Some tumors are difficult to detect and make later surgical procedures painful. In order to improve the accuracy of diagnoses of these ailments, researchers have developed a new approach based on machine learning technology. According to the test results, this new method is much more efficient and more advanced than those used before.
Indeed, the new technique that has been developed is able to accurately detect the rigidity of tumours, whether benign or malignant. The researchers behind this discovery do not intend to stop there and are planning future improvements very soon.
The doctor Hyunjung Yi of the spin convergence research center and the director suyoun lee, from the Center for Neuromorphic Engineering, participated in the study. According to the inventors of this technology, it could be used in other fields.
The incredible precision of the new mechanism of tactile neurons
Unlike previously developed artificial touch neuron approaches, this device is able to assess the consistency of bodies. The new technology is based on neuromorphics, which is a field of research aimed at emulate the human brain’s means of handling stimuli. This recent approach makes it possible to instantly process complex and heavy data suitable for artificial intelligence.
Furthermore, sensory axons pick up external stimuli through receptors and transform them into electrical spikes. The generated graphic diagram therefore changes according to the external data received. Then, the researchers developed an artificial tactile axon mechanism with a basic structure.
This approach combines a pressure receptor and an ovonic threshold switch apparatus to create sensory neural features. The device allows to evaluate in a faster and more precise way, the consistency of the bodies. Therefore, the tests were able to detect breast tumors with a success rate of around 95.8%.
Finally, for better use of the device, the scientists used elastography images of malignant and benign breast tumours.
Ultrasound elastography is still in its infancy
The developed artificial touch axon approach would be able to find and assimilate mechanical properties faster in the years to come. In addition, with the help of follow-up studies, it would even be possible to resolve the concern of the refraction of noise, specific to ultrasound elastography. The researchers also hope the device will prove useful in diagnosing low-power ailments.
Otherwise, applications in the field of robotic surgery where a surgical site needs to be quickly targeted are possible.
SOURCE: MIRA NEWS