Focusnet++: Attentive Aggregated Transformations for Efficient and Accurate Medical Image Segmentation

Chaitanya Kaul, University of Glasgow
Murray-Smith Roderick, University of Glasgow
Nick Pears, University of York, England
Hang Dai, Mohamed bin Zayed University of Artificial Intelligence
Suresh Manandhar, NAAMII, Nepal

IR Deposit conditions: non-described


We propose a new residual block for convolutional neural networks and demonstrate its state-of-the-art performance in medical image segmentation. We combine attention mechanisms with group convolutions to create our group attention mechanism, which forms the fundamental building block of our network, FocusNet++. We employ a hybrid loss based on balanced cross entropy, Tversky loss and the adaptive logarithmic loss to enhance the performance along with fast convergence. Our results show that FocusNet++ achieves state-of-the-art results across various benchmark metrics for the ISIC 2018 melanoma segmentation and the cell nuclei segmentation datasets with fewer parameters and FLOPs.