulated count levels and returned as DICOM.
Analysis Handback
The value of a reconstruction workflow is not that it runs once. It is that the second study on the same scanner starts from a configuration that already exists. Building the adapter is an investment. Every protocol after it inherits the work.
Common Questions About List Mode Reconstruction
List mode is the raw, time-stamped record of every detected event in a PET acquisition, stored before any image is formed. Because it preserves the underlying event stream rather than a single fixed image, it can be reconstructed repeatedly under different parameters. Standard clinical acquisition produces one reconstruction and frequently discards the list mode file afterward, which is why it has to be requested and preserved deliberately.
Reconstructing the same acquisition at selected reduced dose levels of its counts simulates what the image would have looked like at lower administered activity, using data already collected from the patient. That lets a team assess image quality and quantitative performance across a dose range without additional scans or additional exposure. The comparison is only valid when every level comes from the same acquisition and the same reconstruction method, which is the reason for centralizing the work.
Setup is driven by the scanner models in the study rather than the number of patients. Bracken builds a scanner adapter for each model, duplicating that site's reconstruction methodology so outputs stay comparable across the study. That build is a one-time cost per model, and any later protocol using the same scanner model starts from the configuration that already exists. Once a scanner is calibrated, datasets move through quality control and reconstruction on a rolling basis as they arrive. Bracken scopes the schedule during study initiation, once the scanner models and site list are known.
No. Many scanners discard list mode automatically after a standard reconstruction, and not every site has the export capability or the local physics support to retrieve it. Academic centers with research scanners are usually the strongest candidates. Export capability should be verified during site selection, because discovering the gap after activation costs schedule that a dose-finding stage rarely has.
