“I’ll Be Back”: This is Just the Beginning of the AI Revolution

With the release of ChatGPT in 2021, the focus on artificial intelligence (AI) in the worldwide zeitgeist has been around the marvels of generative AI. While generative AI is of course exciting (and a little Skynet terrifying), the workhorse of AI, especially in medical diagnostics, remains machine learning and decision tree analysis of clinical data. Utilizing large datasets to become adept at pattern recognition, these various models have the ability to greatly sharpen our diagnostic capabilities. In the accompanying article, “Machine Learning for Early Discrimination between Lung Cancer and Benign Nodules Using Routine Clinical and Laboratory Data”, Wei et al., utilizing two large datasets, tested several machine learning models in order to develop a model that could be effective in aiding in the diagnosis of lung cancer.¹ In the study, the authors, utilizing basic clinical and laboratory data from 2312 lung cancer patients and 653 control patients, tested several open-source machine learning models and were able to develop a model that was accurate in predicting lung cancer, both with respect to the initial diagnosis and likely stage. …