APPSO‐NN: An adaptive‐probability particle swarm optimization neural network for sensorineural hearing loss detection

As a hearing disorder, sensorineural loss (SNHL) can be effectively detected by magnetic resonance imaging (MRI). However, the manual detection of MRI scanning is subjective, time-consuming, and unpredictable. An accurate automatic computer-aided diagnosis system proposed for SNHL detection, providing reliable references professionals. The first employs wavelet entropy layer to extract features images. Then, neural network as classifier consisting feedforward (FNN) an adaptive-probability PSO (APPSO) algorithm. authors prove rotation-variant property basic particle swarm optimization (PSO) algebraic matrix transformation. unsuitable optimising parameters networks. Thus, in APPSO, integrate new update rules based on all-dimensional variation mechanism into PSO, which improve its searching ability without losing population diversity. compare APPSO-NN with FNN trained five popular evolutionary algorithms. simulation results show that APPSO performs best training FNN. method also compares six state-of-the-art methods. performance sensitivity overall accuracy classification, proves effective promising detection.