An artificial intelligence (AI) algorithm based on 29 different blood parameters may help boost screening for early gastric cancer.
The lack of symptoms in the early stage of gastric cancer leads to delayed clinical presentation and low overall survival rate. Invasive screening, such as esophagogastric duodenoscopy, is effective in decreasing gastric cancer mortality, but adherence is low. Thus, there is a need for minimal or non-invasive gastric cancer screening via routine blood tests, for example. Previous algorithms based on routine blood tests were developed for the prediction of colorectal cancer. However, low sensitivity, specificity, and positive predictive value limited their use.
Researchers from Hong Kong developed a deep learning AI algorithm using data from routine blood tests from more than 190,000 individuals prescribed medications for dyspepsia between 2004 and 2015. A total of 4,790 diagnoses of gastric cancer were identified. The blood tests contained 29 parameters, including complete blood counts, liver function parameters, renal function parameters, and clotting function parameters. Tsz Chun Bryan Wong presented the results of this analysis at the 2023 ASCO Annual Meeting, held June 2-6 in Chicago.
Data from 2004 to 2009 and from 2011 to 2014 were used as a training cohort and the rest were used as a testing cohort. A blood gastric cancer signature was generated using deep-learning AI algorithms. After training and testing, the signature was able to predict gastric cancer with high sensitivity (96%), high specificity (100%), high positive predictive value (99%), and high negative predictive value (100%).
In two validation cohorts (N=24,610 and N=17,058) the signature proved to be highly accurate: 80% sensitivity, 100% specificity, 96% positive predictive value, and 100% negative predictive value.
Based on these outcomes, Wong concluded that “the signature has high accuracy and high sensitivity with low false positive and false negative rates. Therefore, the signature can be used to enhance cost-effectiveness and public participation rates in screening for early gastric cancer.”
Copyright ©2023 Medicom Medical Publishers
