Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis

Md. Masum Billah; Pragati Manandhar; Sarosh Krishan; Alejandro Cedillo; Hergys Rexha; Sébastien Lafond; Kurt K Benke; Sepinoud Azimi; Janan Arslan

Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis

Md. Masum Billah, Pragati Manandhar, Sarosh Krishan, Alejandro Cedillo, Hergys Rexha, Sébastien Lafond, Kurt K Benke, Sepinoud Azimi, Janan Arslan

Research output: Chapter in Book/Conference proceeding › Conference contribution › Scientific › peer-review

Abstract

Despite their predictive capabilities and rapid advancement, the black-box nature of Artificial Intelligence (AI) models, particularly in healthcare, has sparked debate regarding their trustworthiness and accountability. In response, the field of Explainable AI (XAI) has emerged, aiming to create transparent AI technologies. We present a novel approach to enhance AI interpretability by leveraging texture analysis, with a focus on cancer datasets. By focusing on specific texture features and their correlations with a prediction outcome extracted from medical images, our proposed methodology aims to elucidate the underlying mechanics of AI, improve AI trustworthiness, and facilitate human understanding. The code is available at https://github.com/xrai-lib/xai-texture.

Original language	English
Title of host publication	Medical Imaging with Deep Learning (MIDL 2024)
Place of Publication	Paris, France
Publication status	Published - 1 Jul 2024
MoE publication type	A4 Article in a conference publication
Event	Medical Imaging with Deep Learning - Paris Duration: 3 Jul 2024 → …

Publication series

Name	Proceedings of Machine Learning Research
ISSN (Electronic)	2640-3498

Conference

Conference	Medical Imaging with Deep Learning
Abbreviated title	MIDL
City	Paris
Period	03/07/24 → …

Keywords

Artificial Intelligence
Cancer Diagnosis
Explainable AI
Texture Analysis
Medical Imaging

Access to Document

https://hal.science/hal-04562334

EDISS: EMJMD Programme on the Engineering of Data-intensive Intelligent Software Systems
Lafond, S. (Principal Investigator), Azimi Rashti, S. (Principal Investigator), Lilius, J. (Co-Principal Investigator), Strömborg, M. (Coordinator) & Iancu, B. (Co-Investigator)
Education, Audiovisual and Culture Executive Agency - European Commission
01/09/20 → 31/08/26
Project: EU

Cite this

@inproceedings{ecbf45c2cd9b46b48851f4538ebb38c2,

title = "Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis",

abstract = "Despite their predictive capabilities and rapid advancement, the black-box nature of Artificial Intelligence (AI) models, particularly in healthcare, has sparked debate regarding their trustworthiness and accountability. In response, the field of Explainable AI (XAI) has emerged, aiming to create transparent AI technologies. We present a novel approach to enhance AI interpretability by leveraging texture analysis, with a focus on cancer datasets. By focusing on specific texture features and their correlations with a prediction outcome extracted from medical images, our proposed methodology aims to elucidate the underlying mechanics of AI, improve AI trustworthiness, and facilitate human understanding. The code is available at https://github.com/xrai-lib/xai-texture.",

keywords = "Artificial Intelligence, Cancer Diagnosis, Explainable AI, Texture Analysis, Medical Imaging",

author = "Billah, {Md. Masum} and Pragati Manandhar and Sarosh Krishan and Alejandro Cedillo and Hergys Rexha and S{\'e}bastien Lafond and Benke, {Kurt K} and Sepinoud Azimi and Janan Arslan",

year = "2024",

month = jul,

day = "1",

language = "English",

series = "Proceedings of Machine Learning Research",

booktitle = "Medical Imaging with Deep Learning (MIDL 2024)",

note = "Medical Imaging with Deep Learning, MIDL ; Conference date: 03-07-2024",

}

Billah, MM, Manandhar, P, Krishan, S, Cedillo, A, Rexha, H , Lafond, S, Benke, KK, Azimi, S & Arslan, J 2024, Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis. in Medical Imaging with Deep Learning (MIDL 2024). Proceedings of Machine Learning Research, Paris, France, Medical Imaging with Deep Learning, Paris, 03/07/24. <https://hal.science/hal-04562334>

Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis. / Billah, Md. Masum; Manandhar, Pragati; Krishan, Sarosh et al.
Medical Imaging with Deep Learning (MIDL 2024). Paris, France, 2024. (Proceedings of Machine Learning Research).

Research output: Chapter in Book/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis

AU - Billah, Md. Masum

AU - Manandhar, Pragati

AU - Krishan, Sarosh

AU - Cedillo, Alejandro

AU - Rexha, Hergys

AU - Lafond, Sébastien

AU - Benke, Kurt K

AU - Azimi, Sepinoud

AU - Arslan, Janan

PY - 2024/7/1

Y1 - 2024/7/1

N2 - Despite their predictive capabilities and rapid advancement, the black-box nature of Artificial Intelligence (AI) models, particularly in healthcare, has sparked debate regarding their trustworthiness and accountability. In response, the field of Explainable AI (XAI) has emerged, aiming to create transparent AI technologies. We present a novel approach to enhance AI interpretability by leveraging texture analysis, with a focus on cancer datasets. By focusing on specific texture features and their correlations with a prediction outcome extracted from medical images, our proposed methodology aims to elucidate the underlying mechanics of AI, improve AI trustworthiness, and facilitate human understanding. The code is available at https://github.com/xrai-lib/xai-texture.

AB - Despite their predictive capabilities and rapid advancement, the black-box nature of Artificial Intelligence (AI) models, particularly in healthcare, has sparked debate regarding their trustworthiness and accountability. In response, the field of Explainable AI (XAI) has emerged, aiming to create transparent AI technologies. We present a novel approach to enhance AI interpretability by leveraging texture analysis, with a focus on cancer datasets. By focusing on specific texture features and their correlations with a prediction outcome extracted from medical images, our proposed methodology aims to elucidate the underlying mechanics of AI, improve AI trustworthiness, and facilitate human understanding. The code is available at https://github.com/xrai-lib/xai-texture.

KW - Artificial Intelligence

KW - Cancer Diagnosis

KW - Explainable AI

KW - Texture Analysis

KW - Medical Imaging

M3 - Conference contribution

T3 - Proceedings of Machine Learning Research

BT - Medical Imaging with Deep Learning (MIDL 2024)

CY - Paris, France

T2 - Medical Imaging with Deep Learning

Y2 - 3 July 2024

ER -

Explainability in Deep Learning Segmentation Models for Breast Cancer by Analogy with Texture Analysis

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Projects

EDISS: EMJMD Programme on the Engineering of Data-intensive Intelligent Software Systems

Cite this