目的:利用人工神经网络模型对鼻咽癌计划进行剂量预测,根据剂量预测值建立计划质量测度标准(PQM),实现放疗计划个性化定量评估.方法:回顾性分析于福建省肿瘤医院进行放射治疗的鼻咽癌计划114例.提取25项危及器官(OAR)与靶区之间的几何空间关系特征,训练(81例)得到基于人工神经网络的剂量预测模型并测试(23例)验证其准确性.分别用基于剂量限值建立PQM和基于剂量预测值建立PQM两种评估方法对10例临床通过计划进行定量评估,讨论两种评估方法的合理性.结果:11项鼻咽癌主要OAR剂量学指标,预测值与实际值的剂量相关总体平均差值为(-0.07±4.55)Gy,体积相关总体平均差值为-1.06％±3.80％,预测准确性可达90％.10例临床通过的鼻咽癌计划,基于剂量限值建立PQM方法对病例4评估为不合格,其余病例合格.基于剂量预测值建立PQM方法对病例9评估为不合格,其余病例合格.结论:剂量预测引导的鼻咽癌计划质量定量评估方法可以反映计划是否存在继续优化的空间,且该评估方法克服了剂量限值未考虑病例特异性的缺陷,能更科学合理地对放疗计划进行定量评估.

目的:利用循环生成对抗网络模型(CycleGAN)进行锥形束CT(CBCT)图像迁移,生成伪CT(sCT)图像,从而实现CBCT图像的HU值矫正.方法:回顾性分析在福建省肿瘤医院行放射治疗的鼻咽癌患者39例,所有患者均接受临床CT与CBCT扫描.以CBCT图像为基准,采用刚性配准算法对临床CT和CBCT进行配准,获得重采样计划CT(pCT).经阈值分割及形态学处理获取配对影像的外轮廓内部区域作为掩膜,对配对影像进行掩膜操作及归一化预处理.建立CycleGAN神经网络,训练sCT生成模型.基于体素点计算平均绝对误差(MAE)和平均误差(ME),用于比较测试集sCT与pCT之间的差异.结果:测试集的sCT图像与pCT图像相比较,在体外轮廓内的MAE和ME分别为(99.00±15.37)HU和(-24.00±12.64)HU;软组织区域的MAE和ME分别为(48.00±7.45)HU和(-7.00±8.96)HU.结论:CycleGAN能修正CBCT图像的HU值,迁移生成的sCT图像具有与pCT图像近似的HU值及平滑性,可用于放射治疗剂量计算.

目的 利用3D打印技术制作头颈部肿瘤放射治疗的个性化头枕,研究制作过程中的技术难点并进行剂量学测试.方法 利用头颈部肿瘤患者的CT图像在Pinnacle3计划系统中生成患者人头外轮廓及个性化头枕的内外轮廓;通过计算机程序把轮廓生成STL格式的头枕文件,在Autodesk软件结合生成的类人头模型加入带有电离室平衡帽的空腔;打印出头模型和头枕,选用2019月1—5月医院收治的15例头颈部放射治疗患者的临床处方在模体中进行调强计划验证,比较标准头枕与自制3D打印头枕的剂量学差异.结果 在设定40％聚乳酸(PLA)填充时,测量值与计划计算值误差在3％左右.结论 利用3D打印技术可以制作个性化的高度符合的头枕,在40％PLA填充设置时可以得到与标准头枕相近似的结果,符合临床的精度要求.

