Geles: A Novel Imaging Informatics System for Generalizable Lesion Identification in Neuroendocrine Tumors

Geles：一种用于神经内分泌肿瘤普遍病变识别的新型影像信息学系统

• 批准号: 10740578
• 负责人: BENNETT B CHIN
• 金额: $ 38.85万
• 依托单位: UNIVERSITY OF COLORADO DENVER
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-09-19 至 2025-08-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10740578
• 关键词: Acceleration; Address; Algorithms; Artificial Intelligence; Clinical Research; Communities; Data; Data Set; Detection; Development; Devices; Diagnosis; Disease; Eligibility Determination; Emission-Computed Tomography; Enhancing Lesion; Hepatic; Human; Image; Image Analysis; In complete remission; Individual; Informatics; Institution; Learning; Lesion; Liver; Manuals; Medical; Medical Imaging; Medical Research; Methods; Modeling; Molecular Target; Neoplasm Metastasis; Neuroendocrine Tumors; Noise; Patient-Focused Outcomes; Patients; Peptides; Performance; Pharmaceutical Preparations; Positron-Emission Tomography; Progression-Free Survivals; Protocols documentation; Radioisotopes; Regimen; Reproducibility; Research; Residual Neoplasm; Residual state; Resolution; Resource-limited setting; Resources; Site; Somatostatin Receptor; Standardization; System; Systemic Therapy; Technology; Testing; Therapeutic; Training; Variant; X-Ray Computed Tomography; algorithm development; anticancer research; burden of illness; clinical practice; cost; data acquisition; data resource; deep learning; deep learning model; deep neural network; design; gastroenteropancreatic neuroendocrine tumor; generative adversarial network; image reconstruction; image translation; imaging informatics; imaging system; improved; informatics tool; interest; learning strategy; molecular imaging; novel; novel therapeutics; open source; receptor; reconstruction; standard of care; task analysis; therapy development; treatment response; tumor

PROJECT SUMMARY Gastroenteropancreatic neuroendocrine tumors (GEP-NETs) are difficult to detect tumors which commonly present at advanced stages, with the liver as the most common site of metastases. 68Ga and 64Cu DOTATATE positron emission tomography-computed tomography (PET/CT) are the most sensitive methods to identify somatostatin receptor subtype 2 positive GEP-NETs, and targeted peptide radionuclide receptor therapy with 177Lu DOTATATE is the most effective systemic therapy for many patients. Despite the clear advantage in progression-free survival compared to prior standard of care, the vast majority of patients (99%) do not have complete response and require additional therapies. Further development of treatments requires an accurate assessment of the response to therapy. However, there is currently an unmet medical need for automated, standardized quantification of 68Ga DOTATATE positive disease burden, which could have a great impact on novel therapeutic drug regimen development. Deep learning-based approaches have recently been applied to automated lesion detection and quantification, and have achieved state-of-the-art performance. These methods, however, do not consider dataset/domain shifts between training and testing data. In dataset/domain shifts, data used to build and train models might have a significantly different distribution from that used for model testing. Therefore, models without considering domain shifts would not generalize well to unseen data, leading to poor lesion detection performance. In this proposed research, we will develop a novel deep learning-based imaging informatics system, termed Geles, for automated, Generalizable lesion detection for livers in GEP-NET PET/CT imaging. This system will use list-mode data acquisition to produce a large, diverse annotated training dataset, followed by novel adversarial learning to enhance model generalizability. The proposed Geles system will consist of two modules, domain generalization and domain adaptation. Aim 1 will develop an adversarial domain generalization module that is generalizable to unseen domains or resources. This module will build a deep neural network with domain-adversarial learning and extract domain-invariant feature representations for individual lesion identification, so that the system can generalize to unseen domain data, such as PET images from different institutions, devices, imaging protocols, and other variations. Aim 2 will develop a target-oriented domain adaptation module that is automatically adaptable to new specific datasets of interest (i.e., target datasets). Given a small set of unannotated images from a certain target dataset, this module will conduct low-resource unsupervised domain adaptation to further boost the lesion detection performance. Specifically, it will build a novel, augmented generative adversarial network for image-to-image translation in a low-resource setting, so that Geles can take advantage of limited, unannotated specific target data and conduct target-oriented, enhanced lesion detection.

项目摘要 胃肠道神经内分泌肿瘤（GEP-NET）很难检测到通常 出现在高级阶段，肝脏是最常见的转移部位。 68GA和64cu dotatate 正电子发射断层扫描（PET/CT）是识别最敏感的方法 生长抑素受体亚型2阳性GEP NET和靶向肽放射性核素受体治疗 177lu dotatate是许多患者最有效的全身疗法。尽管有明显的优势 与先前的护理标准相比，无进展的生存期，绝大多数患者（99％）没有 完全反应，需要其他疗法。进一步的治疗需要准确 评估对治疗的反应。但是，目前有未满足的自动化需求， 标准化的68GA Dotatate阳性疾病负担，这可能会对 新型的治疗药物方案。基于深度学习的方法最近已应用于 自动病变检测和定量，并实现了最先进的性能。这些方法， 但是，不要考虑训练和测试数据之间的数据集/域移动。在数据集/域移动中，数据 用于构建和火车模型的分布可能与用于模型测试的分布明显不同。 因此，不考虑域移动的模型不会很好地推广到看不见的数据，从而导致差 病变检测性能。在这项拟议的研究中，我们将开发一种新颖的基于学习的成像 信息系统，称为凝胶，用于GEP-NET PET/CT中肝的自动化，可概括的病变检测 成像。该系统将使用列表模式数据采集来生成大型，多样化的注释培训数据集， 然后进行新颖的对抗性学习，以增强模型的推广性。 所提出的凝胶系统将由两个模块组成：域的概括和域的适应性。目标1 将开发一个可推广到看不见的域或资源的对抗域概括模块。 该模块将建立一个深层神经网络，并通过域 - 对抗性学习并提取域不变 单个病变识别的特征表示，因此系统可以推广到看不见的域 数据，例如来自不同机构，设备，成像协议和其他变化的PET图像。目标2将 开发一个面向目标的域适应模块，该模块自动适应于新的特定数据集 兴趣（即目标数据集）。给定一组来自​​某个目标数据集的未注释的图像，此模块 将进行低资源的无监督域的适应性，以进一步提高病变检测性能。 具体而言，它将建立一个新颖的，增强的生成对抗网络，以用于图像到图像翻译 低资源设置，以便凝胶可以利用有限的，未注释的特定目标数据并进行操作 面向目标的，增强的病变检测。