Classification of breast cancer histology images using Convolutional Neural Networks.

Araújo T, Aresta G, Castro E, Rouco J, Aguiar P, Eloy C, Polónia A, Campilho A. · 2017

Read Classification of breast cancer histology images using Convolutional Neural Networks. by Araújo T, Aresta G, Castro E, Rouco J, Aguiar P, Eloy C, Polónia A, Campilho A. (2017) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc5453426.txt.

Description

Araújo and co-authors train a neural network to sort breast biopsy images into diagnostic classes. Their research article appeared in PLoS ONE. Reading stained tissue slides is slow expert work. Specialists agree on a diagnosis only about 75 percent of the time. Older computer aids need hand-built features. Deep learning can learn its own features. The team used a convolutional neural network. It reads detail at several scales at once.

The network sorts images into four classes. They are normal tissue, benign lesion, in situ carcinoma, and invasive carcinoma. A simpler split sorts cancer from non-cancer. Design cues came from how pathologists look. Nuclei detail matters at small scale. Tissue layout matters at large scale. The system reached 77.8 percent accuracy on four classes. It reached 83.3 percent on the two-class split. Sensitivity for cancer cases hit 95.6 percent. Network features also fed a support vector machine.

Catalog record

CREATORAraújo T, Aresta G, Castro E, Rouco J, Aguiar P, Eloy C, Polónia A, Campilho A.
YEAR2017
COLLECTIONPAPERS
FORMATTXT
FILE SIZE37 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC5453426

Browse the full library · About the Tower of Babel