基于组织对称性的超声颅脑异常区域自动检测及人工智能辅助颅内出血识别方法

An Automatic Detection Method for Abnormal Cranial Regions Based on Tissue Symmetry in Ultrasound Combined with an AI-Assisted Recognition Method for Intracranial Hemorrhage

目的:新生儿颅内出血是常见且严重的神经系统疾病,早产、极低体重和缺氧等因素是其主要致病原因。及早诊断和治疗对于减轻神经功能缺损、改善患者预后具有重要意义。方法:设计实现一种基于组织对称性的超声颅脑异常区域自动检测及人工智能辅助颅内出血识别方法,验证该方法的准确率、敏感度和特异度,评估其对提高诊断效率和改善患者预后的应用价值。结果:研究结果显示,超声颅脑异常区域自动检测及人工智能辅助颅内出血识别方法具有较高出血识别的敏感度,可明显提升医生对新生儿颅内出血疾病的诊断效率。结论:颅脑超声因其无创、便捷,适合用于无明显症状的新生儿颅脑疾病筛查,在颅脑容积超声扫查系统基础上引入超声颅脑异常区域自动检测及人工智能辅助颅内出血识别的技术,进一步帮助医生及早发现新生儿颅内出血情况,有效提升新生儿颅内出血的早期诊治率。

Objective:Neonatal intracranial hemorrhage is a common and severe neurological disorder,primarily caused by factors such as prematurity,extremely low birth weight,and hypoxia.Early diagnosis and treatment are crucial for reducing neurological deficits and improving patient outcomes.Methods:This study designed and implemented an automatic detection method for abnormal cranial regions based on tissue symmetry in ultrasound,combined with an AI-assisted recognition method for intracranial hemorrhage.The accuracy,sensitivity,and specificity of this method were verified to evaluate its application value in improving diagnostic efficiency and patient outcomes.Results:The study results indicate that the automatic detection of abnormal cranial regions in ultrasound and the AI-assisted recognition method for intracranial hemorrhage have high sensitivity for hemorrhage detection,significantly enhancing the diagnostic efficiency of doctors for neonatal intracranial hemorrhage.Conclusion:Cranial ultrasound,being non-invasive and convenient,is suitable for screening cranial diseases in newborns without obvious symptoms.Introducing the technology of automatic detection of abnormal cranial regions in ultrasound and AI-assisted recognition of intracranial hemorrhage into the cranial volume ultrasound scanning system further helps doctors to detect neonatal intracranial hemorrhage early,effectively improving the early diagnosis and treatment rates for neonatal intracranial hemorrhage.