Merge pull request #102 from anwai98/u2-tracking-metrics

Minor update to scripts to add missing tracked objects

scripts/automatic_tracks.csv

@@ -1,8 +1,8 @@
 ,Cell_ID,Start,End,Parent_ID
 0,1,0,0,0
-1,2,1,6,0
+1,2,1,6,1
 2,10,1,1,1
-3,3,7,14,0
+3,3,7,14,2
 4,8,7,8,2
 5,4,15,15,3
 6,6,15,23,3
@@ -22,11 +22,11 @@
 20,45,4,23,44
 21,47,4,23,44
 22,50,0,4,0
-23,51,5,13,0
+23,51,5,13,50
 24,62,5,23,50
 25,52,14,15,51
-26,54,14,14,0
-27,55,15,16,0
+26,54,14,14,51
+27,55,15,16,54
 28,60,15,15,54
 29,56,17,22,55
 30,58,17,23,55
@@ -57,3 +57,60 @@
 55,125,4,13,0
 56,126,14,17,125
 57,128,14,23,125
+58,11,0,23,0
+59,12,0,23,0
+60,13,0,23,0
+61,14,0,20,0
+62,15,0,23,0
+63,16,0,23,0
+64,17,0,23,0
+65,18,0,23,0
+66,23,0,23,0
+67,24,0,23,0
+68,29,0,23,0
+69,34,0,23,0
+70,35,0,23,0
+71,36,0,3,0
+72,37,0,23,0
+73,38,0,23,0
+74,39,0,23,0
+75,48,0,23,0
+76,49,0,23,0
+77,67,0,12,0
+78,68,0,23,0
+79,69,0,23,0
+80,70,0,23,0
+81,75,0,23,0
+82,80,0,23,0
+83,81,0,23,0
+84,82,0,23,0
+85,83,0,23,0
+86,84,0,18,0
+87,85,0,23,0
+88,86,0,23,0
+89,91,0,23,0
+90,92,0,23,0
+91,93,0,23,0
+92,94,0,23,0
+93,95,0,1,0
+94,100,0,23,0
+95,101,0,23,0
+96,102,0,23,0
+97,107,0,23,0
+98,108,0,23,0
+99,109,0,23,0
+100,110,0,23,0
+101,115,0,23,0
+102,116,2,23,0
+103,121,2,23,0
+104,122,3,20,0
+105,123,3,23,0
+106,124,4,23,0
+107,129,6,23,0
+108,130,8,9,0
+109,131,8,19,0
+110,132,9,17,0
+111,133,13,23,0
+112,134,18,22,0
+113,135,19,23,0
+114,136,21,23,0

scripts/measure_for_automatic.py

@@ -1,7 +1,6 @@
 import os
 
 import json
-import h5py
 import numpy as np
 import pandas as pd
 import imageio.v3 as imageio
@@ -28,12 +27,17 @@ def _get_tracks_to_isbi():
         parent_id = int(parent_id)
         frames = np.where(segmentation == parent_id)[0]
 
+        if parent_id in track_info:
+            parent_val = track_info[parent_id]['parent']
+        else:
+            parent_val = None
+
         # store the parent's track information
         track_info[parent_id] = {
             'frames': list(np.unique(frames)),
             'daughters': list(daughter_ids),
             'frame_div': frames.max() + 1,
-            'parent': None,
+            'parent': parent_val,
             'label': parent_id,
         }
 
@@ -48,6 +52,20 @@ def _get_tracks_to_isbi():
                 'label': daughter_id
             }
 
+    # now, the next step is to store track info for objects that did not split.
+    for gt_id in np.unique(segmentation)[1:]:
+        if gt_id in track_info:
+            continue
+
+        frames = np.where(segmentation == gt_id)[0]
+        track_info[gt_id] = {
+            'frames': list(np.unique(frames)),
+            'daughters': [],
+            'frame_div': None,
+            'parent': None,
+            'label': gt_id,
+        }
+
     from deepcell_tracking import isbi_utils
     track_df = isbi_utils.trk_to_isbi(track_info)
     track_df.to_csv("automatic_tracks.csv")
@@ -57,7 +75,6 @@ def _get_tracks_to_isbi():
 def _get_tracks_df():
     tracks_path = "automatic_tracks.csv"
     segmentation = imageio.imread(os.path.join(ROOT, "tracking_result.tif"))
-    breakpoint()
     if os.path.exists(tracks_path):
         track_df = pd.read_csv(tracks_path)
     else:
@@ -67,19 +84,17 @@ def _get_tracks_df():
 
 def _get_metrics_for_autotrack(segmentation, seg_df):
     # NOTE: for ground-truth
-    with h5py.File(os.path.join(ROOT, "tracking_micro_sam.h5")) as f:
-        gt = f['labels'][:]
-
+    gt = imageio.imread(os.path.join(ROOT, "tracking_gt_corrected.tif"))
     gt_nodes = _get_node_attributes(gt)
     gt_df = pd.read_csv("gt_tracks.csv")
     gt_G = ctc_to_graph(gt_df, gt_nodes)
     _check_ctc(gt_df, gt_nodes, gt)
     gt_T = TrackingGraph(gt_G, segmentation=gt, name="DynamicNuclearNet-GT")
 
     # NOTE: for segmentation results
-    # calculate node attributes for each detection
     seg_nodes = _get_node_attributes(segmentation)
     seg_G = ctc_to_graph(seg_df, seg_nodes)
+    _check_ctc(seg_df, seg_nodes, segmentation)
     seg_T = TrackingGraph(seg_G, segmentation=segmentation, name="DynamicNuclearNet-autotracking")
 
     ctc_results = run_metrics(