Epidemiological modelling and epidemic threshold analysis in the networks are widely used for the control and prediction of infectious disease spread. Therefore, the prediction of the epidemic threshold in networks is a challenge in epidemiology where the contact network structure fundamentally influences the dynamics of the spread. In this paper, we design and experiment a new general structural and spectral prediction approach of the epidemic threshold. This more captures the full network structure using the number of nodes, the spectral radius, and the energy of graph. With data analytic and data visualization technics, we drive simulations overall on 31 different types and topologies networks. The simulations show similar qualitative and quantitative results between the new structural prediction approach of the epidemic threshold values compared to the earlier MF, HMF and QMF widely used benchmark approaches. The results show that the new approach is similar to the earlier one, further captures the full network structure, and is also accurate. The new approach offers a new general structural and spectral area to analyse the spreading processes in a network. The results are both fundamental and practical interest in improving the control and prediction of spreading processes in networks. So these results can be particularly significant to advise an effective epidemiological control policy.