COVID-19 GLOBAL PERSPECTIVES – THE RAPIDLY EVOLVING ROLE OF RADIOLOGY AND AI
March 18, 2020 [Updated on April 13 and May 18, 2020] – COVID-19 had affected 164 countries and territories around the world and 1 international conveyance (the Diamond Princess cruise ship harboured in Yokohama, Japan), the number of reported cases on March 18th exceeded 200,000 As of May 18, 2020, the Coronavirus is almost everywhere, with over 2.6 million active cases.
This is the seventh known Coronavirus to infect humans . Two other notable examples include severe acute respiratory syndrome (SARS) and the Middle East respiratory syndrome (MERS), the former of which began in southern China and resulted in 774 deaths out of 8,098 infected individuals in 29 countries from November 2002 through July 2003, and the latter of which was first identified in Saudi Arabia and was responsible for 858 deaths among 2,494 individuals in 27 countries since September 2012 [2,3].
As COVID-19 spreads globally, there has been growing interest relating to the role of diagnostic imaging, appropriateness of Chest X-rays and CT scans when it comes to screening, detection and follow up management. There were also initial reports that indicated some technology startups were evaluating their machine learning algorithms to detect COVID-19 specific findings leveraging Chest X-rays and CT scans.
While the appropriateness of Chest X-rays or CT scans to pinpoint COVID-19 pneumonia is being evaluated, a group of Italian experts has been exploring the benefits of bedside ultrasound as one possible alternative for detection of COVID-19 pneumonia.
The data from Italy is preliminary and further studies are needed to confirm the role of lung ultrasound in the diagnosis and management of COVID-19, the authors are strongly recommending the use of bedside ultrasound for early diagnosis of COVID-19 pneumonia in all patients presenting to the emergency department with flu-like symptoms. The publication is available here.
Since there’s limited data, and we are still learning about this disease, and not enough peer-reviewed studies available regarding the role of AI, let’s focus first on the latest updates regarding the appropriateness of radiology imaging when it comes to detection and management of COVID-19.
COVID-19, AI AND DATA ANALYTICS:
Globally, healthcare startups have already begun testing AI algorithms to study COVID-19 pattern recognition or relevant variations that may help with early disease detection. Countries are releasing emergency funds to aid in research and development of diagnostic kits, vaccines and medicines.
There’s already some work in progress internationally to study the role of deep learning algorithms relating to pattern recognition on Chest CT scans. When it comes to detecting abnormalities in lung imaging, focused work has already been done in this space, hence with COVID-19 it may not require a lot of work to further improve the sensitivity and specificity of AI algorithms. However, there’s still more work to be done and international validation required.
On March 11, the ACR® Data Science Institute® (DSI) has published an artificial intelligence (AI) use case on COVID-19 , which is now open for public comment. Expert radiologists rapidly developed the use case to offer the developer community medical context — including necessary inputs, outputs and possible corollary features — for developing an AI solution to detect COVID-19.
Recently, some papers have been published reviewing the role of RT-PCR and Chest CT Scans and how CT can play a vital role in the early detection and management of COVID-19. However, it is worth emphasizing that a patient with reverse-transcription polymerase chain reaction (RT-PCR) confirmed COVID-19 infection may have normal chest CT findings at admission. (Ref: link)
A couple of early CT imaging patient case studies coming out of China in the past months reported COVID-19 was preliminarily diagnosed from CT scans in several patients before they started to show positive RT-PCR test results. Additional COVID-19 radiology research can be found at Special Focus: COVID-19.
AI and Analytics for forecasting the trajectory of COVID-19: As an alternative to epidemiological models for transmission dynamics of Covid-19 in China, a recent paper proposes artificial intelligence (AI)-inspired methods for real-time forecasting of Covid-19 to estimate the size, lengths and ending time of Covid-19 across China.
According to this paper, the accuracy of the AI-based methods for forecasting the trajectory of Covid-19 was high. They predicted that the epidemics of Covid-19 will be over by the middle of April. “If the data are reliable and there are no second transmissions, we can accurately forecast the transmission dynamics of the Covid-19 across the provinces/cities in China. The AI-inspired methods are a powerful tool for helping public health planning and policy-making.”
COVID-19 Open Research Dataset (CORD-19): In a recent briefing, research leaders across tech, academia and the government joined the White House to announce an open data set full of scientific literature on the novel coronavirus. The dataset is freely available for the global research community to apply recent advances in natural language processing and other AI techniques to generate new insights in support of the ongoing fight against this infectious disease.
Supporting our care providers with technology and flexible workflows while they are coping with the COVID-19 challenge:
The need for digital imaging and fast access to radiology scans via remote reading applications is on the rise. Just recently, Agfa HealthCare shared the experience of how it’s Enterprise Imaging Radiologist workstation was deployed speedily within 20 minutes at a Radiologist’s home in the UK who was concerned about being able to work should the need to self-isolate arise. Further details regarding how Agfa HealthCare is supporting its clients with an Enterprise Imaging strategy are posted here.
The Royal College of Radiology UK has also released guidance on information technology requirements for homeworking. RCR resources have been developed to strike a balance between the urgent need for pragmatic home working solutions in response to COVID-19 and the quality of images needed for radiologists to interpret images. Details are available on RCR UK website.
To understand the broader technology and service provider landscape, KLAS Research reached out to vendor community to learn more about what they are doing/can do to help organizations in this crisis. The “COVID-19 Technology & Services Solutions Guide” from KLAS Research is available here.
The COVID-19 situation is fast evolving, more evidence relating to screening, detection, diagnosis and management is being captured. In the meantime, the key takeaway for each and every one of us is to help ease the burden on our health systems by practicing the guidelines being emphasized by local health authorities, so that the care providers at the forefront of care delivery may focus on looking after the ones impacted the most.
Wishing you all, the best of health, well-being and safety.
Published March 18, 2020 – Further updated on April 13 and May 18, 2020