Unraveling a histopathological needle-in-haystack problem

• Verfasst von: Rusche, Daniel [VerfasserIn]
• Verfasst von: Englert, Nils [VerfasserIn]
• Verfasst von: Runz, Marlen [VerfasserIn]
• Verfasst von: Hetjens, Svetlana [VerfasserIn]
• Verfasst von: Langner, Cord [VerfasserIn]
• Verfasst von: Gaiser, Timo [VerfasserIn]
• Verfasst von: Weis, Cleo-Aron Thias [VerfasserIn]
• Titel: Unraveling a histopathological needle-in-haystack problem
• Titelzusatz: exploring the challenges of detecting tumor budding in colorectal carcinoma histology
• Verf.angabe: Daniel Rusche, Nils Englert, Marlen Runz, Svetlana Hetjens, Cord Langner, Timo Gaiser and Cleo-Aron Weis
• E-Jahr: 2024
• Jahr: 22 January 2024
• Umfang: 1-21$t21
• Fussnoten: Gesehen am 17.10.2024
• Titel Quelle: Enthalten in: Applied Sciences
• Ort Quelle: Basel : MDPI, 2011
• Jahr Quelle: 2024
• Band/Heft Quelle: 14(2024), 2, Artikel-ID 949
• ISSN Quelle: 2076-3417
• DOI: doi:10.3390/app14020949
• Datenträger: Online-Ressource
• Sprache: eng
• Sach-SW: budding
• Sach-SW: CANCER
• Sach-SW: classification
• Sach-SW: CRC
• Sach-SW: DIAGNOSIS
• Sach-SW: histopathology
• Sach-SW: PARAMETER
• Sach-SW: supervised segmentation
• K10plus-PPN: 1905980108

Abstract

Background: In this study focusing on colorectal carcinoma (CRC), we address the imperative task of predicting post-surgery treatment needs by identifying crucial tumor features within whole slide images of solid tumors, analogous to locating a needle in a histological haystack. We evaluate two approaches to address this challenge using a small CRC dataset. Methods: First, we explore a conventional tile-level training approach, testing various data augmentation methods to mitigate the memorization effect in a noisy label setting. Second, we examine a multi-instance learning (MIL) approach at the case level, adapting data augmentation techniques to prevent over-fitting in the limited data set context. Results: The tile-level approach proves ineffective due to the limited number of informative image tiles per case. Conversely, the MIL approach demonstrates success for the small dataset when coupled with post-feature vector creation data augmentation techniques. In this setting, the MIL model accurately predicts nodal status corresponding to expert-based budding scores for these cases. Conclusions: This study incorporates data augmentation techniques into a MIL approach, highlighting the effectiveness of the MIL method in detecting predictive factors such as tumor budding, despite the constraints of a limited dataset size.