Zhang et al. introduced a new and non-invasive method of utilizing chest CT scan and machine learning to predict impending lung metastasis earlier than traditional visual assessment by professionals. The researchers analyzed follow-up CT scans from 2,148 patients with breast, colorectal, and esophageal cancers. They developed a robust predictive machine learning model which used signaling features from 3D lung regions. Their results demonstrated that a positive signal from the model created had a significantly higher actual lung metastasis risk across all cancer types as compared to a negative signal patient. Moreover, the model detected these specific signals much before metastasis was visible on standard CT images by approximately0.8 to 1.4 years across the three tumor types. This demonstrates the possible future usage of this model to identify biological changes in the lung before significant visible metastasis to allow for quicker clinical intervention.