AbsCBF_T1Fit (Preclinical)
==========================

**Purpose.** Voxelwise **absolute CBF** from MTI-PASL by fitting T1 curves on **global** vs **selective** series.

Computation
-----------

For each voxel, fit T1 to global and selective TI curves, then:

``CBF = 4980 * (T1_glo / 2250) * (1000/T1_sel - 1000/T1_glo)``  (T1 in ms)

Context I/O
-----------

- **Reads** from ``ctx``:
  - ``AbsData`` *(optional)*: 3D stack (X,Y,N). If absent, the module loads from ``params["data_path"]``.
  - ``savedir`` *(optional)*: output directory (default ``"."``)
- **Writes** to ``ctx``:
  - ``absCBF``: 2D CBF map (X,Y)

Filesystem outputs
------------------

- ``<savedir>/absCBF.npy`` — absolute CBF map
- ``<savedir>/curvefit.png`` — sampled voxel fits for QC

Parameters
----------

====================  ===================  ==============================================
Key                   Type / Default       Meaning
====================  ===================  ==============================================
``data_path``         str / *None*         .npy/.npz path if ``AbsData`` not provided
``data_key``          str / *None*         Key in .npz
``TI_list``           list[float] / *req*  Inversion times (ms)
``sel_index``         list[int] / *req*    Frame indices for selective series
``glo_index``         list[int] / *req*    Frame indices for global series
``mask_path``         str / *None*         2D numpy mask to constrain fitting
``save_curves_every`` int / ``200``        Approx. sampling for diagnostic plot
====================  ===================  ==============================================

YAML example
------------

.. code-block:: yaml

    modules:
      - name: AbsCBF_T1Fit
        params:
          data_path: ./data/abs_input.npz
          data_key: frames
          TI_list: [200, 400, 600, 800, 1000]
          sel_index: [1,3,5,7,9]
          glo_index: [0,2,4,6,8]
          save_curves_every: 300