目的:探讨甲状腺结节良恶性的临床诊断中应用人工智能辅助评分系统联合超声弹性成像诊断的人工智能辅助评分(AI-assisted score,AIAS)评分及应变率比值(ratio of strain rate,SR)。方法:选取130例2020年10月—2021年6月期间于安宁市第...目的:探讨甲状腺结节良恶性的临床诊断中应用人工智能辅助评分系统联合超声弹性成像诊断的人工智能辅助评分(AI-assisted score,AIAS)评分及应变率比值(ratio of strain rate,SR)。方法:选取130例2020年10月—2021年6月期间于安宁市第一人民医院接受甲状腺结节诊断的患者作为本次的研究对象,所有患者均行AIAS辅助评分系统、超声弹性成像、AIAS辅助评分系统联合超声弹性成像诊断,分析并对比三者的AIAS评分与SR。结果:经病理检查确认,130例患者中共308个结节,其中恶性病变185个,良性病变123个;经AIAS辅助评分系统检查发现共273个结节,其中恶性病变159个,良性病变114个;经超声弹性成像检查发现共280个结节,其中恶性病变163个,良性病变117个;经AIAS辅助评分系统联合超声弹性成像检查发现共301个结节,其中恶性病变180个,良性病变121个。AIAS辅助评分系统检查、超声弹性成像检查、联合检查的灵敏度分别为85.95%、88.11%、97.30%,特异度分别为92.68%、95.12%、98.37%,准确度分别为88.64%、90.91%、97.73%;恶性结节的AIAS评分与SR分别为(0.77±0.16)分、(6.21±1.66),均较良性结节的(0.46±0.17)分、(2.58±0.74)显著更高(P<0.05)。分析两种类型结节的图像特征可知,甲状腺右侧叶可见一实性低回声结节,形状不规则,纵横比>1,其内可见多个细点状钙化;超声图像AIAS辅助系统针对图一评分结果显示:评分为0.86,该结节偏恶性,Ti-RADS分类为4C类。结论:以AIAS辅助评分系统+超声弹性成像联合检查的方式诊断良恶性甲状腺结节,能够较好地诊断出甲状腺结节的性质,与病理检查结果具有较高的一致性,值得在临床中推广与使用。展开更多
This paper examined the computer-assisted TRP （Teaching Resource Program）, which is an intelligent writing program. Different from previous research, this study emphasized TRP＇s integrative and holistic functions o...This paper examined the computer-assisted TRP （Teaching Resource Program）, which is an intelligent writing program. Different from previous research, this study emphasized TRP＇s integrative and holistic functions of intelligent analysis, intelligent scoring, and intelligent comment on writing pieces. In what follows, this paper conducted a study of intelligent essay scoring system related to intelligent analysis, intelligent scoring, and intelligent comment of the writing pieces. Also, this paper provided a literature review of existing computer-assisted essay scoring systems and analyzed the advantages and disadvantages of them. Based on these findings in our research, we can conclude that the integrative and holistic computer-assisted TRP can effectively analyze, score, and comment essays, consequently enhancing the writing performance. Some useful implications from this study include that the computer-assisted TRP should be further integrated with writing learning and instruction展开更多
文摘目的:探讨甲状腺结节良恶性的临床诊断中应用人工智能辅助评分系统联合超声弹性成像诊断的人工智能辅助评分(AI-assisted score,AIAS)评分及应变率比值(ratio of strain rate,SR)。方法:选取130例2020年10月—2021年6月期间于安宁市第一人民医院接受甲状腺结节诊断的患者作为本次的研究对象,所有患者均行AIAS辅助评分系统、超声弹性成像、AIAS辅助评分系统联合超声弹性成像诊断,分析并对比三者的AIAS评分与SR。结果:经病理检查确认,130例患者中共308个结节,其中恶性病变185个,良性病变123个;经AIAS辅助评分系统检查发现共273个结节,其中恶性病变159个,良性病变114个;经超声弹性成像检查发现共280个结节,其中恶性病变163个,良性病变117个;经AIAS辅助评分系统联合超声弹性成像检查发现共301个结节,其中恶性病变180个,良性病变121个。AIAS辅助评分系统检查、超声弹性成像检查、联合检查的灵敏度分别为85.95%、88.11%、97.30%,特异度分别为92.68%、95.12%、98.37%,准确度分别为88.64%、90.91%、97.73%;恶性结节的AIAS评分与SR分别为(0.77±0.16)分、(6.21±1.66),均较良性结节的(0.46±0.17)分、(2.58±0.74)显著更高(P<0.05)。分析两种类型结节的图像特征可知,甲状腺右侧叶可见一实性低回声结节,形状不规则,纵横比>1,其内可见多个细点状钙化;超声图像AIAS辅助系统针对图一评分结果显示:评分为0.86,该结节偏恶性,Ti-RADS分类为4C类。结论:以AIAS辅助评分系统+超声弹性成像联合检查的方式诊断良恶性甲状腺结节,能够较好地诊断出甲状腺结节的性质,与病理检查结果具有较高的一致性,值得在临床中推广与使用。
文摘This paper examined the computer-assisted TRP （Teaching Resource Program）, which is an intelligent writing program. Different from previous research, this study emphasized TRP＇s integrative and holistic functions of intelligent analysis, intelligent scoring, and intelligent comment on writing pieces. In what follows, this paper conducted a study of intelligent essay scoring system related to intelligent analysis, intelligent scoring, and intelligent comment of the writing pieces. Also, this paper provided a literature review of existing computer-assisted essay scoring systems and analyzed the advantages and disadvantages of them. Based on these findings in our research, we can conclude that the integrative and holistic computer-assisted TRP can effectively analyze, score, and comment essays, consequently enhancing the writing performance. Some useful implications from this study include that the computer-assisted TRP should be further integrated with writing learning and instruction
