本文選題：pNENs + CT�。� 參考：《濟(jì)南大學(xué)》2017年碩士論文

【摘要】：目的:根據(jù)對(duì)pNENs的CT表現(xiàn)的分析,探討其不同影像學(xué)特點(diǎn)與病理分級(jí)的相關(guān)性,評(píng)估其在臨床診斷、治療及預(yù)后方面的參考價(jià)值。方法:回顧性地研究經(jīng)病理證明為pNENs的31例病人的影像及臨床資料,查看并記下腫瘤的部位、大小、有無鈣化、有無囊變壞死、有無淋巴結(jié)轉(zhuǎn)移、有無遠(yuǎn)處器官轉(zhuǎn)移、平掃及多期增強(qiáng)CT掃描數(shù)值,以及確定病理分級(jí)(包含Ki-67和每10個(gè)高倍視野中的核分裂象數(shù)目,二者病理分級(jí)結(jié)果出現(xiàn)差異時(shí),以分級(jí)高者為準(zhǔn))。全部的觀察結(jié)果均運(yùn)用SPSS17.0版軟件進(jìn)行統(tǒng)計(jì)學(xué)分析:(1)首先分別把不同分級(jí)的平掃加三期增強(qiáng)的CT值用(均數(shù)±標(biāo)準(zhǔn)差)的方法來標(biāo)示,以此對(duì)增強(qiáng)的類型進(jìn)行總體的量化,研究不同分級(jí)的強(qiáng)化方式的不同;(2)比較各個(gè)指標(biāo)的意義以及價(jià)值時(shí),連續(xù)變量采用單因素方差分析,分類變量采用Kruskal-Walls檢驗(yàn)。研究時(shí)首先在三組之間比較各個(gè)指標(biāo)與病理分級(jí)的相關(guān)性,然后在組間兩兩比較比較上述指標(biāo)與病理分級(jí)間的關(guān)系,從而分別篩選出有統(tǒng)計(jì)學(xué)意義的指標(biāo)。最后用ROC曲線和“尤登指數(shù)”來計(jì)算有意義的研究因素的診斷有效率,其指標(biāo)包含ROC曲線下面積、最佳閾值及其敏感性和特異性。結(jié)果:(1)通過對(duì)平掃及多期增強(qiáng)掃描的CT值(均數(shù)±標(biāo)準(zhǔn)差)比較分析,隨病理分級(jí)的增加,總體的強(qiáng)化程度逐漸減弱;(2)經(jīng)統(tǒng)計(jì)學(xué)分析,三組間有顯著性差異的統(tǒng)計(jì)學(xué)指標(biāo)分別為Ta(P=0.008)、Tv(P=0.012)、Tap(P=0.009)、Tvp(P=0.013)、Tav(P=0.024)、有無淋巴結(jié)轉(zhuǎn)移(P=0.001)、長(zhǎng)徑是否大于3cm(P=0.032)、長(zhǎng)徑是否大于4cm(P=0.014)及CT值到達(dá)峰值的時(shí)間(P0.001);(3)經(jīng)過組間的兩兩比較,可用于區(qū)分G1/2與G3的指標(biāo)為Ta、Tv、Tap、Tvp及強(qiáng)化CT值到達(dá)峰值時(shí)間,可用于區(qū)分G1與G2的指標(biāo)為病變長(zhǎng)徑的大小4MX≥3、MX≥4,另外Tav和Tvd還可用于區(qū)分G1與G3;(4)經(jīng)過ROC曲線分析,有統(tǒng)計(jì)學(xué)意義意義的指標(biāo)為4MX≥3(AUC=0.708,P=0.048)、MX≥4(AUC=0.744,P=0.021)、Ta(AUC=0.742,P=0.022)、Tav(AUC=0.727,P=0.031)。經(jīng)過“尤登指數(shù)”分析,Ta的最佳臨界值為107HU(sen=0.667、spe=0.812)、Tav為9HU((sen=0.667、spe=0.812)。結(jié)論:Ta、Tav、4MX≥3、MX≥4是對(duì)病理分級(jí)最為有效的指標(biāo),且AUC均大于0.70,有一定的診斷準(zhǔn)確性,其中準(zhǔn)確性最高的為MX≥4。此外,Ta的最佳閾值為107HU、Tav的最佳閾值為9HU,二者的敏感性和特異性相同,分別為0.667和0.821。因此我們可以認(rèn)為Ta、Tav、4MX≥3、MX≥4對(duì)于判定腫瘤的分級(jí)、指導(dǎo)治療以及預(yù)測(cè)預(yù)后具有一定的價(jià)值。其中強(qiáng)化方式主要區(qū)別低級(jí)別腫瘤(G1/2)與高級(jí)別腫瘤(G3),而大小是區(qū)別G1與G2的主要手段。但是這四個(gè)有意義指標(biāo)的價(jià)值也有一定的限度,因此在判定腫瘤的分級(jí)時(shí)應(yīng)該綜合利用上述指標(biāo),提高診斷結(jié)果的正確性,為治療的有效性提供一定的保證。
[Abstract]:Objective: to study the correlation between different imaging features and pathological grade of pNENs, and to evaluate its reference value in clinical diagnosis, treatment and prognosis. Methods: the imaging and clinical data of 31 patients with pNENs proved by pathology were studied retrospectively. The location, size, calcification, cystic necrosis, lymph node metastasis and distant organ metastasis were observed and recorded. The pathological grade (including Ki-67 and the number of mitotic images per 10 high-power field) was determined by plain scan and multi-phase enhanced CT scan. All the observation results were analyzed statistically by using SPSS17.0 software. (1) first of all, the CT values of different grades of plain scan plus phase 3 enhancement were marked by the method of (mean 鹵standard deviation) respectively, so as to quantify the types of enhancement as a whole. To compare the significance and value of each index, single factor analysis of variance was used for continuous variables and Kruskal-Walls test was used for classifying variables. The correlation between each index and pathological grade was compared among the three groups, and then the relationship between the above indexes and the pathological grade was compared between the two groups. Finally, the diagnostic efficiency of meaningful research factors was calculated by using ROC curve and "Uden index", which included the area under the ROC curve, the optimum threshold, its sensitivity and specificity. Results by comparing and analyzing the CT value (mean 鹵standard deviation) of plain scan and multiphase enhanced scan, the enhancement degree of the whole was gradually weakened with the increase of pathological grade. There were significant differences among the three groups. The statistical indexes of the three groups were as follows: (1) TapP0. 0012 (TapP0. 009), 0. 013 (0. 013), 0. 024 (0. 024), whether the long diameter was greater than 3 cm P0. 032, whether the long diameter was greater than 4 cm P0. 014) and the time when the CT value reached the peak value (P 0. 001) was compared between the two groups. The indexes that can be used to distinguish G1 / 2 from G3 are TapTVP and the peak time of enhanced CT value. The index of distinguishing G1 from G2 is 4MX 鈮，

本文編號(hào)：1936240