Blog

AI COVID-19 An algorithm developed in a collaborative approach

The coronavirus pandemic continues to spread, causing unprecedented system disruption and confronting healthcare professionals around the world with clinical and operational challenges. We believe the COVID-19 pandemic will significantly change healthcare systems. Specifically, it will serve as a catalyst for the development of more effective, efficient and, above all, more humane medicine.

Coronaviruses (CoV) are a large family of viruses that cause diseases ranging from the common cold to more severe diseases such as Middle East Respiratory Syndrome (MERS-CoV) and Severe Acute Respiratory Syndrome (SARS-CoV). COVID-19 is caused by the severe-acute respiratory syndrome Coronavirus 2 (SARS-Cov2) and has a high case fatality rate of up to 4% (1). Due to a high reproduction number and the infectious nature of the disease, tools for rapid testing and evaluation are vital to track and mitigate the spread.

At Siemens Healthineers we understand the urgency and complexity of this situation. In a collaborative approach together with our research network and collaboration partners, we have developed the CT Pneumonia Analysis2 algorithm.

COVID-19: A challenge in diagnosis and therapy planning

COVID-19 is confirmed via a reverse transcription polymerase chain reaction (RT-PCR)3, a laboratory technique to detect genetic material from any pathogen, including a virus. However, the sensitivity of this test may be as low as 60-70%4 and can cause false negatives in the first few days, which means in the best case patients are still treated as if they have the virus, while in the worst case infected patients might be treated as not having the virus, thereby posing a potential risk to other patients. Moreover, a current a shortage of available testing kits is contributing to undetected spread.

Although the role of CT and X-ray for diagnosis is currently being debated, two preliminary studies5 showed that Chest CT imaging of the lung provides improved sensitivity when compared to RT-PCR for patients suspected of having COVID-19. The primary features seen on a lung affected by COVID-19 are peripheral focal or multi-focal ground glass opacities, consolidation and crazy-paving patterns.

Opacity lesions in the peripheral and posterior lungs on CT images are indicative of COVID-19 pneumonia. CT can play an important role in the evaluation of COVID-19 providing advanced imaging evidence

Non-contrast chest Computed Tomography has been useful not only to detect, quantify severity, and assess progression of the disease, but also to support the design of a curative therapy. Similar results can be achieved with Chest X-rays, which can be useful in tracking the disease progression and follow-up.

A collaboration as an answer to the complex situation

Our collaboration partners and our own expert teams below are aware of the urgency and complexity of the current situation. That’s why we decided to jointly focus on the development of our new algorithms to face the COVID-19 pandemic together. Our aspiration to drive innovations forward so people live healthier and longer lives is more valid today than ever before.

(1) Hôpital Foch, Paris, France

(2) Northwell Health, New York, NY, USA

(3) University Hospital Basel, Clinic of Radiology & Nuclear Medicine, Basel, Switzerland

(4) Vancouver General Hospital, Vancouver, Canada

And with our unique AI research and development group in Princeton, our software development center in Bangalore, our product experts of CT in Forchheim, our customer collaboration partners in Paris and our own supercomputer we were able to enhance our AI portfolio with this algorithm, which works on CT images. We are stepping up as a partner to support healthcare systems helping them to deliver high-value care to patients and families and developing with us together the CT pneumonia analysis.

(5) Siemens Healthineers, Bangalore, India Software Development Center

(6) Siemens Healthineers, Forchheim, Germany CT chest Product Development and Manufacturing

(7) Siemens Healthineers, Princeton, NJ, USA Digital Technology and Innovation

(8) Siemens Healthineers, Paris, France AI collaborations

 

How could AI be beneficial for radiology in the context of COVID-19?

To improve COVID-19 therapy, there is an urgent need to increase global knowledge about this new coronavirus. With the inclusion of radiological findings in confirming COVID-19 diagnosis the workload of radiologists has increased. AI-powered analysis of chest scans has the potential to reduce this growing burden on radiologists, who must review and prioritize a rising number of patient chest scans.

Human efficiency and accuracy are compromised as a result of the overwhelming workload. The automation aspect of AI can offer support. AI can be a powerful aid in recognizing the lesions in CT images and even quantitatively characterizing the findings and comparing changes between exams, which is crucial for precision medicine. AI may support clinicians to answer several questions such as patient triaging, diagnosis (in combination with RT-PCR tests and epidemiological risk), assessment of severity and progression, and response to therapy in patients exhibiting COVID-19 symptoms.

 

How does the CT Pneumonia Analysis2 work?

