Datasets:

Angelou0516
/

isles2024

sub_id string	shape string	n_slices int32	display_slice int32	lesion_voxels int64	has_lvo bool	cow_classes int32
sub-stroke0001	512x595x75	75	21	17,314	true	8
sub-stroke0002	512x605x75	75	44	314,986	true	10
sub-stroke0003	512x633x70	70	23	2,217	true	10
sub-stroke0004	512x587x69	69	32	6,275	true	11
sub-stroke0005	512x536x69	69	15	166,539	true	11
sub-stroke0006	512x544x64	64	22	265,018	true	10
sub-stroke0007	512x585x69	69	45	183,873	true	11
sub-stroke0008	512x554x80	80	57	46,810	true	9
sub-stroke0009	512x540x69	69	36	25,028	true	8
sub-stroke0010	512x603x79	79	41	40,117	true	10
sub-stroke0011	512x631x74	74	26	231,784	true	11
sub-stroke0012	512x612x75	75	30	54,952	true	9
sub-stroke0013	512x597x74	74	29	2,726	true	10
sub-stroke0014	512x586x74	74	38	757,886	true	11
sub-stroke0015	512x512x72	72	33	12,173	true	10
sub-stroke0016	512x610x80	80	50	27,941	true	8
sub-stroke0017	512x602x65	65	34	315,470	true	12
sub-stroke0019	512x576x80	80	43	13,497	true	11
sub-stroke0020	512x512x80	80	36	3,668	true	11
sub-stroke0021	512x531x74	74	45	16,374	true	10
sub-stroke0022	512x571x70	70	41	59,870	true	10
sub-stroke0025	512x610x69	69	32	227,916	true	11
sub-stroke0026	512x605x80	80	54	76,632	true	9
sub-stroke0027	512x612x65	65	30	415,331	true	8
sub-stroke0028	512x616x80	80	35	30,610	true	10
sub-stroke0030	512x586x64	64	32	33,553	true	10
sub-stroke0033	512x603x79	79	33	16,991	true	10
sub-stroke0036	512x544x69	69	18	18,696	true	9
sub-stroke0037	512x550x48	48	18	26,256	true	10
sub-stroke0038	512x558x65	65	22	3,169	true	10
sub-stroke0040	512x512x177	177	80	78,223	true	9
sub-stroke0043	512x615x74	74	28	6,372	true	10
sub-stroke0045	512x655x74	74	42	42,946	true	12
sub-stroke0047	512x595x69	69	30	17,023	true	9
sub-stroke0048	512x616x70	70	26	4,202	true	11
sub-stroke0049	512x572x79	79	48	12,045	true	9
sub-stroke0052	512x556x75	75	27	93,111	true	10
sub-stroke0053	512x584x64	64	33	7,094	true	11
sub-stroke0054	512x553x70	70	30	32,561	true	10
sub-stroke0055	512x553x70	70	28	9,728	true	11
sub-stroke0057	512x569x65	65	25	1,524	true	9
sub-stroke0062	512x562x69	69	30	32,802	true	11
sub-stroke0066	512x590x68	68	37	225,651	true	10
sub-stroke0068	512x559x70	70	23	20,347	true	9
sub-stroke0070	512x561x70	70	36	442,954	true	9
sub-stroke0071	512x599x65	65	26	16,676	true	10
sub-stroke0073	512x558x65	65	24	34,136	true	10
sub-stroke0074	512x527x70	70	40	6,402	true	9
sub-stroke0075	512x581x74	74	27	8,043	true	10
sub-stroke0076	512x643x75	75	36	6,520	true	10
sub-stroke0077	512x512x71	71	45	14,663	true	11
sub-stroke0078	512x512x74	74	30	302,578	true	12
sub-stroke0079	512x512x74	74	35	13,527	true	11
sub-stroke0080	512x551x74	74	51	400,774	true	5
sub-stroke0081	512x606x69	69	38	74,853	true	11
sub-stroke0082	512x588x70	70	29	351,602	true	10
sub-stroke0083	512x539x69	69	31	31,576	true	10
sub-stroke0084	512x512x73	73	43	3,217	true	9
sub-stroke0085	512x512x75	75	43	24,558	true	10
sub-stroke0086	512x533x70	70	40	937	true	10
sub-stroke0087	512x578x69	69	30	10,896	true	8
sub-stroke0088	512x512x69	69	46	39,527	true	10
sub-stroke0089	512x512x70	70	39	72,511	true	10
sub-stroke0090	512x533x69	69	34	135,243	true	10
sub-stroke0091	512x638x79	79	41	9,984	true	9
sub-stroke0092	512x594x69	69	38	4,204	true	11
sub-stroke0093	512x590x69	69	19	40,393	true	12
sub-stroke0094	512x652x89	89	61	290	true	9
sub-stroke0095	512x512x71	71	39	358,954	true	9
sub-stroke0096	512x577x74	74	42	263,405	true	10
sub-stroke0097	512x563x69	69	43	172,513	true	9
sub-stroke0098	512x619x74	74	33	300,178	true	9
sub-stroke0099	512x512x69	69	30	3,107	true	10
sub-stroke0100	512x512x71	71	54	1,587	false	12
sub-stroke0101	512x620x64	64	45	97,877	true	10
sub-stroke0102	512x512x68	68	29	12,334	true	10
sub-stroke0103	512x561x75	75	39	25,035	true	10
sub-stroke0104	512x512x69	69	38	52,237	true	11
sub-stroke0105	512x512x75	75	24	37,189	true	11
sub-stroke0106	512x512x69	69	40	10,390	true	11
sub-stroke0107	512x587x75	75	27	295,995	true	11
sub-stroke0108	512x512x69	69	28	25,086	true	12
sub-stroke0109	512x512x72	72	35	3,358	true	13
sub-stroke0110	512x636x69	69	21	27,487	true	12
sub-stroke0111	512x512x69	69	26	62,105	true	10
sub-stroke0112	512x512x76	76	39	13,165	true	10
sub-stroke0113	512x613x74	74	41	98,686	true	9
sub-stroke0114	512x512x65	65	32	151,742	true	9
sub-stroke0115	512x593x64	64	30	322,584	true	9
sub-stroke0116	512x544x74	74	46	114,086	true	11
sub-stroke0117	512x512x69	69	24	23,137	true	10
sub-stroke0118	512x512x74	74	52	59,535	true	9
sub-stroke0119	512x518x69	69	32	775	true	9
sub-stroke0133	512x686x45	45	26	42,975	true	9
sub-stroke0134	512x644x42	42	22	17,696	true	11
sub-stroke0135	512x658x40	40	16	34,810	true	8
sub-stroke0136	512x665x44	44	23	13,379	true	10
sub-stroke0137	512x512x85	85	39	422,537	true	10
sub-stroke0138	512x617x41	41	29	82,760	true	11
sub-stroke0139	512x573x44	44	20	25,921	true	9

