Using artificial intelligence to predict the risk of blood vessel re-narrowing after treatment in people with blocked arteries in the legs

ISRCTN	ISRCTN22632665
DOI	https://doi.org/10.1186/ISRCTN22632665
Sponsor	Shanxi Bethune Hospital (Shanxi Academy of Medical Sciences)
Funder	Science Fundation of Shanxi Bethune Hospital

Submission date: 14/10/2026
Registration date: 15/10/2026
Last edited: 15/10/2026

Recruitment status: No longer recruiting
Overall study status: Completed
Condition category: Circulatory System

Prospectively registered

Protocol

Statistical analysis plan

Results

Individual participant data

Record updated in last year

Plain English summary of protocol

Not provided at time of registration

Contact information

Dr Jing Li
Principal investigator, Scientific, Public

No. 99 Longcheng Street, Xiaodian District
Taiyuan
030032
China

ORCID ID	0009-0009-8191-7251
Phone	+86 13994247485
Email	lijing_7112@163.com

Study information

Primary study design	Observational
Observational study design	Cohort study
Scientific title	Prediction of one‑year restenosis risk among patients with lower extremity arteriosclerosis obliterans receiving interventional therapy: comparing support vector machine with SHapley Additive exPlanations interpretability against K‑nearest neighbor, neural network and discriminant prediction models for model discrimination, calibration and individual‑level risk factor attribution
Study objectives
Ethics approval(s)	Approved 10/10/2023, Shanxi Bethune Hospital (Shanxi Academy of Medical Sciences) (No. 99 Longcheng Street, Xiaodian District, Taiyuan, 030032, China; +86 0351-8368114; lijing@sxbqeh.com.cn), ref: [2022] Lun Shen Medical Device No. (12)
Health condition(s) or problem(s) studied	Prediction of post‑interventional restenosis in patients with lower extremity arteriosclerosis obliterans
Methodology	This is a retrospective observational cohort study. Researchers will collect and extract existing electronic‑medical‑record data from participants diagnosed with lower extremity arteriosclerosis obliterans (LEASO) who received interventional therapy. Collected data will cover demographics, clinical characteristics, lesion‑related indicators, comorbidities, smoking status, and postoperative medication adherence. All participants will receive routine clinical interventional therapy selected according to lesion location and severity, including percutaneous transluminal angioplasty, arterial atherectomy, or endovascular stent angioplasty. After the procedure, participants will receive standard postoperative anticoagulant therapy, antiplatelet therapy, circulation‑improving treatment, as well as guidance for medication and lifestyle modification as per clinical practice. No study‑specific interventions or control groups will be assigned. Participants will undergo scheduled follow‑up assessments every 3 months for 1 year after the intervention. Outcome evaluation will be performed via outpatient review, telephone follow‑up, and computed tomography angiography (CTA). Investigators will assess whether participants develop restenosis, defined as ≥50 % luminal stent stenosis on CTA or recurrent limb symptoms with hemodynamically significant stenosis confirmed by vascular ultrasound or CTA. The collected dataset will be randomly split into a training set and a validation set at a 7:3 ratio. Multivariable logistic regression will screen risk factors. Four machine‑learning algorithms (support vector machine, K‑nearest neighbor, neural network, discriminant model) will build prediction models with ten‑fold cross‑validation for internal performance evaluation. SHAP analysis will be applied to interpret feature importance of the optimal prediction model.
Intervention type	Procedure/Surgery
Primary outcome measure(s)	Restenosis within 1‑year after interventional therapy for lower extremity arteriosclerosis obliterans measured using computed tomography angiography (CTA), vascular ultrasound, outpatient review and telephone follow‑up. Restenosis is defined as either CTA demonstrating ≥50% luminal stent stenosis, or recurrent intermittent claudication/resting pain of the affected limb together with hemodynamically significant stenosis confirmed by vascular ultrasound or CTA. Medical record data are extracted for outcome adjudication at 12‑months after interventional therapy
Key secondary outcome measure(s)
Completion date	30/12/2025

Eligibility

Participant type(s)
Age group	Mixed
Lower age limit	18 Years
Upper age limit	100 Years
Sex	All
Target sample size at registration	215
Total final enrolment	205
Key inclusion criteria	1. Aged ≥18 years 2. Diagnosis of lower extremity arteriosclerosis obliterans (LEASO) confirmed by digital subtraction angiography or computed tomography angiography 3. Received interventional therapy for LEASO 4. Written informed consent obtained from the patient or their family member after study explanation
Key exclusion criteria	1. Previous history of lower‑limb vascular bypass surgery or interventional therapy 2. Cognitive impairment or psychiatric disease 3. Lactation or pregnancy status 4. Acute myocardial infarction, cerebral hemorrhage or cerebral thrombosis within the preceding 3 months 5. Coagulation dysfunction 6. Autoimmune disease 7. Malignant tumor
Date of first enrolment	25/12/2022
Date of final enrolment	25/12/2025

Locations

Countries of recruitment

China

Study participating centres

Results and Publications

Individual participant data (IPD) Intention to share	No

Editorial Notes

14/10/2026: Trial's existence confirmed by Shanxi Bethune Hospital (Shanxi Academy of Medical Sciences).