Prediction and early identification of disease through artificial intelligence (AI)
The potential role of AI-based predictive models in healthcare
More than 70% of today’s medical decisions involve the results of laboratory tests. These tests may also hold the key to earlier identification of patients at risk from complex diseases such as cancer, liver disease, and COVID-19. Because early signs of disease are often evident in laboratory test results, predictive models that leverage AI technology could help identify areas of concern, more likely before any noticeable physical symptoms appear.
By integrating AI into the laboratory data workflow, routine lab results could be combined with other relevant patient information such as age, gender, etc., for use within disease-specific predictive models. By combining this information, labs have the potential to generate disease-specific patient probability scores to help alert physicians to areas of concern and/or potential patient risk or diagnosis. In collaboration with several healthcare institutions, Siemens Healthineers is actively leveraging machine learning and computerized reasoning in the development of AI-driven clinical decision support tools that can be potentially integrated into the existing test-order/result-review workflow.
COVID-19 Severity Algorithm
Access a fully functional Educational Use Only version of the algorithm.
Leveraging AI to help identify potential disease progression in COVID-19 patients
In 2021, Siemens Healthineers partnered with several leading healthcare institutions across the globe to develop an AI-based predictive model. By aggregating deidentified COVID-19 patient data from more than 14,500 COVID-19 patients and leveraging deep machine learning, we created a predictive model using various clinical, demographic, and laboratory data. Based on a potential patient’s lab values and age, the Atellica® COVID-19 Severity Algorithm1 generates a COVID-19 clinical severity score, including projected probability of ventilator use, end-stage organ damage, and 30-day in-hospital mortality.
“After seeing firsthand how seamlessly we were able to integrate the predictive model into our daily laboratory workflow with Atellica Data Manager, I’m confident the algorithm can become an integrated decision support tool that will expand the lab’s contribution to physicians and ultimately aid in critical decision making for enhanced patient care.”
Head of Laboratory Medicine, Hospital Universitario, La Paz, Madrid, Spain
Liver Disease Severity Algorithm
Several factors can lead to liver disease. Excessive alcohol consumption, obesity, diabetes, hepatitis infections, and excessive consumption of medication could all contribute to an inflamed, and eventually fibrotic, liver.
Potential for a predictive model to aid in the early identification of liver disease
Liver disease is potentially curable if identified early and treated appropriately. However, this disease often goes unnoticed until a liver transplant is the only option. An AI-based predictive model to help identify patients at risk of severe liver disease could play a crucial role in early diagnosis of liver cancer. In association with other clinically relevant information, a predictive model could potentially enable early intervention and help to avoid progression to cirrhosis, liver failure, the need for liver transplant, and even mortality.
“The development of simple algorithms to detect clinically significant NASH in at-risk populations would be a major step in tackling this public health burden. By making predictive models like these available within the workflow and integrated into the electronic medical record system, clinicians could better manage therapeutic protocols and more quickly triage patients with more advanced disease to hepatologists.”
Professor of Medicine, Virginia Commonwealth University School of Medicine and Co-head of the Stravitz-Sanyal Institute
Cancer Predictive Algorithm
Potential for predictive models for early identification of cancer risk
In the U.S. alone, it is estimated that there will be 1.9 million new cancer diagnoses and 609,360 cancer deaths in 2022.2 Through earlier assessment of a patient’s cancer risk, combined with other clinically relevant information, AI-based predictive models using routine blood tests have the potential to help physicians more quickly diagnose and deliver effective treatment for cancer patients. In partnership with several leading organizations, we are actively working to create predictive models and ultimately to create a world without fear of cancer.
Related Information
Did this information help you?
1. For educational purposes only. Not for clinical or patient care, diagnosis, treatment, or to cure or prevent any disease. Availability varies by country.
The statements by Siemens Healthineers customers described herein are based on results that were achieved in the customer's unique setting. Because there is no “typical” hospital or laboratory and many variables exist (e.g., hospital size, samples mix, case mix, level of IT and/or automation adoption) there can be no guarantee that other customers will achieve the same results.