Scientists Train AI To Detect Illness In Human Breath

AI To Detect Illness In Human Breath

alfcermed / Pixabay

Unique capabilities of artificial intelligence are getting better as each day passes. A group of researchers from Loughborough University, Western General Hospital, the University of Edinburgh, and the Edinburgh Cancer Centre in the UK have succeeded at training AI to detect illness in human breath. The machine-learning method successfully analyzes compounds in human breath and can even detect terminal diseases like cancer.

“The sense of smell is used by animals and even plants to identify hundreds of different substances that float in the air. But compared to that of other animals, the human sense of smell is far less developed and certainly not used to carry out daily activities,” researcher Andrea Soltoggio wrote on Smithsonian.com. “For this reason, humans aren’t particularly aware of the richness of information that can be transmitted through the air, and can be perceived by a highly sensitive olfactory system.”

The group used NVIDIA Tesla GPU’s to train AI to detect illness in human breath, and the cuDNN-accelerated Keras, and TensorFlow deep learning frameworks. The neural network used data from participants that had different types of cancer and were receiving radiotherapy, Angelika Skarysz, a PhD research student at Loughborough University revealed as per developer.nvidia.

The team also decided to use data augmentation in order to make the neural network more efficient, by increasing the data. According to the team the algorithm was augmented 100 times.

“This is the first successful machine learning attempt at learning ion patterns and detecting compounds from raw GC-MS data,” the team said. “The convolutional neural network achieved the best performance when implemented with two particular features: one-dimensional filters to adapt to the particular structure of GC-MS data, and a three-channel input to read high, medium, and low-intensity signals from the highly variable GC-MS spectrum. The novel approach was shown to discover labelling errors from human experts, suggesting better-than-human average performance.” the researchers were quoted by developer.nvidia.

The researchers also used NVIDA’s GPU to test interference, which in this situation, was scanning different breath samples.

If the neural network develops sufficiently scientists can use AI to detect illnesses in human breath and treat them. The entire project has a lot of potential, although it could turn out to be controversial. The researchers suggest that the system can be used to find substances in the air that could be potentially harmful. Still, it doesn’t necessarily man it will diagnose or “make a decision.” Doctors can use the data it gathered and with the help of this information, make their conclusions.




About the Author

Danica Simic
Danica Simic has been writing ever since she was a child. Before she started writing for ValueWalk she was reviewing laptops, headphones and gaming equipment as well as writing articles about astronomy and game development. Danica is a student of applied and computational physics while also studying software and data engineering. Her hobbies include reading, swimming, drawing and gaming whenever she has free time. - Email her at dsimic@valuewalk.com