Novelty detection for breast cancer image classification

Paweł Cichosz , Dariusz Jagodziński , Mateusz Matysiewicz , Łukasz Neumann , Robert Marek Nowak , Rafał Okuniewski , Witold Oleszkiewicz

Abstract

Using classification learning algorithms for medical applications may require not only refined model creation techniques and careful unbiased model evaluation, but also detecting the risk of misclassification at the time of model application. This is addressed by novelty detection, which identifies instances for which the training set is not sufficiently representative and for which it may be safer to restrain from classification and request a human expert diagnosis. The paper investigates two techniques for isolated instance identification, based on clustering and one-class support vector machines, which represent two different approaches to multidimensional outlier detection. The prediction quality for isolated instances in breast cancer image data is evaluated using the random forest algorithm and found to be substantially inferior to the prediction quality for non-isolated instances. Each of the two techniques is then used to create a novelty detection model which can be combined with a classification model and used at the time of prediction to detect instances for which the latter cannot be reliably applied. Novelty detection is demonstrated to improve random forest prediction quality and argued to deserve further investigation in medical applications
Author Paweł Cichosz ISE
Paweł Cichosz,,
- The Institute of Electronic Systems
, Dariusz Jagodziński ISE
Dariusz Jagodziński,,
- The Institute of Electronic Systems
, Mateusz Matysiewicz
Mateusz Matysiewicz,,
-
, Łukasz Neumann
Łukasz Neumann,,
-
, Robert Marek Nowak ISE
Robert Marek Nowak,,
- The Institute of Electronic Systems
, Rafał Okuniewski
Rafał Okuniewski,,
-
, Witold Oleszkiewicz
Witold Oleszkiewicz,,
-
Pages1003135-1-1003135-12
Publication size in sheets0.5
Book Romaniuk Ryszard (eds.): Proc. SPIE. 10031, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2016, vol. 10031, 2016, SPIE , ISBN 9781510604858, [781510604865 (electronic) ], 1170 p., DOI:10.1117/12.2257157
DOIDOI:10.1117/12.2249183
URL http://dx.doi.org/10.1117/12.2249183
Languageen angielski
File
1003135 (1)_Cichosz_jago.pdf 336.84 KB
Score (nominal)15
ScoreMinisterial score = 15.0, 27-03-2017, BookChapterMatConf
Ministerial score (2013-2016) = 15.0, 27-03-2017, BookChapterMatConf
Citation count*8 (2018-06-20)
Cite
Share Share

Get link to the record
msginfo.png


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back