- Artificial intelligence and machine learning in healthcare can help in disease prevention, diagnosis, as well as in developing customised treatments.
- The current COVID-19 pandemic has sparked new research to use artificial intelligence (AI) for the prediction of oxygen that is required for COVID-19 patients.
- The research was conducted to build an AI tool and a federated learning technique was used as an algorithm.
Pandemics are a serious threat, and the current COVID-19 pandemic is not the first and also may not be the last. Artificial intelligence and machine learning play a significant role in the healthcare sector in fighting the current pandemic and prepare for the upcoming ones. There are many ways machine learning helped combating the COVID-19 pandemic including risk identification, patient diagnosis, diagnosis of existing treatments, prediction on disease spread, predict upcoming pandemic and many more.
Source: © Thekaikoro | Megapixl.com
The current COVID-19 pandemic has triggered new research to use artificial intelligence (AI) for the prediction of oxygen that is required for COVID-19 patients on a global scale. Scientists have successfully developed world’s first artificial intelligence tool that will predict the need of oxygen requirement by using federated learning (FL) technique, which is used to train AI models with data from multiple sources.
RELATED ARTICLE: Telemedicine: A Game-Changer to Combat COVID-19
Let us discuss in detail-
AI tool to predict oxygen requirement
Cambridge-based internationally renowned hospital Addenbrooke’s Hospital, including 20 other worldwide hospitals and healthcare technology leader, NVIDIA, have used AI tool for predicting the oxygen needs of COVID-19 patients on a global scale.
The research was conducted to build an AI tool for the prediction of how much extra oxygen a COVID-19 patient may require in the first days of hospital care. The data from across four continents was used for the study. The federated learning technique was used as an algorithm to investigate chest x-rays and electronic health records of COVID-19 patients from the hospital.
Furthermore, for maintaining patient confidentiality, the patient data was fully anonymised, and an algorithm was sent to each hospital; therefore, no data was shared or left its location.
After determining the algorithm from the data, the research was conducted together to build an AI-tool that could predict the oxygen demands of hospitalised COVID-19 patients worldwide.
The study was named EXAM, which is EMR (electronic medical record)- CXR (chest X-ray)-AI-Model. The study was supported by the NIHR (National Institute for Health Research) Cambridge Biomedical Research Centre (BRC).
The outcomes of ~10,000 COVID-19 patients from worldwide were analysed in the study, including 250 patients who came to Addenbrooke’s Hospital during the first wave of the COVID-19 pandemic in March/April 2020.
With the collaborators across Asia, Europe, North and South America, the EXAM study took two-weeks of artificial intelligence learning to achieve high-quality predictions.
Overall, the use of artificial intelligence and machine learning is transforming the healthcare industry. AI-based prediction models combined with several features to estimate the risk of infection have been developed, which will play a significant role to combat the ongoing COVID-19.