Using Artificial Intelligence to Diagnose Mesothelioma

Scientists from India are working on building a computer program to help diagnose mesothelioma using artificial intelligence.

Diagnosing mesothelioma can be challenging for doctors because it is so rare. Mesothelioma symptoms can also look like symptoms of more common diseases like the flu.

The scientists in this study are hoping to use artificial intelligence and machine learning to help doctors reach an accurate mesothelioma diagnosis more quickly.

Diagnosis and Showing Symptoms

Mesothelioma is a rare cancer caused by exposure to asbestos. It is usually diagnosed after a person goes to a doctor because they are experiencing symptoms.

These symptoms depend on the type of mesothelioma. For mesothelioma in the lungs, these symptoms could include shortness of breath, pain in chest, and fluid retention. For mesothelioma in the abdomen, a person might experience abdominal pain, fatigue, and excess fluid in the abdomen.

There is a long latency period between exposure to asbestos and a mesothelioma diagnosis. This period can be as long as 40 or 50 years. Once a person starts to show symptoms of mesothelioma, it has usually progressed to an advanced stage and is difficult to treat.

Because of the long latency period and aggressive nature of mesothelioma, early diagnosis is key to survival. That’s where artificial intelligence and machine learning come in.

Building a Computer Program

Machine learning is a part of artificial intelligence that uses algorithms to look for patterns in data and make predictions.

The scientists in this study used mesothelioma patient information to test different types of algorithms for predicting a mesothelioma diagnosis.

They found that some machine learning algorithms did a good job of identifying mesothelioma, like a method called “support vector machine”. Other algorithms they tested did not do a good job because they took too much time or were not accurate.

This type of research is cutting-edge, and we need a lot more research to build computer programs that will be helpful for doctors. The scientists in this study hope that more research will “allow for the early diagnosis and consequently, an expected lower mortality rate for [malignant mesothelioma]” in the future.

Source

Kapila D, Panwar S, Raja MKMM, et al. Applications of Neural Network-Based Plan-Cancer Method for Primary Diagnosis of Mesothelioma Cancer. Biomed Res Int. 2023;2023:3164166. Published 2023 Feb 4. doi:10.1155/2023/3164166. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9922178/