Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach

Jean-Baptiste Lamy, Boomadevi Sekar, Gilles Guézennec, Jacques Bouaud, Brigitte Seroussi

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Case-Based Reasoning (CBR) is a form of analogical reasoning in which the solution for a (new) query case is determined using a database of previous known cases with their solutions. Cases similar to the query are retrieved from the database, and then their solutions are adapted to the query. In medicine, a case usually corresponds to a patient and the problem consists of classifying the patient in a class of diagnostic or therapy. Compared to “black box” algorithms such as deep learning, the responses of CBR systems can be justified easily using the similar cases as examples. However, this possibility is often under-exploited and the explanations provided by most CBR systems are limited to the display of the similar cases.
In this paper, we propose a CBR method that can be both executed automatically as an algorithm and presented visually in a user interface for providing visual explanations or for visual reasoning. After retrieving similar cases, a visual interface displays quantitative and qualitative similarities between the query and the similar cases, so as one can easily classify the query through visual reasoning, in a fully explainable manner. It combines a quantitative approach (visualized by a scatter plot based on Multidimensional Scaling in polar coordinates, preserving distances involving the query) and a qualitative approach (set visualization using rainbow boxes). We applied this method to breast cancer management. We showed on three public datasets that our qualitative method has a classification accuracy comparable to k-Nearest Neighbors algorithms, but is better explainable. We also tested the proposed interface during a small user study. Finally, we apply the proposed approach to a real dataset in breast cancer. Medical experts found the visual approach interesting as it explains why cases are similar through the visualization of shared patient characteristics.
LanguageEnglish
Pages42-53
Number of pages12
JournalArtificial Intelligence in Medicine
Volume94
DOIs
Publication statusPublished - 1 Mar 2019

Fingerprint

Case based reasoning
Artificial Intelligence
Artificial intelligence
Breast Neoplasms
Databases
Visualization
Medicine
Learning
User interfaces
Display devices
Datasets
Therapeutics

Keywords

  • Case-based reasoning
  • Visual explanation
  • Explainable Artificial Intelligence
  • Data-driven decision making
  • Multidimensional scaling
  • Breast Cancer
  • Multidimensional Scaling
  • Breast cancer

Cite this

Lamy, Jean-Baptiste ; Sekar, Boomadevi ; Guézennec, Gilles ; Bouaud, Jacques ; Seroussi, Brigitte. / Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach. In: Artificial Intelligence in Medicine. 2019 ; Vol. 94. pp. 42-53.
@article{cf717ec9b94c4c6fbc146ab5e7814640,
title = "Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach",
abstract = "Case-Based Reasoning (CBR) is a form of analogical reasoning in which the solution for a (new) query case is determined using a database of previous known cases with their solutions. Cases similar to the query are retrieved from the database, and then their solutions are adapted to the query. In medicine, a case usually corresponds to a patient and the problem consists of classifying the patient in a class of diagnostic or therapy. Compared to “black box” algorithms such as deep learning, the responses of CBR systems can be justified easily using the similar cases as examples. However, this possibility is often under-exploited and the explanations provided by most CBR systems are limited to the display of the similar cases.In this paper, we propose a CBR method that can be both executed automatically as an algorithm and presented visually in a user interface for providing visual explanations or for visual reasoning. After retrieving similar cases, a visual interface displays quantitative and qualitative similarities between the query and the similar cases, so as one can easily classify the query through visual reasoning, in a fully explainable manner. It combines a quantitative approach (visualized by a scatter plot based on Multidimensional Scaling in polar coordinates, preserving distances involving the query) and a qualitative approach (set visualization using rainbow boxes). We applied this method to breast cancer management. We showed on three public datasets that our qualitative method has a classification accuracy comparable to k-Nearest Neighbors algorithms, but is better explainable. We also tested the proposed interface during a small user study. Finally, we apply the proposed approach to a real dataset in breast cancer. Medical experts found the visual approach interesting as it explains why cases are similar through the visualization of shared patient characteristics.",
keywords = "Case-based reasoning, Visual explanation, Explainable Artificial Intelligence, Data-driven decision making, Multidimensional scaling, Breast Cancer, Multidimensional Scaling, Breast cancer",
author = "Jean-Baptiste Lamy and Boomadevi Sekar and Gilles Gu{\'e}zennec and Jacques Bouaud and Brigitte Seroussi",
year = "2019",
month = "3",
day = "1",
doi = "10.1016/j.artmed.2019.01.001",
language = "English",
volume = "94",
pages = "42--53",
journal = "Artificial Intelligence in Medicine",
issn = "0933-3657",
publisher = "Elsevier",

}

Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach. / Lamy, Jean-Baptiste; Sekar, Boomadevi; Guézennec, Gilles; Bouaud, Jacques; Seroussi, Brigitte.

