A machine learning framework based on a combination of a 1-dimensional U-net and random forests was developed to find consistent patterns in the sensor data characteristic of needle insertions. This ...