AI vendor Infervision's InferRead CT Pneumonia software uses artificial intelligence-assisted diagnosis to improve the overall efficiency of the radiology department. It is being developed in China as a high sensitivity detection aid for novel coronavirus pneumonia (COVID-19).
An older couple walked into the Hubei Provincial Hospital of Integrated Chinese and Western Medicine near their neighborhood with a fever and cough on Dec. 26, 2019. Computed tomography (CT) scans showed lesions in their lungs distinct from anything caused by any other known viral pneumonia. Jixian Zhang, M.D., director of the hospital’s Respiratory Medicine Department, told the couple to get their son over for a check. The son was asymptomatic, but his CT images found similar conditions. On the same day a vendor from the Hua’nan Seafood Market also had a fever and coughing with similar pulmonary manifestations. Zhang started noticing more and more cases alike not long before her professional intuition kicked in that this was the start of a new type of viral infection, having personal experience dealing with the 2003 SARS outbreak. After careful examination and analysis, Zhang and the hospital leaders decided to report the cases to superior authorities. Zhang was later called the “first to report” the outbreak.
Computed Tomography Plays a Key Role in COVID-19 Detection in China
Computed tomography medical imaging is used in detecting the abnormalities in the patients’ lungs. In the images many early novel coronavirus (COVID-19, SARS‐CoV‐2) patients have single or multiple small patchy ground-glass opacities and interlobular septal thickening. The number and area of disease foci grow as the disease progresses. Consolidation and linear shows appear in addition to the ground-glass opacities. Diffuse lesions develop in both lungs when the patient becomes critically ill. Some patients may have what is known as the pale lungs.
In the early days of the outbreak, radiological imaging was not regarded as a way to confirm evidence for COVID-19 cases, instead relying on a positive result of the PCR nucleic test. The supply of the PCR test kits, however, was limited and it took a long time to get the results after collecting the specimen. Many suspected patients had to wait to be tested to confirm their infection. The Chinese health authority soon recognized the gap and changed the diagnostic strategy in their 5th edition of the Guidelines on Diagnosis and Treatment of COVID-19. Radiological features of COVID-19 were included as one of the three determinant clinical manifestations to confirm a suspicious patient. The high efficiency, reliability and accessibility of the radiological diagnostic technology aided identification of the disease in the outbreak. CT image was soon enhanced further by the introduction of artificial intelligence (AI) algorithms that can help quickly triage patients to radiomic COVID-19 signatures in the images.
AI Helps Speed Radiological Review and Diagnosis of COVID-19 Patients
Inclusion of radiological findings in confirming COVID-19 diagnosis instantly increased the workload of the already hectic radiologists and physicians, who had to visually go through up to 300 images of a patient. Human efficiency and accuracy along with the risky physical contact in the exam process are gravely challenged. This was exactly the kind of situation where AI’s strengths find its ways and liberate the medical staff for more intimate care for the patients where human presence and interventions are indispensable and invaluable.
AI may help with both the exam process and reading of the images taken. When handling a patient, a radiologist or technician has to come in physical contact with the patient for instructions about how to position and breathe correctly. AI is used to take the human presence out of the exam room and allow the radiologist to guide the patient through the process contact-free, which also helps to save consumption of personal protection equipment (PPE) which otherwise had to be worn.
As its more valued application, AI with a deep learning algorithm has proven to be a powerful aid in recognizing the lesions in CT images and even quantitatively characterizing the findings and comparing changes between exams, which works at a considerably greater speed and accuracy. Some algorithms may even help differentiate COVID-19 from a regular viral pneumonia. When a CT image suspicious of COVID-19 is detected, the AI alerts the physician and brings the case to the top of the physician’s work list, suggests possible infection, and recommends preset interventions according to the findings. This significantly improves the detection rate and treatment consistency of COVID-19 cases.
VIDEO: How China Leveraged Health IT to Combat COVID-19 — Interview with Jilan Liu, M.D., CEO for the HIMSS Greater China
How Artificial Intelligence Aided Chinese Fight Against COVID-19
The following list summarizes the known AI-assisted radiological applications in China:
About the author: Jilan Liu, M.D., MHA, is the CEO for HIMSS Greater China. She was previously principal and consulting director of the Greater China Joint Commission International (JCI). As chief executive for HIMSS Greater China, Liu and her team made a significant breakthroughs with hospitals and the HIT industry in the greater China region to embrace international HIT standards, certification requirements and to rethink prudent approachs to IT investments and operations.
