In this talk, I will introduce a general quantification framework of ncRNA developed by us recently. For cancer diagnosis including liver cancer, it is well known that the early detection of cancer has been gradually by the initial diagnosis of cancer cell surface proteins to precancerous factors intro-cellular. Around the detection of miR-484(a precancerous biomarker of liver cancer), we developed an artificial intelligent molecular amplification system (AIMA) for visualization of miR-484 in vivo. Based on convolution neural network, the optimal AIMA system can be selected. Using the above reconstruction method, we also developed an vivo quantitative framework of miR-484 and could accurately calculate its concentration and 3D distribution.