A New Artificial Intelligence Application Identifying Lung Cancer Type Has Been Discovered

Artificial intelligence holds great promise for cancer diagnosis and treatment. In September 2018, Nature Medicinepublished a study that developed an artificial intelligence program capable of distinguishing two different lung cancer types [adenocarcinoma and squamous cell carcinoma] with % 97 accuracy. The program can also identify gene mutations associated with lung cancer.

The research used artificial intelligence approaches that imitate connections between nerve cells in the human brain (Google Inception's deep learning with multilayer artificial neural networks). The researchers trained the program using cancer tissue images and genetic analyses from the Cancer Genome Project as the training data.

Related Topic:

• Types of lung cancer (in detail)
• The developed program successfully distinguished adenocarcinoma from squamous cell carcinoma with %97 accuracy.
• Lung cancer-associated genes (EGFR, KRAS and TP53) were identified in cancer tissues with accuracy reaching %86.

Related Topic:

• Treatment of lung cancer with targeted drugs (in detail)

What Does This Study Mean for Us?

• An artificial intelligence program capable of making accurate diagnoses at a pathologist's level was developed. The program also shows potential to reveal previously undiscovered gene mutations in cancer cells.
• Clinical use is being considered after its effectiveness is demonstrated in further clinical studies.
• All medical informatics and technology news