IA-QCNN

The article presents IA-QCNN as a compact and computationally efficient alternative for MRI-based MGMT prediction in GBM. The study claims QCNNs had not previously been directly applied to MGMT promoter methylation prediction from GBM MRI. IA-QCNN combines energy-based slice selection, importance-aware quantum feature weighting, and ring-topology quantum convolution with folding-based pooling. IA-QCNN is proposed as a specialized quantum convolutional neural network for non-invasive MGMT promoter methylation prediction from GBM MRI.