A fast and effective detection framework for whole-slide histopathology image analysis

Pathologists generally pan, focus, zoom and scan tissue biopsies either under microscopes or on digital images for diagnosis. With the rapid development of whole-slide digital scanners for histopathology, computer-assisted digital pathology image analysis has attracted increasing clinical attention. Thus, the working style of pathologists is also beginning to change. Computer-assisted image analysis systems have been developed to help pathologists perform basic examinations. This paper presents a novel lightweight detection framework for automatic tumor detection in whole-slide histopathology images. We develop the Double Magnification Combination (DMC) classifier, which is a modified DenseNet-40 to make patch-level predictions with only 0.3 million parameters. To improve the detection performance of multiple instances, we propose an improved adaptive sampling method with superpixel segmentation and introduce a new heuristic factor, local sampling density, as the convergence condition of iterations. In postprocessing, we use a CNN model with 4 convolutional layers to regulate the patch-level predictions based on the predictions of adjacent sampling points and use linear interpolation to generate a tumor probability heatmap. The entire framework was trained and validated using the dataset from the Camelyon16 Grand Challenge and Hubei Cancer Hospital. In our experiments, the average AUC was 0.95 in the test set for pixel-level detection.

Download Full-text

A Digital Pathology application for whole-slide histopathology image analysis based on genetic algorithm and Convolutional Networks

2016 IEEE Latin American Conference on Computational Intelligence (LA-CCI) ◽

10.1109/la-cci.2016.7885738 ◽

2016 ◽

Cited By ~ 1

Author(s):

Mateo Puerto ◽

Tania Vargas ◽

Angel Cruz-Roa

Keyword(s):

Genetic Algorithm ◽

Image Analysis ◽

Digital Pathology ◽

Convolutional Networks ◽

Histopathology Image Analysis

Download Full-text

Unified framework for development, deployment and testing of image analysis algorithms

10.54294/pq6gf6 ◽

2009 ◽

Author(s):

Alark Joshi ◽

Dustin Scheinost ◽

Hirohito Okuda ◽

Isabella Murphy ◽

Lawrence Staib ◽

...

Keyword(s):

Image Analysis ◽

User Interface ◽

User Interfaces ◽

Rapid Development ◽

Computer Assisted ◽

Yale University ◽

Unified Framework ◽

Considerable Effort ◽

Separate Source ◽

Cross Platform

Developing both graphical and command-line user interfaces for image analysis algorithms requires considerable effort. Generally developers provide limited to very rudimentary user interface controls to their users. These image analysis algorithms can only meet their potential if they can be used easily and frequently by their intended users. Deployment of a large suite of such algorithms on multiple platforms requires that the software be stable and appropriately tested.We present a novel framework that allows for rapid development of image analysis algorithms along with graphical user interface controls. Additionally, our framework allows for simplified nightly testing of the algorithms to ensure stability and cross platform interoperability. It allows for development of complex algorithms by creating a custom pipeline where the output of an algorithm can serve as an input for another algorithm. All of the functionality is encapsulation into the object requiring no separate source code for user interfaces, testing or deployment. This makes our framework ideal for developing novel, stable and easy-to-use algorithms for computer assisted interventions (CAI). The framework has been deployed at the Magnetic Resonance Research Center at Yale University and has been released for public use.

Download Full-text

High-throughput adaptive sampling for whole-slide histopathology image analysis (HASHI) via convolutional neural networks: Application to invasive breast cancer detection

PLoS ONE ◽

10.1371/journal.pone.0196828 ◽

2018 ◽

Vol 13 (5) ◽

pp. e0196828 ◽

Cited By ~ 37

Author(s):

Angel Cruz-Roa ◽

Hannah Gilmore ◽

Ajay Basavanhally ◽

Michael Feldman ◽

Shridar Ganesan ◽

...

Keyword(s):

Breast Cancer ◽

Neural Networks ◽

Image Analysis ◽

Convolutional Neural Networks ◽

High Throughput ◽

Cancer Detection ◽

Invasive Breast Cancer ◽

Adaptive Sampling ◽

Breast Cancer Detection ◽

Histopathology Image Analysis

Download Full-text

Representación de imágenes de histopatología utilizada en tareas de análisis automático: estado del arte

Revista Med ◽

10.18359/rmed.1184 ◽

2014 ◽

Vol 22 (2) ◽

pp. 79 ◽

Cited By ~ 15

Author(s):

John Arevalo ◽

Angel Cruz-Roa ◽

Fabio A. González O

Keyword(s):

Image Analysis ◽

Image Representation ◽

Digital Pathology ◽

Virtual Microscopy ◽

Automatic Analysis ◽

Computer Assisted ◽

Automatic Image Analysis ◽

Histopathology Images ◽

Image Enhancing ◽

Assisted Diagnosis

<p>This paper presents a review of the state-of-the-art in histopathology image representation used in automatic image analysis tasks. Automatic analysis of histopathology images is important for building computer-assisted diagnosis tools, automatic image enhancing systems and virtual microscopy systems, among other applications. Histopathology images have a rich mix of visual patterns with particularities that make them difficult to analyze. The paper discusses these particularities, the acquisition process and the challenges found when doing automatic analysis. Second an overview of recent works and methods addressed to deal with visual content representation in different automatic image analysis tasks is presented. Third an overview of applications of image representation methods in several medical domains and tasks is presented. Finally, the paper concludes with current trends of automatic analysis of histopathology images like digital pathology.</p>

Download Full-text

Biodegradable polyester neuroprostheses promote the reconnection of separated peripheral nerve stubs

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100130109 ◽

1992 ◽

Vol 50 (2) ◽

pp. 1094-1095

Author(s):

Beverly L. Giammara ◽

Jennifer S. Stevenson ◽

Peggy E. Yates ◽

Robert H. Gunderson ◽

Jacob S. Hanker

Keyword(s):

Image Analysis ◽

Basement Membrane ◽

Periodic Acid ◽

Light And Electron Microscopy ◽

Computer Assisted ◽

Type Iii Collagen ◽

Image Analysis System ◽

Biodegradable Polyester ◽

Adult Rats ◽

Type Iv

An 11mm length of sciatic nerve was removed from 10 anesthetized adult rats and replaced by a biodegradable polyester Vicryl™ mesh sleeve which was then injected with the basement membrane gel, Matrigel™. It was noted that leg sensation and movement were much improved after 30 to 45 days and upon sacrifice nerve reconnection was noted in all animals. Epoxy sections of the repaired nerves were compared with those of the excised segments by the use of a variation of the PAS reaction, the PATS reaction, developed in our laboratories for light and electron microscopy. This microwave-accelerated technique employs periodic acid, thiocarbohydrazide and silver methenamine. It stains basement membrane or Type IV collagen brown and type III collagen (reticulin), axons, Schwann cells, endoneurium and perineurium black. Epoxy sections of repaired and excised nerves were also compared by toluidine blue (tb) staining. Comparison of the sections of control and repaired nerves was done by computer-assisted microscopic image analysis using an Olympus CUE-2 Image Analysis System.

Download Full-text