Tuple index out of range error during post-processing step in Nucleus Instance Segmentation task

[akihikoueda]

• TIA Toolbox version: 1.6.0
• Python version: 3.10.15
• Operating System: Ubuntu 22.04.5

Description

I attempted to perform a Nucleus Instance Segmentation task using the mapde-conic model pre-trained on the CoNIC Dataset and tried generating patches. While the predictions work well, an error occurs during the post-processing step. Could you kindly advise on how to address this issue?

What I Did

[Execution Code]

img_file_name_lists = ["sample_tile.png"]

from tiatoolbox.models import IOSegmentorConfig
ioconfig = IOSegmentorConfig(
    input_resolutions=[
        {'units': 'mpp', 'resolution': 0.25}
    ],
    output_resolutions=[
        {'units': 'mpp', 'resolution': 0.25},
        {'units': 'mpp', 'resolution': 0.25},
        {'units': 'mpp', 'resolution': 0.25},
        {'units': 'mpp', 'resolution': 0.25}
    ],
    margin=26,
    tile_shape=[1024, 1024],
    patch_input_shape=(252, 252),
    patch_output_shape=(150, 150),
    stride_shape=(150, 150),
    save_resolution={'units': 'mpp', 'resolution': 0.25}
)

# Tile prediction
inst_segmentor = NucleusInstanceSegmentor(
    pretrained_model="mapde-conic",
    num_loader_workers=2,
    num_postproc_workers=2,
    batch_size=4,
)

tile_output = inst_segmentor.predict(
    img_file_name_lists,
    save_dir="sample_tile_results/",
    mode="tile",
    on_gpu=False,
    crash_on_exception=False,
    ioconfig=ioconfig
)
"""

[Error Code]
"""
Traceback (most recent call last):
  File "/home/ueda/.pyenv/versions/3.10.15/lib/python3.10/concurrent/futures/process.py", line 246, in _process_worker
    r = call_item.fn(*call_item.args, **call_item.kwargs)
  File "/home/ueda/kobe/lib/python3.10/site-packages/tiatoolbox/models/engine/nucleus_instance_segmentor.py", line 281, in _process_tile_predictions
    head_predictions = [v[idx][0] for v in predictions]
  File "/home/ueda/kobe/lib/python3.10/site-packages/tiatoolbox/models/engine/nucleus_instance_segmentor.py", line 281, in <listcomp>
    head_predictions = [v[idx][0] for v in predictions]
IndexError: tuple index out of range

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/home/ueda/kobe/lib/python3.10/site-packages/tiatoolbox/models/engine/semantic_segmentor.py", line 1225, in _predict_wsi_handle_exception
    self._predict_one_wsi(wsi_idx, ioconfig, str(wsi_save_path), mode)
  File "/home/ueda/kobe/lib/python3.10/site-packages/tiatoolbox/models/engine/nucleus_instance_segmentor.py", line 753, in _predict_one_wsi
    self._merge_post_process_results()
  File "/home/ueda/kobe/lib/python3.10/site-packages/tiatoolbox/models/engine/nucleus_instance_segmentor.py", line 805, in _merge_post_process_results
    raise future.exception()  # noqa: RSE102
IndexError: tuple index out of range

[shaneahmed]

mapde-conic is a nucleus detection model. It is not an instance segmentation model. We are working on Nucleus detection model in #538