A survey on deep learning in medical image analysis

“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

https://bura.brunel.ac.uk/bitstream/2438/14221/1/FullText.pdf

A survey of deep neural network architectures and their applications

MUCKE aims to mine a large volume of images, to structure them conceptually and to use this conceptual structuring in order to improve large-scale image retrieval. The last decade witnessed important progress concerning low-level image representations. However, there are a number problems which need to be solved in order to unleash the full potential of image mining in applications. The central problem with low-level representations is the mismatch between them and the human interpretation of image content. This problem can be instantiated, for instance, by the incapability of existing descriptors to capture spatial relationships between the concepts represented or by their incapability to convey an explanation of why two images are similar in a content-based image retrieval framework. We start by assessing existing local descriptors for image classification and by proposing to use co-occurrence matrices to better capture spatial relationships in images. The main focus in MUCKE is on cleaning large scale Web image corpora and on proposing image representations which are closer to the human interpretation of images. Consequently, we introduce methods which tackle these two problems and compare results to state of the art methods. Note: some aspects of this deliverable are withheld at this time as they are pending review. Please contact the authors for a preview.

http://ieeexplore.ieee.org/iel2/4524/12820/00592460.pdf

Image Processing

A scheme is developed for classifying the types of motion perceived by a humanlike robot. It is assumed that the robot receives visual images of the scene using a perspective system model. Equations, theorems, concepts, clues, etc., relating the objects, their positions, and their motion to their images on the focal plane are presented. >

http://ieeexplore.ieee.org/iel2/815/3148/00100651.pdf

Robot vision

Small molecule inhibitors for cancer metabolism: promising prospects to be explored

In this paper, we present a multi-scale approach based on superpixel classification for optic cup localization. Our approach provides 3 major contributions. First, a contrast enhancement scheme is proposed to reduce illumination influence and enhance feature discrimination. Second, features are extracted from multiple superpixels scales for richer description of the optic cup. Third, a unique cup is localized by integrating the multi-scales together using majority voting. Our approach was validated on a clinical online dataset, ORIGA-light, of 650 population-based images. Overall, our approach is able to achieve a 0.248 non-overlap ratio (m1) and a 0.085 absolute CDR error (δ). Experimental results also shows that our multi-scale approach has a complementary effect to increase performance stability, and is able to achieve a higher accuracy when compared with the previous state-of-the-art superpixel-based method.

Multi-scale superpixel classification for optic cup localization

Stimuli-responsive hydrogels have attracted much attention over the past decade for potential bioengineering applications such as wound dressing and drug delivery. In this work, a pH and temperature dual-responsive microgel-embedded hydrogel has been fabricated by incorporating poly(N-isopropylacrylamide-co-acrylic acid) (PNIPAAm-co-AAc) based microgel particles into polyacrylamide (PAAm)/chitosan (CS) semi-interpenetrating polymer network (semi-IPN), denoted as microgel@PAM/CS. The resultant hydrogel possesses excellent mechanical properties including stretchability, compressibility, and elasticity. In addition, the microgel@PAM/CS hydrogels can tightly adhere to the surfaces of a variety of tissues such as porcine skin, kidney, intestine, liver, and heart. Moreover, it shows controlled dual-drug release profile of both bovine serum albumin (BSA) (as a model protein) and sulfamethoxazole (SMZ), an antibiotic. Excellent antimicrobial properties are obtained for SMZ-loaded microgel@PAM/CS hydrogels. Compared with traditional drug administration methods such as by mouth, injection, and inhalation, the microgel@PAM/CS hydrogels possess advantages such as higher drug loading efficiency (by more than 80%) and controllable and sustained (over 48 h) release. The microgel@PAM/CS hydrogels can significantly enhance the wound healing process. This work provides a facile approach for the fabrication of multifunctional stimuli-responsive microparticle-embedded hydrogels with semi-IPN structures, and the as-prepared microgel@PAM/CS hydrogels have great potential for applications as smart wound dressing materials in biomedical engineering.

A pH and Temperature Dual-Responsive Microgel-Embedded, Adhesive, and Tough Hydrogel for Drug Delivery and Wound Healing.

Endoscopic video processing could facilitate pre-operative planning, intra-operative image guidance and generation of post-operative analysis of the surgical procedure. However, most of the current methods are still based on a single frame of image analysis, which makes the results of the previous frame images independent of each other and causes vibration. In this paper, we propose an temporal context framework for endoscopy artefact segmentation and detection. The framework extends the general segmentation and detection model to the form based on temporal input, and we add a Temporal Context Transformer(TCT) after the encoder of the model to improve the model’s ability to construct temporal context features. the experiments of the EndoCV 2022 challenge dataset that this framework can improve the robustness of the model.

Temporal Context Framework for Endoscopy Artefact Segmentation and Detection

Optical coherence tomography (OCT) is an emerging biomedical imaging technique. In recent years, it is applied in diagnosing eye diseases such as glaucoma and age-related macular degeneration. 3D scans can be obtained by stacking 2D OCT scans into a volume. However, due to the relatively low scan speed, there exists severe misalignment between the 2D scans. In this paper, we propose an exhaustive experimental study on macular OCT image alignment using intensity based image registrations techniques. More than 10 combinations of registration metric and transformations have been explored. Experimental results show that the combinations of sum of squared differences and normalized correlation coefficients and rigid transform are the most effective.

http://ieeexplore.ieee.org/iel7/7033073/7047454/07047596.pdf

Jiang Liu

Papers

Small molecule inhibitors for cancer metabolism: promising prospects to be explored

Multi-scale superpixel classification for optic cup localization

A pH and Temperature Dual-Responsive Microgel-Embedded, Adhesive, and Tough Hydrogel for Drug Delivery and Wound Healing.

Temporal Context Framework for Endoscopy Artefact Segmentation and Detection

A comparative study on macular optical coherence tomography image alignment