Pancreatic cancer is a lethal condition with poor outcomes and an increasing incidence, being the 4th leading cause of death by cancer in Europe with less than 5 months of average survival time after diagnosis. The lethal nature of this disease requires an early differential diagnosis of pancreatic cysts, which are identified in up to 16% of normal subjects, and some of which will evolve into pancreatic cancer. However, in autopsy studies, the prevalence of these pancreatic cysts indicates that it may be as high as 24.3% [1].
Radiologists' discrepancies on evaluation may result in an inconclusive diagnostic, which leads to the real need of implementing automatic quantitative image analysis that can accurately predict which of these lesions are likely to progress to pancreatic ductal adenocarcinoma.
At Sycai Technologies, we have developed an artificial intelligence (AI)-based tool able to detect, identify and classify pancreatic cystic lesions on abdominal computed tomography (CT) scans with a high degree of precision.
By integrating our tool into hospitals, all CT scans performed will be analyzed guaranteeing an early differential diagnosis of pancreatic cysts and enabling the identification of those lesions in which pancreatic abnormalities are not even expected, and that today go undetected. This is particularly important for this kind of lesions, as most of them are found by accident when performing tests to other organs.
Additionally, if a non-suspicious cyst is found, periodic tests must be repeated every 6-12 months. Therefore, an early classification of the pancreatic cysts, together with a computer-aided prediction model, will avoid unnecessary CT scans and will reduce the waiting lists.
[1] Kimura, W. et al. Analysis of small cystic lesions of the pancreas. Int J Pancreatol. (1995)