Computer-assisted Grading and Risk Stratification of Follicular Lymphoma

滤泡性淋巴瘤的计算机辅助分级和风险分层

• 批准号: 8215904
• 负责人: Metin Nafi Gurcan
• 金额: $ 30.73万
• 依托单位: OHIO STATE UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-05-01 至 2014-07-17
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8215904
• 关键词: Accounting; Adopted; Adult; Aggressive Clinical Course; Agreement; Area; B-Cell Lymphomas; Biologic Characteristic; Biological Markers; Cancer and Leukemia Group B; Cells; Centroblast; Classification; Clinic; Clinical; Clinical Trials; Collection; Communities; Computer Assisted; Computers; Data; Data Set; Databases; Detection; Development; Diagnosis; Diagnostic; Disease; Evaluation; Failure; Follicular Lymphoma; Genetic Markers; Goals; Health; Histopathology; Image; Image Analysis; Indolent Clinical Course; Institution; Lymphoma; Morphology; Non-Hodgkin' s Lymphoma; Ohio; Outcome; Pathologist; Pathology; Patients; Performance; Private Practice; Process; Radiology Specialty; Reader; Research; Research Personnel; Resources; Risk; Scanning; Selection Bias; Site; Slide; Software Tools; Stratification; System; T-Lymphocyte; Techniques; Testing; Tissues; Toxic effect; United States; Universities; World Health Organization; base; chemotherapy; clinical practice; computer aided detection; computer generated; computerized; design; digital; experience; high standard; improved; lymph nodes; macrophage; monocyte; outcome forecast; public health relevance; software development; tool; tumor

DESCRIPTION (provided by applicant): Follicular lymphoma (FL) is the second most common non-Hodgkin's lymphoma. Several treatment options exist today, but these are costly and include significant toxicities. No biological or genetic markers are available in clinical practice for reliable risk stratification of follicular lymphomas and the choice of appropriate treatment depends heavily on morphology-based histological grading. In a system adopted by the World Health Organization (WHO), follicular lymphomas are stratified into three grades depending on the average count of centroblasts in ten randomly selected, standard high-power fields (HPFs). Follicular lymphomas with low histological grades show an indolent clinical course with long average survival, but are considered incurable with currently available therapies. In contrast, high-grade follicular lymphomas have an aggressive clinical course and are rapidly fatal if not treated with aggressive chemotherapy. However, in contrast to low-grade follicular lymphoma, high-grade FL may be cured with aggressive chemotherapy. Currently, the inter-reader agreement between pathologists in grading FL is extremely low. In a multi-site study, the agreement among experts for the various grades of follicular lymphoma varied between 61% and 73%. Since only ten HPFs are used by the pathologist for practical reasons, this system may be prone to selection bias in cases that show significant differences in various areas of a section. The primary goal of this project is to develop an effective computer-aided system to assist pathologists in making diagnostic decisions about histological grading of follicular lymphoma. It is important to note that this project aims to provide supplementary information to the pathologist as he or she carries out the classification process; this is not an attempt to automate the classification process. To achieve this objective, ten board-certified hematopathologists (with experience in grading follicular lymphoma) from The Ohio State University, Cleveland Clinic, Vanderbilt University, and private practice will participate in the creation of the database that will contain digitized follicular lymphoma slide images, as well as the associated truth for the development and evaluation of the computer-aided follicular lymphoma grading system. After extensive evaluation of the system with the collected datasets and outcome data, as well as datasets from the Cancer and Leukemia Group B (CALGB) trials, the developed system will be installed at participating pathologists' institutions, and the developed software will be made available to the research community as a shareable resource. PUBLIC HEALTH RELEVANCE: Follicular lymphoma (FL) is the second most common non-Hodgkin's lymphoma. This project aims to provide supplementary information to the pathologist for the grading of the tumor using computerized image analysis techniques. The supplementary information will be useful for better diagnosis, prognosis and treatment of this disease.

描述（由申请人提供）：滤泡性淋巴瘤（FL）是第二常见的非霍奇金淋巴瘤。目前存在几种治疗方案，但这些治疗方案成本高昂，而且有明显的毒性。在临床实践中，没有生物学或遗传标记可用于可靠的滤泡性淋巴瘤风险分层，适当治疗的选择在很大程度上取决于基于形态学的组织学分级。在世界卫生组织（WHO）采用的系统中，滤泡性淋巴瘤根据随机选择的10个标准高倍视场（HPFs）中成中心细胞的平均计数分为三个等级。组织学分级低的滤泡性淋巴瘤临床病程缓慢，平均生存期长，但目前可用的治疗方法被认为是无法治愈的。相比之下，高级别滤泡性淋巴瘤具有侵袭性的临床病程，如果不进行积极的化疗治疗，可迅速致命。然而，与低级别滤泡性淋巴瘤相比，高级别滤泡性淋巴瘤可以通过积极的化疗治愈。目前，病理医师对FL分级的一致性非常低。在一项多地点研究中，专家对不同级别滤泡性淋巴瘤的一致性在61%到73%之间。由于实际原因，病理学家只使用10个hpf，因此在切片的各个区域显示显着差异的情况下，该系统可能容易产生选择偏差。本计画的主要目标是发展一套有效的电脑辅助系统，以协助病理学家对滤泡性淋巴瘤的组织学分级作出诊断决定。重要的是要注意，这个项目旨在为病理学家提供补充信息，因为他或她进行分类过程；这并不是试图使分类过程自动化。为了实现这一目标，来自俄亥俄州立大学、克利夫兰诊所、范德比尔特大学和私人诊所的10名经认证的血液病理学家（具有滤泡性淋巴瘤分级经验）将参与创建数据库，该数据库将包含数字化滤泡性淋巴瘤幻灯片图像，以及用于开发和评估计算机辅助滤泡性淋巴瘤分级系统的相关事实。在使用收集的数据集和结果数据以及来自癌症和白血病B组（CALGB）试验的数据集对系统进行广泛评估后，开发的系统将安装在参与的病理学家机构中，开发的软件将作为可共享资源提供给研究界。公共卫生相关性：滤泡性淋巴瘤（FL）是第二常见的非霍奇金淋巴瘤。本项目旨在为病理学家使用计算机图像分析技术对肿瘤进行分级提供补充信息。补充资料将有助于本病更好的诊断、预后和治疗。