Traditional and machine learning models for predicting haemorrhagic transformation in ischaemic stroke: a systematic review and meta-analysis

Table 1 Summary of baseline characteristics of included studies

	Model development studies (n = 67)	Model validation studies (n = 33)
Publication year
2008–2013	9	6
2014–2019	13	20
2020–2024	45	7
Type of study
Derivation	6	0
Derivation and validation	61	0
Validation	0	33
Type of prediction model
Traditional model	44	33
Machine learning model	23	0
Study population
General ischaemic stroke	11	5
Ischaemic stroke with an indication for anticoagulation	0	1
Ischaemic stroke with thrombectomy	14	4
Ischaemic stroke with thrombolysis	35	23
Ischaemic stroke with thrombolysis or thrombectomy	2
Ischaemic stroke without thrombolysis or thrombectomy	5	0
Multicentre
Yes	38	20
No	27	13
Not reported	2
Type of haemorrhagic transformation
Any haemorrhagic transformation	39	11
Radiological category
Haemorrhagic infarction	2	0
Parenchymal haemorrhage	6	1
Clinical category
aICH	2	2
sICH	37	29
NINDS	13	15
SITS-MOST	3	12
ECASS II	16	16
ECASS III	5	2
IST-3	1	0
HBC	4	1

Abbreviations: aICH Asymptomatic intracerebral haemorrhage, sICH Symptomatic intracerebral haemorrhage, NINDS National Institutes of Neurological Diseases and Stroke, SITS-MOST Safe Implementation of Thrombolysis in Stroke-Monitoring Study, ECASS II European-Australian Cooperative Acute Stroke Study II, ECASS III European-Australian Cooperative Acute Stroke Study III, IST-3 The Third International Stroke Trial, HBC Heidelberg Bleeding Classification

ISSN: 2046-4053