3D Tracking for OCT Image Fusion in Gynecological Diagnosis

描述(由申请人提供)：这项第一阶段SBIR提案的目标是开发一种导航和图像融合系统，以改进对宫颈癌和癌前病变的诊断。该系统将结合光学相干断层扫描(OCT)成像与传统阴道镜检查(有或没有乙酰白化)和/或荧光成像，以允许对可疑区域进行非常详细的形态分析。这两个免费的成像系统将通过实时三维空间跟踪像内窥镜一样的OCT成像探头进行组合。一旦知道了宫颈相对于3D跟踪系统帧的位置，并且已经记录了获取的宫颈图像(例如来自镜)，软件将持续更新宫颈图像上OCT探头的位置，提供宏观成像和OCT图像数据之间的配准，并为OCT操作员提供探索可疑区域的指导。同样的联合配准对于基于手动扫描的视频率OCT采集创建广域体积数据集也是有用的，我们也将检验这一假设。在这个项目中，我们将开发一个概念验证原型，它将在幻影中进行表征和测试，并在绵羊宫颈阴道模型中进行演示。公共卫生相关性：2008年，美国预计将诊断出大约11,070例浸润性宫颈癌，据估计，非浸润性宫颈癌(原位癌)的发病率约为[http://www.cancer.gov/cancertopics/types/cervical].的四倍2008年，美国预计将有大约3870名妇女死于宫颈癌。尽管与宫颈癌相关的死亡率在过去50年中大幅下降，但宫颈癌前病变的发病率却在上升，而且有人认为，如果不改进目前的筛查和诊断技术，与宫颈癌相关的死亡率可能会上升，特别是如果人类乳头状瘤病毒(HPV)血清阳性率继续上升的话。宫颈癌前期病变的早期发现可以对宫颈癌死亡率产生巨大影响，将五年存活率从70%提高到约100%。本项目致力于开发一种新的成像系统，以提高宫颈组织的诊断和定性。

DESCRIPTION (provided by applicant): The goal of this Phase I SBIR proposal is to develop a navigation and image fusion system for improved diagnosis of cervical cancers and pre-cancers. The system will combine optical coherence tomography (OCT) imaging with traditional colposcopy (with or without aceto-whitening) and/or fluorescence imaging to allow highly detailed morphological analysis of suspicious regions. These two complimentary imaging systems will be combined through real-time three-dimensional spatial tracking of a small endoscope-like OCT imaging probe. Once the location of the cervix is known relative to the frame of the 3D tracking system and an acquired cervical image (e.g. from colposcope) has been registered, software will continually update the position of the OCT probe on the cervical image providing registration between macro imaging and OCT image data and providing guidance to the OCT operator for exploration of suspicious regions. This same co-registration may also be useful for creating wide-field volume data sets based on manually scanned video-rate OCT acquisition, and we will test this hypothesis as well. During this project, we will develop a proof-of-concept prototype which will be characterized and tested in phantoms and demonstrated in an ovine cervicovaginal model. PUBLIC HEALTH RELEVANCE: In 2008, approximately 11,070 cases of invasive cervical cancer are expected to be diagnosed in the United States, and it is estimated that incidence of noninvasive cervical cancer (carcinoma in situ) is about four times as high [http://www.cancer.gov/cancertopics/types/cervical]. Approximately 3,870 women are expected to die from cervical cancer in the United States during 2008. Though mortality rates associated with cervical cancer have decreased dramatically over the last 50 years, the incidence of cervical precancers has increased, and it has been suggested that without improvement in current screening and diagnostic techniques, mortality rates associated with cervical carcinoma could increase, especially if human papilloma virus (HPV) seroprevalence continues to rise. Early detection of cervical precancerous lesions can have a dramatic impact on cervical cancer mortality rates, increasing five- year survival rates from 70% to about 100%. This project focuses on development of a novel imaging system for improving diagnosis and characterization of cervical tissue.