Micro force sensing in various clinical scenarios is a challenging issue to be addressed. It is highly difficult to trade off the size, cost, and measurement accuracy of a micro force sensing system. In this article, a compact and affordable micro force sensing system enhanced by deep neural network is proposed. A three-axis force sensor is designed and fabricated with a footprint of only 14 mm and is employed to transform the force on the material structure into more accurate distance information. Such a sensor configuration can be seamlessly interfaced with the distal end of our in-house compliant and flexible continuum robot. On top of that, the RNN-LSTM network is exploited to augment the micro force sensing capability of the distal end-effector of the robot, which addresses the limitation on the nonlinear force issue of the continuum robot and the material itself. The RNN-LSTM network alone can be employed to perform force curve fitting for specific interventional tasks. The results indicate that more than 90% accuracy has been achieved, and the network can be applied to large-scale continuum robot-assisted interventional scenario deployment and teleoperation force perception.