In: Artificial Intelligence in Medicine, Vol. 94, 01.03.2019, p. 42-53.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach

AU - Lamy, Jean-Baptiste

AU - Sekar, Boomadevi

AU - Guézennec, Gilles

AU - Bouaud, Jacques

AU - Seroussi, Brigitte

PY - 2019/3/1

Y1 - 2019/3/1

N2 - Case-Based Reasoning (CBR) is a form of analogical reasoning in which the solution for a (new) query case is determined using a database of previous known cases with their solutions. Cases similar to the query are retrieved from the database, and then their solutions are adapted to the query. In medicine, a case usually corresponds to a patient and the problem consists of classifying the patient in a class of diagnostic or therapy. Compared to “black box” algorithms such as deep learning, the responses of CBR systems can be justified easily using the similar cases as examples. However, this possibility is often under-exploited and the explanations provided by most CBR systems are limited to the display of the similar cases.In this paper, we propose a CBR method that can be both executed automatically as an algorithm and presented visually in a user interface for providing visual explanations or for visual reasoning. After retrieving similar cases, a visual interface displays quantitative and qualitative similarities between the query and the similar cases, so as one can easily classify the query through visual reasoning, in a fully explainable manner. It combines a quantitative approach (visualized by a scatter plot based on Multidimensional Scaling in polar coordinates, preserving distances involving the query) and a qualitative approach (set visualization using rainbow boxes). We applied this method to breast cancer management. We showed on three public datasets that our qualitative method has a classification accuracy comparable to k-Nearest Neighbors algorithms, but is better explainable. We also tested the proposed interface during a small user study. Finally, we apply the proposed approach to a real dataset in breast cancer. Medical experts found the visual approach interesting as it explains why cases are similar through the visualization of shared patient characteristics.

AB - Case-Based Reasoning (CBR) is a form of analogical reasoning in which the solution for a (new) query case is determined using a database of previous known cases with their solutions. Cases similar to the query are retrieved from the database, and then their solutions are adapted to the query. In medicine, a case usually corresponds to a patient and the problem consists of classifying the patient in a class of diagnostic or therapy. Compared to “black box” algorithms such as deep learning, the responses of CBR systems can be justified easily using the similar cases as examples. However, this possibility is often under-exploited and the explanations provided by most CBR systems are limited to the display of the similar cases.In this paper, we propose a CBR method that can be both executed automatically as an algorithm and presented visually in a user interface for providing visual explanations or for visual reasoning. After retrieving similar cases, a visual interface displays quantitative and qualitative similarities between the query and the similar cases, so as one can easily classify the query through visual reasoning, in a fully explainable manner. It combines a quantitative approach (visualized by a scatter plot based on Multidimensional Scaling in polar coordinates, preserving distances involving the query) and a qualitative approach (set visualization using rainbow boxes). We applied this method to breast cancer management. We showed on three public datasets that our qualitative method has a classification accuracy comparable to k-Nearest Neighbors algorithms, but is better explainable. We also tested the proposed interface during a small user study. Finally, we apply the proposed approach to a real dataset in breast cancer. Medical experts found the visual approach interesting as it explains why cases are similar through the visualization of shared patient characteristics.

KW - Case-based reasoning

KW - Visual explanation

KW - Explainable Artificial Intelligence

KW - Data-driven decision making

KW - Multidimensional scaling

KW - Breast Cancer

KW - Multidimensional Scaling

KW - Breast cancer

U2 - 10.1016/j.artmed.2019.01.001

DO - 10.1016/j.artmed.2019.01.001

M3 - Article

VL - 94

SP - 42

EP - 53

JO - Artificial Intelligence in Medicine

T2 - Artificial Intelligence in Medicine

JF - Artificial Intelligence in Medicine

SN - 0933-3657

ER -