AI model accurately identifies cancer development on CT scans

The development of the AI model

The development of the AI algorithm involved utilizing CT scans of roughly 500 patients who had large lung nodules. The next stage of testing involves evaluating the accuracy of the technology on patients with large lung nodules in a clinical setting to predict their risk of lung cancer. 

Additionally, the AI model has the potential to assist doctors in making faster decisions regarding patients with growths that are currently considered medium-risk. When combined with the Herder test, the AI model was capable of identifying high-risk patients within this group.

The AI tool has been found to perform more effectively and efficiently than current cancer diagnosis methods, according to the study published in the Lancet's eBioMedicine journal. The researchers believe that this AI tool could potentially enhance the early detection of cancer and increase the success of treatment by identifying high-risk patients and fast-tracking them for earlier intervention. 

Potential benefits of the AI tool

Cancer, which is responsible for almost one in six deaths globally, leads to approximately 10 million deaths every year, as per the World Health Organization. Lung cancer is the leading cause of cancer-related deaths worldwide and is responsible for one-fifth (21 percent) of cancer fatalities in the UK. Early diagnosis of the disease leads to more effective treatment outcomes, however, recent statistics indicate that over 60 percent of lung cancers in England are diagnosed at stage three or four. Therefore, the experts involved in this study point out that this underscores the urgent need for new initiatives to accelerate the detection of lung cancer.

With the aim of accelerating cancer detection by expediting the treatment process and streamlining CT scan analysis, researchers anticipate that the AI tool will achieve this goal in due course. However, the team emphasized that the study is in its nascent stages and more comprehensive testing will be necessary before implementing the model into healthcare systems.