Table 5 Initial class distribution concerning each target endpoint and time window, after creating the learning examples considering 3, 4, and 5 consecutive snapshots of patient historical assessments.

From: Triclustering-based classification of longitudinal data for prognostic prediction: targeting relevant clinical endpoints in amyotrophic lateral sclerosis

  

90 days

180 days

365 days

N

Y

N

Y

N

Y

C1

3 CS

2640 (94%)

176 (6%)

2228 (83%)

445 (17%)

1868 (85%)

331 (15%)

4 CS

2229 (93%)

176 (7%)

1839 (81%)

423 (19%)

1571 (85%)

282 (15%)

5 CS

1912 (92%)

175 (8%)

1537 (79%)

408 (21%)

1338 (84%)

255 (16%)

C2

3 CS

3533 (97%)

128 (3%)

3304 (92%)

285 (8%)

3093 (94%)

199 (6%)

4 CS

3045 (96%)

127 (4%)

2822 (91%)

279 (9%)

2668 (94%)

176 (6%)

5 CS

2647 (95%)

127 (5%)

2434 (90%)

269 (10%)

2317 (94%)

159 (6%)

C3

3 CS

4058 (98%)

62 (2%)

3888 (96%)

183 (4%)

3669 (97%)

126 (3%)

4 CS

3474 (98%)

62 (2%)

3308 (95%)

179 (5%)

3144 (97%)

97 (3%)

5 CS

3001 (98%)

62 (2%)

2846 (94%)

168 (6%)

2721 (97%)

75 (3%)

C4

3 CS

1692 (87%)

263 (13%)

1298 (71%)

523 (29%)

1005 (72%)

388 (28%)

4 CS

1415 (84%)

263 (16%)

1041 (67%)

504 (33%)

833 (70%)

356 (30%)

5 CS

1226 (82%)

263 (18%)

866 (64%)

490 (36%)

714 (68%)

329 (32%)

C5

3 CS

3246 (96%)

146 (4%)

2907 (88%)

398 (12%)

2479 (89%)

309 (11%)

4 CS

2730 (95%)

146 (5%)

2402 (86%)

387 (14%)

2066 (89%)

266 (11%)

5 CS

2328 (94%)

146 (6%)

2009 (84%)

378 (16%)

1750 (89%)

223 (11%)

  1. C1, need for NIV; C2, need for an auxiliary communication device; C3, need for PEG; C4, need for a caregiver, and C5, need for a wheelchair.