CASIA OpenIR  > 毕业生  > 博士学位论文
光学分子影像术中微创成像技术研究和临床应用
毛亚敏
学位类型工学博士
导师田捷
2018-05
学位授予单位中国科学院研究生院
学位授予地点北京
关键词光学分子影像 微创手术 内窥镜 近红外荧光成像 手术导航 肺癌
摘要 光学分子影像技术是成像领域近年来发展起来的新生力量,其通过分子探针标记病变区域或重要组织,基于光学成像连续动态的采集光学分子信号。作为一种客观评价病变组织的成像方式,光学分子影像技术为医生提供分子细胞水平信息,用于识别病变组织,使医生能够在手术过程中进行准确、客观、有效的判断,是一种术中实时在体的高灵敏度成像方式。
在微创手术中,如何进行精准的肿瘤定位和肿瘤边界划分是临床上面临的挑战性问题。伴随着电视辅助胸腔镜手术的普及,手动触诊病灶区域被限制,使得外科医生主要依赖于视觉检查。但是作为术中的主要成像方式,白光内窥成像缺乏肿瘤和正常组织之间的对比,微创手术亟需一种高灵敏的术中成像方式。相比于传统的白光观测,光学分子影像为微小病灶的识别提供了一种有力的观测工具,有潜力提升手术效果。本文立足于解决肺癌微创手术的实际临床应用问题,不仅在理论算法上对基于光学分子影像的术中微创成像方法进行了深入研究,还构建了一套基于光学分子影像的术中微创成像设备,并基于二者设计了针对肺癌微创手术的预临床和临床试验。论文的主要工作和创新点归纳如下:
  1. 提出一种基于光学分子影像的术中微创成像方法,即多通道光学实时成像方法,有效提高了术中成像的灵敏度。针对肿瘤微创手术难以实现多通道实时成像导航切除的关键科学问题,研究术中激发荧光分子成像与普通白光微创内窥成像的融合成像方法,结合小波变换和自适应阈值实现多光谱图像的精准融合, 同时进行了激发荧光成像的优化。在确保多通道光学动态成像实时融合(达到20 fps)的基础上,有效提高了成像灵敏度(能探测到0.01 μM浓度的indocyanine green,(ICG))。经过文献对比,该方法能够探测到的最小荧光探针浓度(ICG浓度)比已经发表的成像方法提高了10倍。
  2. 构建了一套基于光学分子影像的术中微创成像设备,通过设计多通道光路并优化多种光学元件,有效提高了成像设备的信背比。针对肿瘤微创手术对于成像设备高信背比和轻量化需求的关键技术问题,设计集光学激发、白光光学探测、近红外荧光探测于一体的多通道共享光路;在此基础上,优化和改造成像元件、分光元件、滤光元件等光学元件。从而在保证手持光学成像元件的体积(仅23*11*4 cm3)和重量(仅1.15 kg)轻量化设计的同时,提升了系统的光学成像信背比(对0.01 μM浓度 ICG,信背比≥5 dB)。经过多种成像仿体和动物实验证实该设备相较于德国Karl Storz公司(国际主流微创设备厂商)的商品化成像设备,在成像灵敏度上提高了两个数量级。
  3. 设计了基于光学分子影像术中微创成像方法和成像设备的临床转化试验,在肺癌微创术中实时精准导航胸外科医生进行肺癌切除手术。针对肺癌微创手术难以在术中精准定位微小多发肺结节、划分肺段边界的关键临床问题,设计了基于术中光学分子影像导航技术的肺癌结节探测和肺段切除术的临床试验方案。该方案在美国clinical trial上注册公开(注册号:NCT02611245)。通过36例肺癌患者的临床试验证实,光学分子影像术中导航的肺结节探测灵敏度达到88.7%,阳性预测率达到92.6%。此外在7例(19.4%)患者身上精确的探测到9个术前CT诊断未能发现的额外肺结节,并实现高精度的目标肺段成像。
本研究围绕光学分子影像的术中微创成像方法、成像设备和临床应用开展研究,三个方面相辅相成,串成一条主线。提出了一种基于光学分子影像的多通道成像实时融合方法,发表在光学工程类杂志Journal of Biomedical Optics上(国际光学工程学会会刊);构建了一套基于光学分子影像的术中微创成像设备,获得授权国家发明专利1项,申请美国发明专利1项,并获得第43届日内瓦国际发明博览会金奖;设计了基于光学分子影像的肺癌结节探测临床试验方案,在北京大学人民医院开展肺癌微创临床实验,取得了显著的临床效果,相关研究发表在临床外科类期刊European Journal of Cardio-Thoracic Surgery上(中科院外科分区二区)。此外,受邀在世界分子影像大会WMIC和国际光学工程学会年会SPIE Photonics West上做口头报告。相关研究成果表明了本人在博士期间科研工作的创新性,以及相关光学分子影像成像方法和设备的临床转化应用潜力。
其他摘要 In recent years, optical molecular imaging is a new developing technology in optical imaging field. It is based on molecular probe technology to mark the lesions or important tissues and optical imaging technology to collect fluorescence signals. As an imaging method to objectively evaluate diseased tissues, optical molecular imaging provides surgeons with information for identifying suspected lesions. It enables surgeons to perform more accurate, objective and effective judgment during surgery. Optical molecular imaging provides new ways for non-invasive imaging in vivo.
