Prostate Boundary Detection in Ultrasound Images

超声图像中的前列腺边界检测

• 批准号: 6549120
• 负责人: SAYAN PATHAK
• 金额: $ 37.45万
• 依托单位: INSIGHTFUL CORPORATION
• 依托单位国家: 美国
• 财政年份: 2000
• 资助国家: 美国
• 起止时间: 2000-09-01 至 2004-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6549120
• 关键词: bioimaging /biomedical imaging; biomedical automation; clinical research; computer assisted diagnosis; computer graphics /printing; computer program /software; diagnosis design /evaluation; human data; image processing; male reproductive system disorder diagnosis; prostate disorder; rectum /anus; ultrasound therapy

DESCRIPTION (provided by applicant): The goal of this proposal is to improve the effectiveness of transrectal ultrasound (TRUS) in diagnosing and treating prostate diseases. Prostate boundaries in TRUS images provide the anatomical description needed for diagnosis, for treatment planning, and for follow-up. Our proposed research enables computer aided prostate delineation using "just-enough interaction" with the expert. In phase I, we have embedded several segmentation algorithms within a software environment allowing interaction with the physicians. The variability in (semi-)automated prostate delineation compared to manual delineation was reduced by half. Also, the delineation time compared to manual outlining was reduced by a factor of 2. Based on the feedback from several companies and end-users, who have shown interest in our technology, we find that in the past year, emerging intraoperative clinical procedures and post-treatment monitoring techniques have necessitated further automation and generated the need for prostate specific multimodality image registration algorithms. In phase II, we aim to further reduce human interaction in prostate delineation without compromising on algorithm performance by developing and validating more advanced segmentation algorithms. We will automate post-implant monitoring by segmenting the prostate from post-implant TRUS images by developing advanced segmentation and registration algorithms. We will validate the algorithms on large clinical dataset. PROPOSED COMMERCIAL APPLICATIONS: Our technology is of great interest to vendors of prostate cancer treatment planning systems, be it in the area of prostate brachyltherapy or in prostate cryotherapy. The movement of radiation treatment planning systems to inter-operative settings make automated delineation technology even more valuable to these vendors. This technology is also of interest to manufacturers of ultrasound machines.

描述（由申请人提供）：该提案的目的是提高直肠超声（TRU）在诊断和治疗前列腺疾病中的有效性。 TRUS图像中的前列腺边界提供了诊断，治疗计划和随访所需的解剖描述。我们提出的研究使计算机协助的前列腺描述使用与专家的“恰好互动”。 在第一阶段，我们将几种分割算法嵌入了软件环境中，允许与医生相互作用。与手动描述相比，（半）自动前列腺描述的变异性降低了一半。同样，与手动概述相比，描述时间减少了2倍。 基于对我们的技术表现出兴趣的几家公司和最终用户的反馈，我们发现在过去的一年中，术中新兴的术中临床程序和处理后监测技术已经需要进一步自动化，并产生了对前列腺特定多模化的需求图像注册算法。在第二阶段，我们旨在通过开发和验证更先进的分割算法来进一步减少前列腺描述中的人类相互作用，而不会损害算法性能。我们将通过开发高级分割和注册算法从植入后TRUS图像分割前列腺来自动化植入后监测。我们将验证大型临床数据集上的算法。 拟议的商业应用： 我们的技术对前列腺癌治疗计划系统的供应商非常感兴趣，无论是在前列腺胆管治疗还是前列腺冷冻疗法中。辐射处理计划系统上术间设置的运动使自动描述技术对这些供应商更有价值。超声机器制造商也感兴趣。