Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble

Aadish Shaikh; Bicky Yadav; Dhruv Thejas Kj; G. Ignisha Rajathi

doi:10.1109/NQComp68334.2026.11497687

Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble

Aadish Shaikh^*
, Bicky Yadav
, Dhruv Thejas Kj
, G. Ignisha Rajathi

^*Corresponding author for this work

School of Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Ovarian cancer remains the most lethal gynecologic malignancy and creates a major burden on the health of women globally. For effective treatment, accurate subtyping is imperative; however, most deep learning models overestimate their performance by incorporating the common and easily identifiable HGSC into a 5-class problem. In this respect, we are interested in a more challenging 4-class problem involving rare subtypes: clear cell, endometrioid, low-grade serous, and mucinous carcinomas. We propose a weighted ensemble that combines Swin Transformer and ConvNeXt-Tiny. It achieves state-of-the-art accuracy of 96.46% and a macro F1-score of 96.12%. This work demonstrates that hybrid CNN-Transformer ensembles provide a strong solution to the problem of detailed ovarian cancer subtype classification.

Original language	English
Title of host publication	Proceedings of International Conference on Next-Gen Quantum and Advanced Computing
Subtitle of host publication	Algorithms, Security, and Beyond, NQComp 2026
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	118-123
Number of pages	6
ISBN (Electronic)	9798331559359
DOIs	https://doi.org/10.1109/NQComp68334.2026.11497687
Publication status	Published - 2026
Event	2026 International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026 - Bangalore, India Duration: 05-03-2026 → 06-03-2026

Publication series

Name	Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026

Conference

Conference	2026 International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026
Country/Territory	India
City	Bangalore
Period	05-03-26 → 06-03-26

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

All Science Journal Classification (ASJC) codes

Artificial Intelligence
Information Systems
Computer Networks and Communications
Hardware and Architecture

Access to Document

10.1109/NQComp68334.2026.11497687

Cite this

Shaikh, A., Yadav, B., Kj, D. T., & Ignisha Rajathi, G. (2026). Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble. In Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026 (pp. 118-123). (Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/NQComp68334.2026.11497687

Shaikh, Aadish ; Yadav, Bicky ; Kj, Dhruv Thejas et al. / Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble. Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026. Institute of Electrical and Electronics Engineers Inc., 2026. pp. 118-123 (Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026).

@inproceedings{62c8014256e24500b5b17f3284ccea33,

title = "Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble",

abstract = "Ovarian cancer remains the most lethal gynecologic malignancy and creates a major burden on the health of women globally. For effective treatment, accurate subtyping is imperative; however, most deep learning models overestimate their performance by incorporating the common and easily identifiable HGSC into a 5-class problem. In this respect, we are interested in a more challenging 4-class problem involving rare subtypes: clear cell, endometrioid, low-grade serous, and mucinous carcinomas. We propose a weighted ensemble that combines Swin Transformer and ConvNeXt-Tiny. It achieves state-of-the-art accuracy of 96.46\% and a macro F1-score of 96.12\%. This work demonstrates that hybrid CNN-Transformer ensembles provide a strong solution to the problem of detailed ovarian cancer subtype classification.",

author = "Aadish Shaikh and Bicky Yadav and Kj, \{Dhruv Thejas\} and \{Ignisha Rajathi\}, G.",

note = "Publisher Copyright: {\textcopyright} 2026 IEEE.; 2026 International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026 ; Conference date: 05-03-2026 Through 06-03-2026",

year = "2026",

doi = "10.1109/NQComp68334.2026.11497687",

language = "English",

series = "Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "118--123",

booktitle = "Proceedings of International Conference on Next-Gen Quantum and Advanced Computing",

address = "United States",

}

Shaikh, A, Yadav, B, Kj, DT & Ignisha Rajathi, G 2026, Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble. in Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026. Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026, Institute of Electrical and Electronics Engineers Inc., pp. 118-123, 2026 International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026, Bangalore, India, 05-03-26. https://doi.org/10.1109/NQComp68334.2026.11497687

Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble. / Shaikh, Aadish; Yadav, Bicky; Kj, Dhruv Thejas et al.
Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026. Institute of Electrical and Electronics Engineers Inc., 2026. p. 118-123 (Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble

AU - Shaikh, Aadish

AU - Yadav, Bicky

AU - Kj, Dhruv Thejas

AU - Ignisha Rajathi, G.

PY - 2026

Y1 - 2026

N2 - Ovarian cancer remains the most lethal gynecologic malignancy and creates a major burden on the health of women globally. For effective treatment, accurate subtyping is imperative; however, most deep learning models overestimate their performance by incorporating the common and easily identifiable HGSC into a 5-class problem. In this respect, we are interested in a more challenging 4-class problem involving rare subtypes: clear cell, endometrioid, low-grade serous, and mucinous carcinomas. We propose a weighted ensemble that combines Swin Transformer and ConvNeXt-Tiny. It achieves state-of-the-art accuracy of 96.46% and a macro F1-score of 96.12%. This work demonstrates that hybrid CNN-Transformer ensembles provide a strong solution to the problem of detailed ovarian cancer subtype classification.

AB - Ovarian cancer remains the most lethal gynecologic malignancy and creates a major burden on the health of women globally. For effective treatment, accurate subtyping is imperative; however, most deep learning models overestimate their performance by incorporating the common and easily identifiable HGSC into a 5-class problem. In this respect, we are interested in a more challenging 4-class problem involving rare subtypes: clear cell, endometrioid, low-grade serous, and mucinous carcinomas. We propose a weighted ensemble that combines Swin Transformer and ConvNeXt-Tiny. It achieves state-of-the-art accuracy of 96.46% and a macro F1-score of 96.12%. This work demonstrates that hybrid CNN-Transformer ensembles provide a strong solution to the problem of detailed ovarian cancer subtype classification.

UR - https://www.scopus.com/pages/publications/105041546940

UR - https://www.scopus.com/pages/publications/105041546940#tab=citedBy

U2 - 10.1109/NQComp68334.2026.11497687

DO - 10.1109/NQComp68334.2026.11497687

M3 - Conference contribution

AN - SCOPUS:105041546940

T3 - Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026

SP - 118

EP - 123

BT - Proceedings of International Conference on Next-Gen Quantum and Advanced Computing

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2026 International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026

Y2 - 5 March 2026 through 6 March 2026

ER -

Shaikh A, Yadav B, Kj DT, Ignisha Rajathi G. Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble. In Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026. Institute of Electrical and Electronics Engineers Inc. 2026. p. 118-123. (Proceedings of International Conference on Next-Gen Quantum and Advanced Computing: Algorithms, Security, and Beyond, NQComp 2026). doi: 10.1109/NQComp68334.2026.11497687

Histopathological Classification of Rare Ovarian Cancer using a Swin-ConvNeXt Ensemble

Abstract

Publication series

Conference

UN SDGs

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this