In the minimally invasive surgery, how to perform accurate tumor localization and tumor boundary identification is a clinically challenging problem. With the popularity of television-assisted thoracoscopic surgery, the manual palpation of the lesion area is limited, so that surgeons rely mainly on visual inspection. However, as the main imaging method during surgery, white-light endoscopic imaging lacks contrast between tumor and normal tissues, so that minimally invasive surgery needs a highly sensitive intraoperative imaging technique. Compared to traditional white-light observations, optical molecular imaging provides a powerful real-time imaging tool for the identification of micro tumors, which has the potential to improve surgical outcomes.
This research focuses on the study of multi-channel imaging method and prototype system construction based on intraoperative optical molecular imaging. It aims to solve practical application problems in minimally invasive surgery for lung cancer with in-depth studies of theory and algorithms, as well as a design of intraoperative minimally invasive imaging prototype system. The main work and innovations of the thesis are summarized as follows:
1. An intraoperative minimally invasive imaging method based on optical molecular imaging is proposed, namely multi-channel real-time optical imaging method, which effectively improves the sensitivity of intraoperative imaging. In order to solve the key scientific problems of lacking multi-channel real-time imaging technique for the minimally invasive surgery navigation, we studies the fusion imaging methods of intraoperative fluorescence molecular imaging and conventional white-light endoscopic imaging. The multi-channel imaging are realized by combining wavelet transform algorithm and adaptive threshold algorithm. Furthermore, we also studies the optimization of fluorescence molecular imaging. We not only ensure real-time fusion of multi-channel optical dynamic imaging (up to 20 fps), but also improve imaging sensitivity (indocyanine green (ICG) can be detected at a concentration of 0.01 μM). After comparison of the literature, the minimal detection concentration of ICG was about 10 times higher than that of the published imaging method.
2. An intraoperative minimally invasive imaging device based on optical molecular imaging was constructed. By designing multi-channel optical paths and optimizing various optical components, the signal-to-noise ratio of imaging devices was effectively improved. In order to solve the key technical issues of developing high signal-to-noise ratio and lightweight imaging equipment for minimally invasive surgery, we design a multi-channel optical path integrating optical excitation, white-light optical detection, and near-infrared fluorescence detection. Furthermore, optimization and retrofitting of imaging elements, spectroscopic elements, filter elements and other optical elements are implemented. As a result, we not only ensure the lightweight design of the handheld optical imaging device (only 23*11*4 cm3) and weight (1.15 kg only), but also improve the signal-to-noise ratio of the optical imaging system (for a concentration of 0.01 μM ICG, the signal-to-noise ratio ≥ 5 dB). A variety of imaging phantom experiments and animal experiments have confirmed that the device has improved imaging sensitivity by two orders of magnitude compared to commercial imaging device from the Karl Storz (a leading international endoscopic manufacturer).
3. Clinical transformation experiments based on minimally invasive imaging methods and imaging devices were designed to accurately navigate the minimally invasive lung cancer surgery. In order to solve the key clinical problems of multiple pulmonary nodules detection and segment boundaries identification, we design clinical trials of lung cancer nodule detection and segmentectomy based on intraoperative optical molecular imaging navigation technology. This program is registered in American clinical trial database (registration number: NCT02611245). This study enrolled 36 patients, the detection sensitivity of pulmonary nodules with optical molecular imaging was 88.7% and the positive predictive rate was 92.6%. In addition, the application of optical molecular imaging method led to the detection of 9 additional nodules which were missed using traditional detection methods (1mm computed tomography scan and white-light thoracoscopic exploration) in 7 patients (19.4%).
This study focuses on the intraoperative minimally invasive imaging method, device, and clinical applications based on optical molecular imaging method. A multi-channel real-time optical imaging method was proposed, which was published in the Journal of Biomedical Optics; An optical molecular imaging equipment was constructed, which was awarded one national invention patent, applied for one American invention patent, and won the gold medal at the 43rd International Exhibition of Inventions of Geneva; A clinical transformation experiments based on optical molecular imaging was designed to accurately navigate the minimally invasive lung cancer surgery in the People's Hospital of Peking University and achieved remarkable clinical results, which was published in the European Journal of Cardio-Thoracic Surgery. In addition, I was invited to give an oral report at WMIC and SPIE Photonics West. These research results have demonstrated the innovation of my research work during the Ph.D. period and the potential for clinical transformation of optical molecular imaging methods and device.
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/21517
专题毕业生_博士学位论文
作者单位中国科学院自动化研究所
推荐引用方式
GB/T 7714
毛亚敏. 光学分子影像术中微创成像技术研究和临床应用[D]. 北京. 中国科学院研究生院,2018.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
光学分子影像术中微创成像技术研究和临床应(4938KB)学位论文 暂不开放CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[毛亚敏]的文章
百度学术
百度学术中相似的文章
[毛亚敏]的文章
必应学术
必应学术中相似的文章
[毛亚敏]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。