Related COVID-19 Content:
VIDEO: Imaging COVID-19 With Point-of-Care Ultrasound (POCUS) — Interview with emergency physician Mike Stone, M.D.,
VIDEO: What Cardiologists Need to Know about COVID-19 — Interview with Thomas Maddox, M.D.
2. 天翼云联合上海联影推出AI肺炎诊断系统 砍柴网 2020-03-05 https://baijiahao.baidu.com/s?id=1660304154838043708&wfr=spider&for=pc
3. 鉴别新冠肺炎CT影像 这个团队20天搭建出AI辅助系统 中国科技网 2020-03-09 http://www.stdaily.com/kjrb/kjrbbm/2020-03/09/content_896703.shtml
4. NCP版肺炎AI系统，5分钟完成疑似病例影像诊断 北晚新视觉网 2020-03-06 https://baijiahao.baidu.com/s?id=1660391418090724336&wfr=spider&for=pc
5. 依图科技助力疫情防控 腾讯网 2020-02-20 https://new.qq.com/omn/20200220/20200220A04CGL00.html?pc
6. 3秒内完成新冠肺炎定量筛查！全国首个抗疫AI影像系统是怎样炼成的？ 量子位QbitAI 2020-02-22 https://www.thepaper.cn/newsDetail_forward_6067440
7. 柏视新动态||新型冠状病毒肺炎辅助阅片系统 柏视医疗官网 2019-02-14 https://www.pvmedtech.com/news/post-14.html
8. 原标题：东软医疗与钟南山团队联手推出“火眼AI” 中国日报 2020-02-17 http://cnews.chinadaily.com.cn/a/202002/17/WS5e49e193a3107bb6b57a00ed.html
9. 体素科技CT多病种加入“防疫” 减少院内交叉感染 20200303 http://sh.people.com.cn/n2/2020/0303/c134768-33847064.html
10. 腾讯AI部署湖北最大方舱医院 助力秒级识别新冠肺炎 2020-02-22 中国日报网 https://baijiahao.baidu.com/s?id=1659206548310668824&wfr=spider&for=pc
11. 抗击疫情 | 视见科技新冠肺炎CT+AI上线深圳三院等多家医院 视见科技官网 2020-02-25 https://www.imsightmed.com/news/detail/id/66.html
12. 抗击疫情 | 视见AI携手医生共抗战“疫”视见科技官网 2020-02-12 https://www.imsightmed.com/news/detail/id/65.html
13. 点内科技新冠肺炎AI-赋能医疗防控战“疫”腾讯网 2020-02-18 https://new.qq.com/omn/20200218/20200218A0D83600.html?pc
14. 华为云推出新冠肺炎AI辅助诊断服务，CT量化结果秒级输出 华为云官网 2020-02-10 https://www.huaweicloud.com/news/2020/20200210195951099.html
15. 上海人工智能研究院携手健培科技共同发布新冠肺炎影像云，AI助力疫情防控 健培科技官网 http://www.jianpeicn.com/about-jianpei/news/3055/
16. 新冠肺炎影像云检测平台发布两周显威力！健培科技官网 http://www.jianpeicn.com/about-jianpei/news/3099/
17. 中大五院携手汇医慧影 新冠AI辅助筛查 “科研+临床”双管齐下 北国网 2020-03-03 http://shangjie.lnd.com.cn/jishi/2020/03/03/440848.html
18. 中科院完成“新冠肺炎影像辅助诊断平台”建设 科大讯飞参与负责 2020-03-02 http://finance.ifeng.com/c/7uW1AYNM6cx
19. 深睿医疗新冠肺炎增强版AI系统发往湖北 驰援一线 沈睿医疗健康号 https://www.cn-healthcare.com/articlewm/20200205/wap-content-1086087.html
20. 以AI驰援，商汤肺部智能分析产品助力多地科技抗“疫” 商汤科技官网 2020-02-24 https://www.sensetime.com/News/view/id/203.html
21. 平安智慧医疗火线支援全国1500+家医院新冠肺炎CT阅片 扬子晚报网2020-02-24http://www.yangtse.com/zncontent/303855.html
22. 新闻媒体深度专访：PereDoc以科技助力防疫抗疫 PereDoc官网 2020-02-24 http://www.peredoc.com/home/news/desc/id/70.html
23. 基于百度飞桨的连心医疗肺炎筛查和预评估AI系统投入使用 网易科技 2020-02-29 http://tech.163.com/20/0229/10/F6I0P3I200098IEO.html
24. 助力基层疫情防控！零氪科技“新冠肺炎AI辅诊系统”多地上线 零氪科技健康号 2020-02-17 https://www.cn-healthcare.com/articlewm/20200213/wap-content-1088000.html