Artificial intelligence models were able to predict lung cancer risk based on three variables.
ADVERTISEMENTArtificial intelligence (AI) models were found to be as good or better at predicting a person’s risk of lung cancer compared to the best models available, a new study showed.
Researchers from University College London and the University of Cambridge in the UK used data from the UK Biobank and the US National Lung Screening Trial to develop models to predict the risk of a person getting lung cancer in the next five years.
They used datasets to experiment with more than 60 different machine learning models to see which was the most effective at predicting a person’s risk for the disease.
Of these, they combined four and were able to predict lung cancer risk with the same or improved accuracy compared to the best available models. The findings were published in the journal PLOS Medicine on Tuesday.
The models were fed just three variables: a person’s age, how many years they smoked, and the average number of cigarettes smoked per day.
“Our study shows that artificial intelligence can be used to accurately predict lung cancer risk using just three pieces of information that would be easy to gather during routine GP appointments, online or via apps,” said Dr Tom Callender, from University College London’s School of Medicine, in a statement.
“This approach has the potential to greatly simplify population level screening for lung cancer and help to make it a reality”.
Lung cancer is the leading cause of cancer deaths worldwide, according to the World Health Organization (WHO), and smoking is the leading cause of lung cancer, accounting for 85 per cent of all cases.
It caused an estimated 1.8 million deaths in 2020, WHO says.
Symptoms of lung cancer include chest pain, shortness of breath, a cough that doesn’t go away, coughing up blood, fatigue, weight loss, and recurrent lung infections.
Detecting lung cancer in the early stages of the disease can lead to a better outcome as there are more effective treatments available.
In 2020, the UK Lung Cancer Coalition said that the most common route for diagnosing lung cancer remained emergency hospital admission, despite this being too late.
The UK government announced in June that they would roll out a targeted national lung cancer screening programme to help detect cancer sooner. People aged 55 to 74 with a history of smoking will be invited for screenings.
The European Union updated its cancer recommendations last year, emphasising that countries should target high-risk profiles, including heavy smokers and ex-smokers who used to smoke heavily.
The researchers state in the study that the risk assessment for lung cancer screening can be simplified without reducing performance. This could improve the effectiveness of national screening programmes.
“This research is a prime example of how machine learning tools such as AutoPrognosis, combined with innovative clinical researchers, can make a real impact in healthcare at a population level,” said Mihaela van der Schaar, an author of the study from the University of Cambridge.
One limitation of the study was that it was based on retrospective data from the UK and the US, so other data from more regions should be considered, the researchers added.
Other research on using AI to help with early detection of lung cancer has looked at using these tools to help interpret CT scans of the chest and look for biological markers or gene mutations.