The algorithm is designed to automatically identify and quantify abnormal patterns in the lungs, enabling simple to use analysis of non-contrasted chest CT scans for research purposes. The system identifies lungs, lobes and abnormalities associated with pneumonia. It also computes high opacity abnormalities as they were shown to correlate with severe symptoms. The results could be used to analyze the severity and the progression of abnormalities in patients exhibiting COVID-19 symptoms. First, the abnormal patterns in the lungs are automatically segmented in 3D.

Second, disease severity measures are computed: two global and two lobe-wise measures.

Percentage of Opacity (POO): percentage of predicted volume of abnormalities compared to the total lung volume.

Percentage of High Opacity (POHO): percentage of predicted high opacity volume compared to the predicted volume of abnormalities

Lung Severity Score (LSS): measures the extent of abnormalities across each lobe

Lung High Opacity Score (LHOS): measures the extent of abnormalities with high opacity for each lobe

The POO and LSS quantify the extent of lung involvement and the distribution of involvement across lobes, respectively. Given that high opacity abnormalities (i.e., consolidations) were shown to correlate with severe symptoms, two new severity measures POHO and LHOS were introduced to complement the assessment. They quantify the extent of high opacity abnormalities and their distribution of involvement across lobes, respectively. The performance of the method in estimating POO, LSS, POHO, and LHOS is evaluated on a database of 100 COVID-19 cases and 100 controls from multiple institutions from Canada, Europe, and the U.S. Ground truth is established by computing the same measures from manual annotations of the lesions, lungs, and lobes.

 

Learn more about system requirements and apply the COVID-19 CT pneumonia analysis in your research environment for free.

To make it as easy as possible for you to use our new algorithm, we provide it to you via all our reading and interpretation solutions you use already in your existing workflows.

 

syngo.via openApps CT pneumonia analysis: With OpenApps, syngo.via connects you to the Siemens  Helathineers Digital Marketplace and its innovative offerings. You can benefit from the integrated solution that is your new way to organize and discover apps from Siemens Healthineers and other vendors – directly on your syngo.via. Simply click on the shopping cart icon in the upper-right corner of your syngo.via VB30A or higher. A Frontier License is not needed to access the CT Pneumonia Analysis prototype.

syngo.via Frontier CT pneumonia analysis: synio.via Frontier allows you to strengthen your clinical opinion leadership with easy access to numerous post-processing prototypes that are seamlessly integrated with your routine syngo.via environment. The new CT Pneumonia Analysis2 prototype is included in the dedicated syngo.via Frontier Prototype Store, which is continuously enriched with new contributions.

AI-Rad Companion Research CT pneumonia2 analysis: The AI-RAD Companion, our family of AI-powered augmented workflow solutions, running on the teamplay digital health platform, helps to reduce the burden of basic repetitive tasks and may increase your diagnostic precision when interpreting medical images. Its solutions provide automatic post-processing of imaging datasets through AI-powered algorithms. The automation of routine workflows with repetitive tasks and high case volumes helps you to ease your daily workflow – so that you can focus on more critical issues.

 

Beyond research, AI algorithms may indicate signs of COVID-19 on chest X-rays embedded in the clinical workflow.

Artificial Intelligence plays an important role today and its support can contribute to radiologist’s work. The AI-Rad Companion Chest X-ray6, a new part of the AI-Rad Companion product family is another solution where deep learning is embedded in a dedicated software. The AI-Rad Companion Chest X-ray automatically characterizes radiographic findings for a lung X-ray (Upright patient position and PA direction). It works like a second or third reader to support radiologists in differential diagnosis and clinical decision-making. It is capable of characterizing pneumothorax, pulmonary lesions (nodules, masses, and granuloma), atelectasis, consolidation, and pleural effusion. Atelectasis and consolidations are findings that show a correlation with signs of pneumonia that is caused by the SARS-CoV-2 virus.

 

1Wilson N, Kvalsvig A, Barnard LT, Baker MG. Case-Fatality Risk Estimates for COVID-19 Calculated by Using a Lag Time for Fatality. Emerg Infect Dis. 2020;26(6).

2For Research Use Only. Not for use in diagnostic procedures.

3CDC. CDC Clinical Criteria [Internet]. 2020. Available from: https://www.cdc.gov/coronavirus/2019-nCoV/hcp/clinical-criteria.html, 20.04.2020.

4Kanne JP, Little BP, Chung JH, Elicker BM, Ketai LH. Essentials for radiologists on COVID-19: an update—radiology scientific expert panel. Radiology. 2020;200527.

5https://pubs.rsna.org/doi/10.1148/radiol.2020201365, 22.04.2020

6AI-Rad Companion Chest X-ray is currently not for sales in the United States and other selected countries

Leave a Reply

Your email address will not be published. Required fields are marked *