目的 基于治疗过程中的摆位误差进行六维模拟矫正,评估鼻咽癌调强放射治疗过程中的剂量误差变化.方法 选取2018年9月在福建省肿瘤医院行调强放射治疗的鼻咽癌患者5例,为每例患者在医院鼻咽癌首次摆位误差数据库(636例)中随机提取33次摆位误差;根据这组摆位误差在Pinnacle3计划系统中,利用误差六维校正方法,进行每一治疗分次的模拟,并自编程序进行剂量总合成;评估治疗分次剂量、总剂量与相应的计划剂量的差异.结果5例患者误差校正后的靶区总剂量差异较小(基本<50 cGy),危及器官中脊髓、脑干有明显的剂量增加,最大可达到254 cGy和346 cGy;腮腺除了1例有剂量减少外,其他均有增加,最大达到102 cGy.结论 基于福建省肿瘤医院摆位误差数据库进行摆位误差校正后剂量合成,所有靶区总剂量的变化幅度均较小,危及器官剂量差异较大,应充分评估脊髓、脑干、腮腺的剂量,控制摆位误差,减少对危及器官的影响.

目的:研究宫颈癌容积旋转调强放疗过程中,利用图像引导技术评估膀胱与直肠因平移摆位误差因素对受照剂量的影响.方法:入组2017年2～4月的25例行容积旋转调强放疗的宫颈癌病人.在计划CT图像和首次摆位的锥形束CT(CBCT)图像中勾画出膀胱和直肠.然后根据首次摆位误差,对膀胱和直肠在前后、左右、头脚方向上进行相应的平移.分别比较计划设计与加入平移摆位误差下的膀胱和直肠剂量差异,并用配对t检验分析差异.结果:在计划CT与CBCT图像上勾画出的直肠体积为(33.0±16.0)和(21.8±16.5)cm3,膀胱体积为(387.1±222.3)和(227.8±107.3)cm3.在计划CT图像与考虑平移摆位误差的CBCT图像上,直肠的V50为(17.27±13.42)％和(23.9±26.58)％,膀胱的V50为(33.59±12.2)％和(44.27±22.61)％,直肠的最大剂量为(5 324±383)和(5 370±441)cGy,膀胱的最大剂量为(5 547±365)和(5 513±328) cGy.直肠和膀胱的V50均有统计学意义(t=-3.60,P＜0.01;t=-2.98,P＜0.01),直肠和膀胱的最大剂量均无统计学意义(t=-1.18,P=0.20;t=0.66,P=.51).结论:宫颈癌放射治疗中,膀胱和直肠在计划CT和首次治疗时的体积与剂量均有较大差异,故应该多次采集图像来评估剂量分布,才能准确评价放射治疗的真正疗效.

Objective To realize the automatic identification of setup error data in the treatment of nasopharyngeal carcinoma by adjusting the image feature matching method. Methods Based on the Matlab platform, the error image and the inpatient number of the patient were automatically programmed according to the image acquisition tool (CBCT) before IMRT. Into the group selected from October 2017 to November, our hospital received treatment of nasopharyngeal carcinoma in 45 cases of newly diagnosed patients. Perform program stability and accuracy testing. Results The results of the program run through a manual inspection, the hospital number and positioning error of 100％ identified. The average run time per case is about 0.25 seconds. Conclusion Using the program compiled in this study, data such as setup error and patient's hospitalization number can be quickly and accurately obtained. Fully meet the clinical data record requirements.

目的 根据螺旋断层放射治疗(TOMO)过程中每日进行的兆伏(MV)级CT扫描的图像研究鼻咽癌患者外体形变化和摆位误差变化的相关性.方法 入组10例行TOMO的鼻咽癌患者,每例患者每次治疗前进行MV级CT扫描,共有千伏(KV)级CT图像10组,MV级图像322组,然后进行图像配准获得摆位误差.分别在MV级CT图像和模拟CT图像上勾画出患者的头部外轮廓,并将所有MVCT上的外轮廓通过配准映射至KVCT图像上,分别计算相应重叠的体积,根据体积数据分析体形变化与摆位误差的相关性.结果 在放射治疗整个过程中,10例患者中7例有消瘦的情况发生,主要开始在治疗15～20次之间.相关分析中,10例患者中仅3例Y轴平移有相关性,患者编号为1,3,10,相关系数为0.622,-0.76,-0.53(P<0.05).结论 从螺旋断层放射治疗前到放射治疗15～20次期间,鼻咽癌患者头部有明显体积变化,但体积变化与摆位误差没有强相关性.

