pipeline.yaml

# Beamforming config for the OpenH-RF eSAF rotational 3D ultrasound sub-dataset.
#
# Each acquisition is raw RF channel data (n_ch = 1) from a single normal
# plane-wave transmit per rotation frame on the JP_Linear_68 linear array
# (68 elements, 10 MHz, fs 40 MHz, c 1490 m/s). The RF is demodulated before
# delay-and-sum beamforming; the imaging grid is sized automatically per file
# from the probe aperture and the recorded axial depth, so the same config
# works for every acquisition (targets span 20-130 mm).
#
# reconstruct.py loads this whole processing chain with zea.Config.from_path
# and runs it unchanged (single track -> single pipeline.yaml).

parameters:
  dynamic_range: [-40, 0]
  apply_lens_correction: false

pipeline:
  operations:
    - name: cast
      params:
        dtype: float32
    - name: demodulate
    - name: beamform
      params:
        beamformer: delay_and_sum
        num_patches: 200
    - name: envelope_detect
    - name: normalize
    - name: log_compress