Hi
I was wondering if there's something wrong with the way I use TractoEval. Here's the scripts I use:
,the candidate trk file is produced on a 2mm DWI image , and ref trk file is the original groundtruth.
And the output is like that:
I can see that there are 25 bundles being recognized.
However, the final mearsures_aggregated.json indicates all bundles get a 0.0 of DICE,OL while OR is null. Am I use the pipeline in a wrong way?
{"subjects": {"sub-1006": {"FAT_L": {"dice": 0.0, "overlap": 0.0, "overreach": null, "streamlines_count": 406, "volume": 16120.0, "volume_endpoints": 2080.0, "avg_length": 77.33497619628906, "std_length": 5.265476703643799, "min_length": 60.000003814697266, "max_length": 93.99998474121094, "span": 33.59373092651367, "curl": 2.302065714744796, "diameter": 16.29106963173533, "elongation": 4.7470778742261945, "surface_area": 12528.0, "end_surface_area_head": 2448.0, "end_surface_area_tail": 1344.0, "radius_head": 7.771808505735393, "radius_tail": 6.235815091705328, "irregularity": 3.16523759842401, "irregularity_of_end_surface_head": 0.07751444496477941, "irregularity_of_end_surface_tail": 0.0908943862423208, "mean_curvature": 0.13636115123632506, "fractal_dimension": 2.1065881765927696}, "UF_L": {"dice": 0.0, "overlap": 0.0, "overreach": null, "streamlines_count": 871, "volume": 29072.0, "volume_endpoints": 5056.0, "avg_length": 89.68082427978516, "std_length": 9.4235258102417, "min_length": 61.999996185302734, "max_length": 114.00001525878906, "span": 20.870037078857422, "curl": 4.297109005648922, "diameter": 20.31621118553331, "elongation": 4.41424946122064, "surface_area": 21472.0, "end_surface_area_head": 5992.0, "end_surface_area_tail": 3920.0, "radius_head": 13.978698877556546, "radius_tail": 12.685019746555582, "irregularity": 3.751287205694235, "irregularity_of_end_surface_head": 0.1024499067037899, "irregularity_of_end_surface_tail": 0.1289573502551238, "mean_curvature": 0.1733832083189118, "fractal_dimension": 2.165310453463346}, ......, "scores": {"dice": 0.0, "overlap": 0.0, "overreach": NaN, "total_streamlines": 80736}}
Hi
I was wondering if there's something wrong with the way I use TractoEval. Here's the scripts I use:
,the candidate trk file is produced on a 2mm DWI image , and ref trk file is the original groundtruth.
And the output is like that:
I can see that there are 25 bundles being recognized.
However, the final mearsures_aggregated.json indicates all bundles get a 0.0 of DICE,OL while OR is null. Am I use the pipeline in a wrong way?
{"subjects": {"sub-1006": {"FAT_L": {"dice": 0.0, "overlap": 0.0, "overreach": null, "streamlines_count": 406, "volume": 16120.0, "volume_endpoints": 2080.0, "avg_length": 77.33497619628906, "std_length": 5.265476703643799, "min_length": 60.000003814697266, "max_length": 93.99998474121094, "span": 33.59373092651367, "curl": 2.302065714744796, "diameter": 16.29106963173533, "elongation": 4.7470778742261945, "surface_area": 12528.0, "end_surface_area_head": 2448.0, "end_surface_area_tail": 1344.0, "radius_head": 7.771808505735393, "radius_tail": 6.235815091705328, "irregularity": 3.16523759842401, "irregularity_of_end_surface_head": 0.07751444496477941, "irregularity_of_end_surface_tail": 0.0908943862423208, "mean_curvature": 0.13636115123632506, "fractal_dimension": 2.1065881765927696}, "UF_L": {"dice": 0.0, "overlap": 0.0, "overreach": null, "streamlines_count": 871, "volume": 29072.0, "volume_endpoints": 5056.0, "avg_length": 89.68082427978516, "std_length": 9.4235258102417, "min_length": 61.999996185302734, "max_length": 114.00001525878906, "span": 20.870037078857422, "curl": 4.297109005648922, "diameter": 20.31621118553331, "elongation": 4.41424946122064, "surface_area": 21472.0, "end_surface_area_head": 5992.0, "end_surface_area_tail": 3920.0, "radius_head": 13.978698877556546, "radius_tail": 12.685019746555582, "irregularity": 3.751287205694235, "irregularity_of_end_surface_head": 0.1024499067037899, "irregularity_of_end_surface_tail": 0.1289573502551238, "mean_curvature": 0.1733832083189118, "fractal_dimension": 2.165310453463346}, ......, "scores": {"dice": 0.0, "overlap": 0.0, "overreach": NaN, "total_streamlines": 80736}}