End of preview. Expand in Data Studio

ISLES'24 - Ischemic Stroke Lesion Segmentation Challenge 2024

A real-world, longitudinal, multimodal acute-stroke dataset: acute admission CT (NCCT, CTA, 4D CT-perfusion + derived perfusion maps), follow-up MRI (DWI/ADC), and clinical tabular data, with the final infarct lesion as the segmentation target.

This repository mirrors the public TRAINING set only (149 cases). The 98-case ISLES'24 test set is withheld by the organizers for challenge scoring on Grand Challenge and is not part of any public release.

Provenance

Official source: Zenodo record 16748089 - "ISLES'24 - A Real-World Longitudinal Multimodal Stroke Dataset", train.7z, CC BY-NC-SA 4.0 (author-provided; not a third-party re-host).
Challenge: https://isles-24.grand-challenge.org/
Code / data-loading reference: https://github.com/ezequieldlrosa/isles24
This mirror is an unmodified raw copy of the released volumes (no resampling, no intensity changes), with two documented exceptions below.

Counts & faithfulness notes

149 training cases (sub-stroke0001 ... sub-stroke0189, non-contiguous): University Hospital Munich (TUM) + University Hospital Zurich.
The companion paper reports N=150 for training; the public Zenodo release ships 149 (one case excluded upstream). This mirror = the 149 released.
Dropped duplicate: the upstream release contains one stray uncompressed file sub-stroke0142_ses-02_lesion-msk.nii (a native-space leftover). Subject 0142 also has the standard co-registered *_space-ncct_lesion-msk.nii.gz like every other case, so this mirror keeps the standard GT and omits the stray.

