AI-Powered Real-Time Detection and Quantitative Analysis of Bilateral acute and sub-acute Infarcts Using Non-Contrast Computed Tomography Imaging: A Deep Learning-Based Approach
试验速览
- 阶段
- 不适用
- 状态
- 尚未招募
- 发起方
- 入组人数
- 100
- 试验地点
- 1
- 主要终点
- Diagnostic accuracy of AI in detecting bilateral MCA infarcts on non-contrast CT, measured by sensitivity, specificity, PPV, NPV, and AUC-ROC.
研究概览
简要总结
This study aims to evaluate the diagnostic accuracy of an AI-driven real-time detection system for acute and sub-acute infarcts using non-contrast CT images. The study will assess the sensitivity, specificity, and clinical utility of AI-based infarct detection compared to expert radiologist interpretation. By leveraging deep learning algorithms, the system provides automated infarct detection, real-time image analysis, and AI-generated reports, assisting in early stroke diagnosis and management. The study will be conducted as a prospective observational diagnostic accuracy study, comparing AI results with the gold standard (radiologist reports and follow-up MRI-DWI findings). The findings will help determine the effectiveness and reliability of AI in infarct detection, particularly in resource-limited settings where radiology expertise may be scarce.
研究设计
- 研究类型
- Observational
入排标准
- 年龄范围
- 18.00 Year(s) 至 90.00 Year(s)（—）
- 性别
- All
入选标准
- •Patients aged 18 years and above.
- •Presenting with acute or subacute neurological symptoms suggestive of stroke (e.g., weakness, slurred speech, visual disturbances, altered sensorium).
- •Undergoing non-contrast CT (NCCT) of the brain as part of initial evaluation.
- •Undergoing MRI with diffusion-weighted imaging (DWI) within 72 hours of the NCCT for diagnostic correlation.
- •Symptom onset within 1 to 72 hours prior to imaging.
- •Informed consent obtained from the patient or legally authorized representative.
排除标准
- •Poor-quality NCCT images due to motion artifacts or technical issues.
- •Presence of non-ischemic pathology on CT such as hemorrhage, tumors, trauma, or postoperative changes.
- •History of prior stroke in the same vascular territory.
- •Inability to undergo MRI, including contraindications such as pacemakers, metallic implants, severe claustrophobia, or clinical instability.
- •Patients who have received reperfusion therapy (e.g., thrombolysis or thrombectomy) before the baseline NCCT.
- •Incomplete clinical, laboratory, or follow-up imaging data that prevents accurate analysis.
结局指标
主要结局
Diagnostic accuracy of AI in detecting bilateral MCA infarcts on non-contrast CT, measured by sensitivity, specificity, PPV, NPV, and AUC-ROC.
时间窗: Real-time analysis immediately after CT scan
次要结局
- 1)Diagnostic accuracy of AI in detecting infarcts across ASPECTS severity categories (0–4, 5–7, 8–10), measured by sensitivity.(2)Time-to-diagnosis comparison between AI & radiologists, measured in minutes from CT scan completion to diagnosis report generation.)
研究者
DrAbdul Majith Seeni Mohammed
Saveetha medical college and hospital, Saveetha institute of medical and technical sciences