Call for collaboration for development of AI support for COVID-19 diagnosis and prognosis from imaging and clinical data, University of Cambridge, UK

Project team: The team consists of PIs and researchers who are members of the Cambridge Mathematical Imaging in Healthcare Centre (CMIH), Addenbrookes and Royal Papworth Hospital, namely:

Prof. Carola-Bibiane Schönlieb (Centre for Mathematical Sciences, University of Cambridge)
Prof. Evis Sala (Department of Radiology, University of Cambridge)
Dr Zhongzhao Teng (Department of Radiology, University of Cambridge)
Dr Michael Roberts (Researcher in deep learning from CT scans)
Dr Thomas White (MRC Epidemiology Unit, University of Cambridge)
Dr Mishal Patel (Imaging and data analytics expert)
Dr Muhunthan Thillai (Clinical lead for interstitial lung disease)
Dr Alessandro Ruggiero (Royal Papworth Hospital, University of Cambridge)
Dr Lorena Escudero (Department of Radiology, University of Cambridge)
Dr Judith Babar (Thoracic Imaging Lead and Consultant Radiologist, Addenbrookes)

Project proposal: To support a rapid diagnosis of COVID-19 and patient prognosis, our team of Cambridge researchers are eager to bring expertise in AI for imaging (in the CMIH) together with expertise in radiology and clinical applications (in Addenbrookes and Papworth) to develop an algorithm that can rapidly diagnose and suggest a prognosis to the clinician based on information from imaging and clinical data from various demographies. We are aware of several attempts and advances in this direction in several parts of the world. We believe, however, that these individual, localised efforts are not enough to face the current global crisis. There is a need for truly joined-up effort to provide a free, open-source tool to hospitals all over the world.

Therefore, we are reaching out to our colleagues in China for help in this endeavour. China not only has precious imaging data of suspected COVID-19 patients and their demographics, but also made advances in the idea of deep learning for diagnosing COVID-19 on the basis of the chest x-rays or CT scans. We would be very grateful for your support in one or both of:
• getting access to x-ray or CT data of individuals suspected and/or diagnosed with COVID-19;
• getting access to deep learning algorithms developed in China which can diagnose from x-ray or CT scans; sharing the algorithms in a form so that our Cambridge team can re-train them on the local population.

What we are asking for at the moment only involves a small set of researchers but, depending on what is possible, could eventually span the whole world and develop into a global effort. Your support and help are greatly appreciated.

Colleagues from China, please contact Dr Zhongzhao Teng, Department of Radiology, University of Cambridge, Email: zt215@cam.ac.uk.

英国剑桥大学呼吁国际合作开发针对X射线或CT扫描和临床信息的COVID-19 AI快速诊断和愈后预测工具

团队成员：由来自剑桥大学，剑桥大学附属Addenbrookes和剑桥大学附属皇家Papworth医院项目负责人和研究人员组成，包括：
Carola-Bibiane Schönlieb教授 (剑桥大学，数学科学中心)
Evis Sala 教授 (放射系，剑桥大学)
滕忠照 博士 (放射系，剑桥大学)
Michael Roberts博士 (影像数据分析专家)
Thomas White博士 (MRC 流行病学中心, 剑桥大学)
Mishal Patel 博士 (影像数据分析专家)
Muhunthan Thillai 博士 (间质性肺疾病专家)
Alessandro Ruggiero 博士 (皇家Papworth医院, 剑桥大学)
Lorena Escudero 博士 (放射系，剑桥大学)
Judith Babar 放射系 (胸部影像, Addenbrookes医院，剑桥大学)

项目：为了支持全球范围内的COVID-19的快速诊断和愈后预测，我们剑桥大学的研究人员团队渴望整合人工智能，放射学和临床，开发出基于X射线或CT扫描和临床信息的算法以支持不同国家的快速诊断工具。我们注意到某些地方在这方面已经做了尝试，并取得进步和成果。但是，我们认为，单独的，局部的努力不足以应对当前的全球危机。我们需要真正的共同努力为世界各地的医院提供免费的开源快速诊断工具。

为此，我们正在努力寻求中国科学家和医疗工作者的帮助。中国不仅拥有宝贵的可疑和确诊的COVID-19患者的影像数据和其他临床信息，而且在基于胸部X射线或CT扫描诊断COVID-19的深度学习方面也取得了进展。我们非常感谢您对以下一项或两项的支持：
• 获得疑似和/或确诊的COVID-19的患者的X射线或CT数据；
• 获得在中国开发的可以通过X射线或CT扫描诊断的深度学习算法；以某种形式共享算法，以便我们的剑桥团队可以对世界不同地方来的数据进行重新训练。
目前，我们所要求的是小范围的研究人员参与，但视情况而定，最终会演变为一项全球努力。非常感谢您的支持和帮助。

来自中国的同行，请联系滕忠照博士，剑桥大学放射系，zt215@cam.ac.uk