Structure (BIDS)

<sub>/                      sub-stroke0001 ... sub-stroke0189  (149)
  ses-01  = acute admission CT
  ses-02  = follow-up MRI (2-9 days later)

raw_data/<sub>/ses-01/                     # native acquisition space
  *_ncct.nii.gz   *_cta.nii.gz   *_ctp.nii.gz (4D, 55 timepoints)
  perfusion-maps/ *_{cbf,cbv,mtt,tmax}.nii.gz

derivatives/<sub>/                         # ALL co-registered to NCCT (space-ncct)
  ses-01/ *_space-ncct_cta.nii.gz  *_space-ncct_ctp.nii.gz (4D)
          *_space-ncct_lvo-msk.nii.gz      # large-vessel occlusion (CTA), binary
          *_space-ncct_cow-msk.nii.gz      # Circle-of-Willis anatomy (CTA), multi-class
          perfusion-maps/ *_space-ncct_{cbf,cbv,mtt,tmax}.nii.gz
  ses-02/ *_space-ncct_dwi.nii.gz  *_space-ncct_adc.nii.gz
          *_space-ncct_lesion-msk.nii.gz   # <-- GROUND TRUTH (final infarct), binary

phenotype/<sub>/ses-*/ *_demographic_baseline.csv  *_outcome.csv

Co-registration: within each subject, the raw NCCT and every derivatives/*space-ncct* volume (CTA, perfusion maps, DWI/ADC, all masks) share an identical voxel grid (shape, spacing, affine). The raw NCCT itself is the reference space (hence there is no space-ncct_ncct). So a fully aligned input+label stack = raw ncct + the derivatives space-ncct modalities + space-ncct_lesion-msk.

Ground truth

Mask	Space	Type	Role
`space-ncct_lesion-msk` (ses-02)	NCCT	binary {0,1}	Gold standard - final infarct lesion, the challenge target
`space-ncct_lvo-msk` (ses-01)	NCCT	binary {0,1}	Auxiliary - large-vessel occlusion (on CTA)
`space-ncct_cow-msk` (ses-01)	NCCT	multi-class	Auxiliary - Circle-of-Willis anatomy (on CTA)

The gold-standard lesion mask is derived from the follow-up DWI: masks were auto-generated by DeepISLES (the productized ISLES'22 winning ensemble) and then quality-controlled / manually corrected and verified by two neuroradiologists (>10 years' experience). The neuroradiologist-verified DWI lesion is the reference standard; it is provided co-registered into NCCT space.

Cross-dataset overlap (leakage note)

No patient reuse from ISLES'22 (a different, MRI-only DWI cohort). ISLES'24 is a new acute-stroke CT cohort. Modality and patients differ.
There is center-level overlap (Munich contributes to both editions) and a methodological dependency (ISLES'24 GT masks are produced by DeepISLES, which was trained on ISLES'22) - but neither is shared imaging.
No shared TCIA / BraTS / Medical Segmentation Decathlon lineage. Identifiers are challenge-internal BIDS sub-stroke#### IDs only.

License

CC BY-NC-SA 4.0 (per the Zenodo deposit). Non-commercial, share-alike, attribution required.

Citation

@article{delarosa2024isles24,
  title   = {ISLES'24: Final Infarct Prediction with Multimodal Imaging and Clinical Data. Where Do We Stand?},
  author  = {de la Rosa, Ezequiel and Su, Ruisheng and Reyes, Mauricio and Riedel, Eda Otman and Baazaoui, Hakim and Wiest, Roland and Kofler, Florian and Kirschke, Jan S. and Wiestler, Benedikt and Menze, Bjoern},
  journal = {arXiv preprint arXiv:2408.10966},
  year    = {2024}
}

Data: Zenodo record 16748089 (de la Rosa et al., 2025). Please also credit the ISLES'24 challenge organizers and the contributing hospitals (TUM Munich, USZ Zurich).

Downloads last month: 1,821

Paper for Angelou0516/isles2024

ISLES'24: Final Infarct Prediction with Multimodal Imaging and Clinical Data. Where Do We Stand?

Paper • 2408.10966 • Published Aug 20, 2024