Under the guidance of "green city circle" concept, building the ecological security pattern of Wuhan Metropolitan Area contributes to realizing the synergistic development of regional economy, society, and ecology. The multi-indicator overlay analysis adopted in traditional studies loses the original information embedded in each factor and can only identify homogeneous ecological sources. This study incorporated a Self-Organizing Mapping Neural Network model (SOM) into the importance-sensitivity-connectivity framework, which could delineate ecologically functional areas based on ecosystem service portfolios and then identify multi-type ecological sources. After modifying the ecological resistance surface by integrating natural factors and human influence, the Minimum Cumulative Resistance model (MCR) was applied to form the multi-feature ecological security pattern. The main conclusions are as follows:there were two types of typical "high importance-low sensitivity-high connectivity" ecological sources in Wuhan Metropolitan Area, including 9756 km²grain production area (Xiaogan, Tianmen, Qianjiang, and Xiantao) and 6791 km²carbon sink area (northeast of Huanggang and southwest of Xianning), which should be regarded as the ecological security core area. Besides, low-connected grain production sources were located in the suburbs of Wuhan and Ezhou, and low-connected carbon sink sources were located in the southeast of Huangshi, which were suitable as ecological protection functional areas. Furthermore, 39 ecological corridors were identified, which could be divided into grain production connecting axis, carbon sink connecting axis and cross category connecting axis according to the source attributes at both ends. In summary, a multi-feature ecological security pattern of "One Ring, Three Wings and Multiple Connections" were established in Wuhan Metropolitan Area. This pattern organically integrated the concept of ecological function zoning, which could help the city circle to adopt appropriate protection and restoration strategies according to local conditions.