A Deep Learning-Based Model for Tactile Understanding on Haptic Data Percutaneous Needle Treatment

Abstract

Tactile understanding during surgery is essential in medical simulation. To improve a remote surgical operation one step further, in this paper, we develop a sequence classification technique, categorising different tissues, evaluating on biomechanics data. The importance of the proposed model is emphasised when problems such as a delay is occurring during simulation. Monitoring, predicting, and understanding the sense of tissue which is supposed to be involved in operation is vital during surgery. To achieve this, different deep structural techniques are investigated to find the effect of deep features for tactile and kinaesthetic understanding. The experimental results reveal that residual networks outperform others with respect to different terms. The results are accurate and fast which enables the technique to perform in real-time.