Objective To investigate the precision of full six-degree target shift corrections using the ArcCHECK system.Metbods Fourteen patients receiving intensity-modulated radiotherapy (IMRT) for nasopharyngeal carcinoma (NPC) in Fujian Medical University Cancer Hospital from May to September,2015 were selected.The first treatment setup errors were obtained using cone-beam computed tomography.The setup errors were simulated in ArcCHECK,and the full six-degree target shift corrections was used to correct the errors.The plans without and with setup errors and the plan with corrected setup errors were taken.The paired t-test was used to compare dose to agreement (DTA) and Gamma passing rates between the plan without setup errors and the plan with setup errors and plan with corrected setup errors.Results The DTA and Gamma passing rates were (96.76± 1.57)％ and (98.35±0.92)％ for the plan without setup errors,(59± 21.42) ％ and (62.86± 21.63) ％ for the plan with setup errors,and (91.41± 4.82) ％ and (94.11±4.33)％ for the plan with corrected setup errors.There were significant differences between the plan without setup errors and the plan with setup errors and plan with corrected setup errors in DTA passing rate (t=6.64 and 5.13,both P＜0.05) and Gamma passing rate (t=6.15 and 4.19,both P＜0.05).Conclusions The full six-degree target shift corrections can be used in IMRT for NPC,with good results in correcting setup errors and improving the precision for IMRT dose distribution.

目的 研究局部晚期鼻咽癌患者在新辅助化疗联合同步放化疗过程中肿瘤及腮腺解剖形态变化规律.方法 选择2014年5—9月收治的15例初诊鼻咽癌患者.患者均接受新辅助化疗联合同步放化疗,放疗采用调强放疗.新辅助化疗前(P1点)、后(P2点)、放疗剂量至40 Gy(P3点)及放疗结束同日(P4点)分别行4次定位CT扫描,分别勾画出肿瘤靶区及危及器官.获取各组CT图像中肿瘤和腮腺的体积,对比不同组之间的变化.结果 相对于P1点体积,鼻咽原发灶和颈部转移淋巴结体积在P2,P3,P4点分别缩小至(55.74±9.27)%及(48.96±26.17)%,(37.24±12.37)%及(21.01±13.09)%,(27.22±6.88)%及(13.33±10.89)%.腮腺相对于P1点,在P2点没有明显变化,P3,P4点分别缩小至(67.61±9.70)%,(67.10±10.58)%.结论 对鼻咽原发灶及颈部转移淋巴结通过放疗前勾画区域进行计划设计对于肿瘤患者有较大收益.放疗中腮腺体积也明显缩小,在放疗累计剂量达40 Gy时调整放疗计划能够使腮腺获益较大.

目的:探析护理干预对肺癌放化疗患者癌因性疲乏的影响.方法:选取笔者所在科接收治疗的肺癌放化疗患者84例为研究对象,按照入院时间顺序分两组,每组42例.予以对照组常规护理,予以观察组护理干预,对比两组患者的临床护理效果.结果:观察组患者的癌性疲乏改善情况优于对照组,差异有统计学意义(P<0.05);观察组生活质量改善情况优于对照组,差异有统计学意义(P<0.05).结论:对肺癌化疗患者应用护理干预,可有效改善癌性疲乏现象,使患者的生活质量与日常生活水平得到提高.

目的:观察鼻咽癌调强放疗患者口腔黏膜反应采用护理干预的效果.方法 :将2015年1-6月到我院治疗的80例鼻咽癌调强放疗患者随机分为对照组和观察组,各40例.对照组采用常规护理的方法,观察组在常规护理的基础上加用护理干预,观察两种护理方法对鼻咽癌调强放疗患者口腔黏膜反应方面的治疗效果.结果 :观察组患者的口腔黏膜反应明显轻于对照组,差异具有统计学意义(P<0.05).结论 :对鼻咽癌调强放疗患者实施护理干预,能帮助患者达到预定目标,提高患者治疗依从性,帮助患者恢复健康.

