The research on crosstalk simulation using integer-order memristive synapses has made great progress recently. However, most existing models still adopt a single-memristor structure, which constrains synaptic weight modulation and makes it difficult to represent both excitatory and inhibitory synaptic connections in a unified manner. These models also often fail to capture the inherent memory effects and non-local dynamic properties of biological neurons. To address these issues, we introduce a fractional-order memristive bridge synapse model for crosstalk coupling in this study. Combining Hindmarsh-Rose (HR) and FitzHugh-Nagumo (FN) neurons and using fractional calculus, we construct an 8D heterogeneous coupled neural network, which is termed the fractional-order memristive bridge crosstalk-coupled neural network (FMBCCNN). Here, a major innovation is the integration of a fractional-order memristive bridge structure that mimics synaptic connections in a bridge configuration. This design can achieve both historical memory characteristics and bidirectional synaptic weight regulation, thereby overcoming limitations of traditional coupling forms.

Using dynamical analysis tools such as phase portraits, bifurcation diagrams, and Lyapunov exponents, we systematically investigate how synaptic and crosstalk strengths influence system behavior under traditional fractional-order condition. The results reveal diverse dynamical behaviors, including attractor coexistence, forward and reverse period-doubling bifurcations, and chaotic crises. Further analysis shows that the system maintains continuous periodic motion over broader parameter range and exhibits clear parameter hysteresis under the more generalized condition of non-uniform fractional orders compared with those under the traditional condition. Although local dynamic patterns remain similar, the corresponding parameter intervals are substantially widened. In addition, the system displays more intensive and obvious alternation phenomenon between periodic and chaotic behaviors. We also simulate the effect of varying the fractional-order derivative, offering a more general mathematical characterization of neuronal firing activity.

Finally, the chaotic sequences generated by the system are applied to an image encryption algorithm combined with bit-plane decomposition and DNA encoding. Security analysis confirms that the encrypted images have pixel correlation coefficients all below 0.01 in horizontal, vertical, and diagonal directions, information entropy greater than 7.999, and a key space of 2²⁰⁸⁰. These results verify the excellent encryption performance and reliability of the proposed scheme and the generated sequences.