arXiv subjects

Dag Johansen

Publications and source records attributed to Dag Johansen.

23 records · Page 2Linked to original sources

arXiv2019

ResUNet++: An Advanced Architecture for Medical Image Segmentation

Accurate computer-aided polyp detection and segmentation during colonoscopy examinations can help endoscopists resect abnormal tissue and thereby decrease chances of polyps growing into cancer. Towards developing a fully automated model for pixel-wise polyp segmentation, we propose ResUNet++, which is an improved ResUNet architecture for colonoscopic image segmentation. Our experimental evaluations show that the suggested architecture produces good segmentation results on publicly available datasets. Furthermore, ResUNet++ significantly outperforms U-Net and ResUNet, two key state-of-the-art deep learning architectures, by achieving high evaluation scores with a dice coefficient of 81.33%, and a mean Intersection over Union (mIoU) of 79.27% for the Kvasir-SEG dataset and a dice coefficient of 79.55%, and a mIoU of 79.62% with CVC-612 dataset.

eess.IV↗

arXiv2019

Kvasir-SEG: A Segmented Polyp Dataset

Pixel-wise image segmentation is a highly demanding task in medical-image analysis. In practice, it is difficult to find annotated medical images with corresponding segmentation masks. In this paper, we present Kvasir-SEG: an open-access dataset of gastrointestinal polyp images and corresponding segmentation masks, manually annotated by a medical doctor and then verified by an experienced gastroenterologist. Moreover, we also generated the bounding boxes of the polyp regions with the help of segmentation masks. We demonstrate the use of our dataset with a traditional segmentation approach and a modern deep-learning based Convolutional Neural Network (CNN) approach. The dataset will be of value for researchers to reproduce results and compare methods. By adding segmentation masks to the Kvasir dataset, which only provide frame-wise annotations, we enable multimedia and computer vision researchers to contribute in the field of polyp segmentation and automatic analysis of colonoscopy images.

eess.IV↗

arXiv2019

A Survey of Automatic Methods for Nutritional Assessment

Nutritional assessment is key in order to make decisions about the nature and cause of nutrition related health issues that affect an individual. The systematic process of collecting and interpreting relevant nutrition information, however, is still in its technological infancy. Despite technological advances in storage and analysis of nutritional data, methods for collecting data are largely unchanged over the past two decades. It is well documented that these methods have issues that cause under-reporting. Meanwhile, new developments in wearable biometric logging devices have seen increased traction among individuals. This is sometimes referred to as the Quantified Self movement. One part of this movement is the development of technological means for objectively collecting nutritional data. Nutritional assessment, however, is about to be heavily impacted by emerging computer science technologies, and this survey provides an overview of promising technology approaches supporting nutritional assessment. Both academic and commercial systems are reviewed and categorized.

cs.CY↗

arXiv2018

The Medico-Task 2018: Disease Detection in the Gastrointestinal Tract using Global Features and Deep Learning

In this paper, we present our approach for the 2018 Medico Task classifying diseases in the gastrointestinal tract. We have proposed a system based on global features and deep neural networks. The best approach combines two neural networks, and the reproducible experimental results signify the efficiency of the proposed model with an accuracy rate of 95.80%, a precision of 95.87%, and an F1-score of 95.80%.

cs.LG↗

arXiv2012

Secure Abstraction with Code Capabilities

We propose embedding executable code fragments in cryptographically protected capabilities to enable flexible discretionary access control in cloud-like computing infrastructures. We are developing this as part of a sports analytics application that runs on a federation of public and enterprise clouds. The capability mechanism is implemented completely in user space. Using a novel combination of X.509 certificates and Javscript code, the capabilities support restricted delegation, confinement, revocation, and rights amplification for secure abstraction.

cs.CR↗