Technology is rising to the challenge of helping us detect signs of breast cancer as early as possible. A team of researchers from South Korea has developed a device that, in tandem with a specialized AI algorithm, can detect early breast cancer with a 96% accuracy. The study was carried out with data obtained from 201 breast mass biopsies, out of which 66 malignant and 66 benign masses were used for the study.
It relies on a tactile sensor that can measure the hardness or softness of a given object. The team reports this is the first time an accurate tactile sensor has ever been developed.
A mechanical touch
"Humans can easily distinguish the hardness level [of an object], but it is very difficult to make a device that can distinguish hardness," explains Lee Hyun-Jung, Vice President of and Senior Research Fellow at the Korea Institute of Science and Technology in a press video. "However, it is significant that we have further developed a technology that can quantify [this property]."
The sensor itself looks unassuming -- like an ordinary computer chip. However, when exposed to pressure, this sensor generates an electrical current of different frequencies. This signal can then be interpreted to infer how much pressure it is experiencing. If the force being applied to the sensor to press it against an object is known, this pressure serves as a proxy for the hardness of the object.
The team's work is significant as they were able to study how this frequency changes in relation to the hardness of an object, and create software that is able to interpret it. This latter part of the process is handled by an artificial intelligence algorithm.
Put together, they explain, these two elements can help detect breast cancers early, pain-free, and with very high accuracy. That being said, the team is confident that other types of cancer can be diagnosed using this tool, and that it can further help in accurately mapping the boundary of tumors in robotic surgery, where doctors cannot or should not touch the surgical site to determine this themselves.
The paper "An Artificial Tactile Neuron Enabling Spiking Representation of Stiffness and Disease Diagnosis" has been published in the journal Advanced Materials.
