SBIR Phase I Topic 402: Artificial Intelligence-Aided Imaging for Cancer Prevention, Diagnosis, and Monitoring

SBIR 第一阶段主题 402：用于癌症预防、诊断和监测的人工智能辅助成像

• 批准号: 10265917
• 负责人: Xin Qi
• 金额: $ 39.23万
• 依托单位: REVEAL BIOSCIENCES, INC.
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-16 至 2021-06-15
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10265917
• 关键词: Advanced Development; Algorithmic Software; Artificial Intelligence; Biological Sciences; Biopsy; Computer software; Contracts; Data; Data Sources; Decision Making; Diagnosis; Diagnostic; Disease Progression; Gleason Grade for Prostate Cancer; Human; Image; Individual; Malignant neoplasm of prostate; Medical Imaging; Monitor; Pathologist; Patient-Focused Outcomes; Patients; Phase; Physicians; Pilot Projects; Population; Positioning Attribute; Procedures; Prostate; Prostatectomy; Provider; Recurrence; Risk; Small Business Innovation Research Grant; Testing; Time; TimeLine; Validation; aggressive therapy; base; cancer imaging; cancer prevention; design; disorder risk; high risk; imaging modality; imaging platform; imaging software; improved; patient stratification; prediction algorithm; prognostic assays; prototype; tool; usability

Current diagnostic approaches cannot determine which patients with early prostate cancer are most at risk for progression or recurrence, which leaves patients and physicians in the difficult position of choosing between active surveillance or aggressive therapy based on population risks associated with subjectively assigned Gleason scores. Stratifying patients into high-risk and low-risk groups at the time of initial biopsy or prostatectomy would provide critical information that providers and patients need to make decisions about prostate-sparing approaches or more aggressive treatment options that could improve patient outcomes. In this Phase I contract, Reveal Biosciences proposes to advance the development of imageDx: Prostate, an image-based, artificial-intelligence (AI)-enabled prognostic test designed to stratify patients with early stage prostate cancer according to their risk for disease progression or recurrence. Objectives include 1) selecting an imaging platform for test validation, 2) refining a user interface for the tool, 3) conducting a pilot study to demonstrate the feasibility of a risk prediction algorithm, 4) developing standard operating procedures for the tool, and 5) developing a regulatory strategy for test approval by the FDA.

目前的诊断方法无法确定哪些早期前列腺癌患者最有进展或复发的风险，这使得患者和医生处于根据与主观指定的格里森评分相关的人群风险在主动监测或积极治疗之间做出选择的困难境地。在初次活检或前列腺切除术时将患者分为高风险和低风险组将为提供者和患者做出有关前列腺保留方法或更积极的治疗方案的决策所需的关键信息来改善患者的预后。在这份第一期合同中，Reveal Biosciences 提议推进 imageDx：前列腺的开发，这是一种基于图像、人工智能 (AI) 的预后测试，旨在根据早期前列腺癌患者的疾病进展或病情风险对他们进行分层。复发。目标包括 1) 选择用于测试验证的成像平台，2) 完善该工具的用户界面，3) 进行试点研究以证明风险预测算法的可行性，4) 为该工具开发标准操作程序，以及 5 ）制定 FDA 测试批准的监管策略。