ObjectiveDuring registration of cone-beam CT images and planning CT, this paper investigated the effect of setup errors for the different registration-algorithms and multiple regions-of-interest for nasopharyngeal carcinoma for intensity-modulated radiotherapy (IMRT).Methods 20 nasopharyngeal carcinoma patients were enroled for IMRT. Each patient received Cone-beam CT (CBCT) during the first fraction of treatment. The registration-algorithms were selected bone and gray registration. There were three kinds of regions-of-interest, including between brow ridge to top teeth, between bottom teeth to neck and whole region. Six methods were selected to match CBCT and planning CT and get setup errors. The differences to six methods were evaluated.Results There were statisticaly significant differences between the whole of bone algorithm to the whole of gray algorithm in y,Rx,Ry (P<0.05). There were statisticaly significant differences between the whole of bone algorithm to the bottom teeth of gray algorithm in x,z,Rx (P<0.05). Conclusion There were statisticaly significant differences when the registration-algorithms included neck. We should pay attention to the neck setup errors for nasopharyngeal carcinoma radiotherapy and improve the precise of treatment for neck region.

ObjectiveAccording different CT images to analysis the changes of volume for head and neck, parotid gland and gross target during treatment for nasopharyngeal carcinoma.Methods 15 patients were enroled for nasopharyngeal carcinoma during treatment. For each patients, there were 4 serials CT being scanned, including previously treatment, previously radiotherapy, the dose were 45~53 Gy and after radiotherapy. We contoured the body of head and neck, neck, target and parotid gland, then calculated the volume. According the changes of volumes, we evaluated the largest change phase and hope nursing intervention decrease the changes..Results During the dose were 45~53 Gy, there are significant differences for head and neck, neck, target and parotid to previously radiotherapy. The t values were 4.26, 13.23, 13.23, 5.86, 3.93, 5.93(P<0.05) for neck, left parotid gland, right parotid gland, gross target volume, left neck gross target volume and right neck gross target volume.Conclusion During delivered 45~53 Gy,there were the large changes of volume to previously radiotherapy for region of interest (ROIs) for nasopharyngeal carcinoma treatment. In clinical practice, nursing intervention to decrease the changes is important for precise radiotherapy. Otherwise, we should take adaptive radiotherapy.

Objective: To investigate the frequency of setup correction for cervical cancer during image-guided radiotherapy ( IG-RT) in intensity-modulated radiotherapy ( IMRT) .Methods: 17 patients with cervical cancer were treated by IMRT and IGRT.Cone-beam CT ( CBCT) were acquired at the first, second and third fraction.CBCT and planning CT were registered and got the setup errors.The errors were done the statistic analysis using paired t-test.Results:Using absolute value of setup errors, The aver-age value of y-axis shift ( superior-inferior direction) was obviously larger than other axes.There are no significant in statistics for any two groups of setup error.Conclusion:We must pay attention to the shift of y-axis for cervical cancer during image-guided ra-diotherapy.The setup error was at random each fraction.So we must scan more times using CBCT and correct setup errors to get accu-rate treatment.

目的：基于已有的头部和颈部为中心的摆位误差数据，研究鼻咽癌不同段的靶区扩边方案。方法：入组鼻咽癌患者20例，每位患者扫描CBCT的次数为6～9次，配准获得摆位误差；同时入组100例以颈部为中心的患者，每人仅扫描1次CBCT，获取患者的摆位误差。依据这些误差计算出头部和颈部的肿瘤扩边的大小。这两组病例均使用调强放射治疗。结果：头部肿瘤靶区扩边 X， Y， Z 方向各为：1.71mm，1.93mm，1.32mm。颈部靶区扩边X， Y， Z方向分别为：3.62mm，4.26mm，3.74mm。结论：头部比颈部的肿瘤PTV需要的扩边明显要小。鼻咽癌在治疗过程中，鼻咽部（头部）肿瘤扩边2 mm可以达到治疗要求。颈部需要扩边5 mm才能达到控制肿瘤的摆位误差的目的。

