Native T1 mapping

Native T1 mapping provides myocardial and blood-pool tissue characterization without contrast. In CardiacNexus, this modality feeds Myocardial tissue characterization and remains separate from cine-derived structure, function, and strain pages.

Modality
Native T1 mapping
UKB source
Data Field 20214
Pipeline step
ShMOLLI T1 map QC, myocardial/blood-pool T1 extraction, and optional aggregate blood-pool correction
Outputs
native_T1.csv, native_T1_corrected.csv, visualization/ventricle/native_t1*.png
Maturity
Source-audited modality page

Current pipeline role

The current source expects shmolli_t1map.nii.gz and seg_shmolli_t1map.nii.gz. Segmentation labels are myocardium, LV blood pool, and RV blood pool. The subject is skipped when the ShMOLLI segmentation fails quality control or when the T1 helper cannot calculate the required regions.

Uncorrected rows are emitted by the subject-level native T1 route. Corrected rows are emitted later from the aggregate native T1 CSV using cohort-level blood-pool correction coefficients.

Current outputs

Output familyCurrent rowsSource pathNotes
Native myocardial T1global, IVS, and free-wall myocardial T1 rowssrc/feature_extraction/Native_T1_20214/eval_native_t1.pyvalues currently formatted to two decimals
Native blood-pool T1LV and RV blood-pool T1 rowseval_native_t1.pyused directly and as correction inputs
Blood-corrected native T1corrected global, IVS, and free-wall rowssrc/feature_extraction/Combined_Features/eval_native_t1_corrected.pycohort-dependent aggregate correction
QC and visualizationnative_t1.png, native_t1_ivs_fw_blood.pngeval_native_t1.pysegmentation and regional overlay context

QC and interpretation caveats

Native T1 values are sensitive to scanner, field strength, sequence, reconstruction, motion, segmentation, partial volume, and blood-pool contamination. Corrected and uncorrected rows should not be pooled unless method labels and correction behavior are explicit.

The corrected route depends on the current aggregate cohort because correction coefficients are fitted from available native T1 rows. Changing cohort composition or missingness can change corrected values even when a subject's uncorrected native T1 row is unchanged.

Source audit

  • Current input names, ShMOLLI label map, QC requirement, output labels, visualization paths, and helper boundary were checked against eval_native_t1.py.
  • Corrected output labels and cohort-level correction behavior were checked against eval_native_t1_corrected.py.
  • docs/data/output_column_inventory.yml records native_T1 and native_T1_corrected artifact contracts.
  • Textbook context boundary: broad clinical textbook context is not surfaced here because page-specific T1 mapping, ShMOLLI, and correction sources are sufficient for draft rollout.
Native T1 mapping context from the CardiacNexus manuscript image cache
Native T1 mapping context figure from the manuscript image cache, cited there to Barison et al. 2022, shown to orient the acquisition/source-image concept behind native T1 phenotype rows. CardiacNexus output labels, correction behavior, and scanner/method caveats remain defined in the linked phenotype page.Source: manuscript image cache citing Barison et al. 2022; permission and exact source-panel review pending for draft use.
  • Displayed figure provenance is recorded in docs/data/figure_provenance.yml; permission and exact reuse wording remain pending for public release.