When confronted with a real-world problem, heuristic knowledge and experience can guide the solution of a specific technical problem as the key step toward innovation. In particular, a heuristic prototype must be used correctly to cue the technical problem that exists in a particular situation. The present study selected an innovative paradigm and scientific innovation materials to investigate the neural basis of insight induced by heuristic prototypes using event-related functional magnetic resonance imaging (fMRI). The day prior to undergoing fMRI scanning, participants were asked to solve 42 difficult technical problems that scientists might have already encountered but were unknown to the participants. In the subsequent fMRI experiment, the same participants were randomly presented with 84 prototypes classified into two types: related prototypes (RPs), which were useful for solving previously encountered problems, and unrelated prototypes (UPs), which sometimes did not contribute to problem solving. While being scanned, participants were asked to assess whether a prototype is relevant to any of the technical problems. This study comprised two conditions: solving technical problems when presented with a related heuristic prototype and failing to solve technical problems using unrelated heuristic prototypes. The authors assumed that the regions significantly activated by the RP condition, compared with the UP condition, reflected brain activity related to the role of heuristic prototypes in scientific insight. fMRI data showed that the left dorsolateral prefrontal gyrus (left DLFPC, BA9) and the left angular gyrus (left AG, BA39) were more significantly activated when presented with RPs than with UPs. The results suggest that the DLPFC may be involved in the automatic retrieval of technical problems and breaking of mental sets. Moreover, the left AG may be involved in forming novel associations between technical problems and related prototypes.