Objective: To investigate the correlation of patient ’ s weight and changes of parotid glands between before and after in-tensity-modulated radiotherapy ( IMRT) for nasopharyngeal carcinoma.Methods:57 patients with nasopharyngeal carcinoma were treated by IMRT.Computed tomography ( CT) was acquired before and after treatment.The each slide of parotid glands was contoured independently by physician.Volume changes of parotid glands , DICE index and weight of patient were done the statistic analysis.The correlation of the indexes were also studied.Results:There were significant in statistics to parotid glands volume.Right and left parot-id glands both were t =2.00 (P<0.01) .There were significant in statistics to the weight of patients.That was also t=2.00 (P<0.01 ) .When the change rates of patient ’ s weight was≤5%; ≥5%and <10%; ≥10%, the mean change rates of right and left volume were 31.34%, 26.54%;34.76%, 34.54%;36.32%, 28.59%, respectively.When the change rates of patient ’ s weight was ≤5%; ≥5%and≤10%; ≥10%, the mean DICE index of right and left volume were 0.662, 0.642;0.597, 0.593;0.56, 0.562.Conclusion: The patient ’ s weight and the volume and DICE index of parotid glands have some correlation during IMRT treatment.Physician can replan using the change rate of patient ’ s weight as a guide.

目的:通过量化的方法研究调强放射治疗前后的腮腺体积和空间位置的变化.方法:入组40例经过调强放射治疗后的鼻咽癌患者.治疗前后各扫描获得一组CT图像.临床医生在这两组图像中分别勾画出腮腺.通过Pinnacle自带功能把治疗前的数据映射至治疗后的CT图像上.统计分析腮腺治疗前后的体积变化.同时使用体积差异法和DICE相似度法来评价腮腺空间位置的变化情况.结果:左右腮腺前后体积对比有显著意义,其中左腮腺t=1 1.32 (P＜0.01),其中右腮腺t=12.41(P＜0.01).使用体积差异法评价,左腮腺平均变化为(34.62±13.14)％,体积缩小范围3.01％～58.93％;右腮腺平均为(34.3±12.54)％,体积缩小范围为5.61％～59.39％.DICE参数评价,左腮腺平均为0.59±0.11,变化范围为0.35～0.73;右腮腺DICE值平均为0.66±0.11,变化范围为0.29～0.83.结论:放射治疗后腮腺相对于治疗前有明显的缩小,空间位置有较大的变化.放射治疗中采用量化标准评估腮腺体积的变化,医生可以及时修改放射治疗计划来避免腮腺超过剂量限值.

目的 通过鼻咽癌治疗前后的靶区变化来研究原发肿瘤残留的规律.方法:入组49例经过调强放射治疗后的鼻咽癌病例.治疗前后各扫描一组CT图像和MRI图像一组.治疗接近结束的时均发现有肿瘤残留.临床医生分别在两组CT图像中勾画出鼻咽部原发肿瘤的体积.在计划Pinnacle计划系统中统计出靶区前后的体积,来初步研究其规律.结果:其中按分期肿瘤退缩分别为T1,T2,T3,T4分别为0.95,0.55,1.32,2.18 cm3/天,相应的退缩率为3.09,1.79,4.44,7.00％/天.肿瘤总得平均缩小为1.34 cm3/天.结论:肿瘤的残留与分期的肿瘤的分期相关性不明显,肿瘤在放射治疗中有明显的缩小,临床中应注意调整方案来增加肿瘤残留部分的剂量,以期更好提高治愈率及减少放射治疗的副反应.

恶性淋巴瘤是比较常见的来源于淋巴造血组织的恶性肿瘤.我科自2000年以来,有8例单克隆性B细胞淋巴瘤患者单纯采用美罗华(Rituximab)治疗一个疗程后肿瘤明显缩小,病情得到控制.由于美罗华有不同程度的毒副作用,现将护理体会总结如下.