【摘要】 目的评价乳腺超声人工智能设备,对乳腺结节良恶性的诊断价值。方法采集具有明确病理学诊断结果的乳腺结节超声图像400张,其中乳腺良性结节200张,乳腺恶性结节200张,运用超声人工智能对400张超声图像进行诊断,并对诊断结果进行统计分析。结果乳腺超声人工智能的灵敏度为96.06%、特异度为97.46%,具有良好的真实性,与病理结果一致性程度极好,Kappa值为0.94,具有极好的可靠性。结论超声人工智能有利于避免超声诊断存在的主观性缺陷,一定程度上实现了超声诊断的量化及标准化,值得临床推广。更多还原

【Abstract】 Objective The authors of this article have led the development of breast ultrasound artificial intelligence equipment,this article is aimed at studying the diagnostic value of breast and nodules.Methods 400 ultrasound images of breast nodules with clear pathological diagnosis were collected,including 200 benign nodules of the breast and 200 malignant nodules of the breast.Ultrasound artificial intelligence was used to diagnose 400 ultrasound images,and the diagnosis results were statistically analyzed.Results The sensitivity of breast ultrasound artificial intelligence is as high as 96.06%,the specificity is 97.46%,It has good authenticity and excellent consistency with pathological results.The Kappa value is 0.94,which has excellent reliability.Conclusions Ultrasonic artificial intelligence has a high accuracy rate for the diagnosis of benign and malignant breast nodules,which is conducive to avoiding the subjective defects of ultrasound diagnosis.To some extent,it has realized the quantification and standardization of ultrasound diagnosis,which is worthy of clinical promotion.更多还原