# FocusNet: An attention-based Fully Convolutional Network for Medical Image Segmentation

## TL;DR

FocusNet introduces an attention mechanism within convolutional neural networks using autoencoder-generated feature maps, achieving competitive results in medical image segmentation tasks like skin cancer and lung lesion detection.

## Contribution

The paper presents a novel attention integration method in CNNs using autoencoder features, enhancing segmentation performance in medical imaging.

## Key findings

- Achieves highly competitive segmentation accuracy.
- Effective in skin cancer and lung lesion datasets.
- Outperforms or matches U-Net variants.

## Abstract

We propose a novel technique to incorporate attention within convolutional neural networks using feature maps generated by a separate convolutional autoencoder. Our attention architecture is well suited for incorporation with deep convolutional networks. We evaluate our model on benchmark segmentation datasets in skin cancer segmentation and lung lesion segmentation. Results show highly competitive performance when compared with U-Net and it's residual variant.