Aim: Brain wave detection can predict the level of neurological recovery in cardiac arrest patients. The aim of this study was to develop a method to detect brain waves to determine the level of neurological recovery in patients with cardiac arrest. Methods: We constructed a transformer architecture with a learnable filter, which was embedded between the embedding layer and the encoder layer of the transformer. The parameters of the learnable filter were optimized using the Adaptive Moment Estimate (Adam) algorithm, which adaptive reduces noise. The transformer encoder module was then able to capture the long-term dependencies of the PCG signal, allowing the network to learn more effective features from the input signal. Finally, the softmax function was used in the output layer to obtain the discrete probabilities of CPC appearing from 1 to 5, and the final prediction was made based on these probabilities. Results: Our (CQUPT_MH_Team) method uses the datasets of the challenge, 0.06 for 12 hours, 0.22 for 24 hours, 0.13 for 48 hours, and 0.33 for 72 hours. Conclusion: The results are quite impressive, and there is still a lot of room for improvement.