Radiomics-Informed Deep Curriculum Learning for Breast Cancer Diagnosis

Nebbia, Giacomo; Dadsetan, Saba; Arefan, Dooman; Zuley, Margarita L.; Sumkin, Jules H.; Huang, Heng; Wu, Shandong

doi:10.1007/978-3-030-87240-3_61

Cited by 5 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Adding a per-sample difficulty weight to the CE loss can be advantageous to penalize difficult samples more than easy ones 15 . Using this motivation, we propose to use the class label information from KNN features to extract a difficulty score for every sample in the training dataset.…”

Section: Dwce Lossmentioning

confidence: 99%

Enhancing downstream classification of breast abnormalities in contrast enhanced spectral mammography using a neighborhood representation loss

Bhandary Panambur,

Madhu,

Bayer

et al. 2024

Medical Imaging 2024: Computer-Aided Diagnosis

View full text Add to dashboard Cite

Contrast-enhanced spectral mammography (CESM) has proven to be a beneficial tool for additional management and diagnosis of breast cancer in patients. Deep learning-based automated tools for analyzing CESM images and detecting abnormalities can be crucial in reducing the workload of radiologists. Even though self-supervised learning (SSL) can significantly influence the quality of learned representations resulting in increased classification performances, it requires large amounts of data often unavailable in the CESM domain. In this work, we propose a novel loss function named Difficulty-Weighted Neighborhood Representation (DWNR) utilizing the SSL-based representations designed explicitly for the downstream classification of the region of interest (ROI) in CESM images in a small dataset setting. We propose utilizing the neighborhood representations extracted from an SSL backbone to guide the loss function in learning robust representations of the dataset's normal, benign, and malignant ROIs. We formulate the DWNR loss using a combination of a neighborhood representation-guided triplet (NR-triplet) loss and a difficulty-weighted cross-entropy (DWCE) loss. The NR-triplet loss leverages the nearest and farthest neighbor feature representations, extracted from a self-supervised backbone, as positive and negative samples. Meanwhile, the DWCE loss exploits the class information from the nearest neighbors to perform per-sample difficulty weighting. The experimental results on the publicly available CDD-CESM dataset demonstrate that the proposed DWNR loss outperforms traditional loss functions such as Cross-Entropy (CE) and Focal loss by a significant gain of more than 2% absolute mean F1-scores for both SSL-based and ImageNet-SSL-based transfer learning. In addition, we demonstrate the efficacy of employing a DWNR-based ROI classifier for comprehensive CESM whole image analysis in BI-RADS scoring and cancer diagnosis. The results reveal a gain in mean F1-scores of more than 1% in both tasks when using the DWNR loss models for CNN initialization, indicating the significance of the proposed method to whole image analysis. Based on our analysis, incorporating DWNR loss during transfer learning could positively impact automated breast cancer diagnosis using the CESM imaging modality. This novel loss function opens new possibilities for leveraging deep learning in CESM, particularly in scenarios with limited labeled data availability.

show abstract

Section: Dwce Lossmentioning

confidence: 99%

Enhancing downstream classification of breast abnormalities in contrast enhanced spectral mammography using a neighborhood representation loss

Bhandary Panambur,

Madhu,

Bayer

et al. 2024

Medical Imaging 2024: Computer-Aided Diagnosis

View full text Add to dashboard Cite

show abstract

“…Mammogram-based breast cancer studies can also be categorized based on their utilization of input data into several groups: those focusing on unimodal data, which may encompass whole mammograms [19]- [22], [25], [28], [29], [32], [33], [33], [35], [37]- [39], [41], [44], [48], [52]- [56] or lesion crops [18], [27], [34], [36], [40], [51], [57]; others concentrating on multi-view mammograms [17], [30], [50], [58], [58]- [63]; some addressing both whole mammograms and lesion crops [26], [31], [42]; and additional investigations involving the integration of multiple modalities, such as the combination of mammograms with ultrasounds or electronic health records (EHRs) [36], [49], [54], [64]- [72]. In this work, we focus on two modalities: mammogram-based lesion crops and EHR data containing radiographic features and clinical patient metadata for breast cancer prediction.…”

Section: A Mammogram-based Breast Cancer Predictionmentioning

confidence: 99%

Frozen Large-scale Pretrained Vision-Language Models are an Effective Foundational Backbone for Enhancing Multimodal Breast Cancer Risk Assessment

Vo,

Wang,

Wong

et al. 2024

Preprint

View full text Add to dashboard Cite

Breast cancer is a pervasive global health concern among women. Leveraging multimodal data from enterprise patient databases-including Picture Archiving and Communication Systems (PACS) and Electronic Health Records (EHRs)-holds promise for improving risk assessment. This study introduces a multimodal deep-learning model leveraging mammogram datasets to evaluate breast cancer risk. Our approach integrates frozen large-scale pretrained vision-language models, showcasing superior performance and stability compared to traditional imagetabular models across two public breast cancer datasets. The model consistently outperforms conventional full finetuning methods by using frozen pretrained vision-language models alongside a lightweight trainable classifier. The observed improvements are significant. In the CBIS-DDSM dataset, the Area Under the Curve (AUC) increases from 0.867 to 0.902 during validation and from 0.803 to 0.830 for the official test set. Within the EMBED dataset, AUC improves from 0.780 to 0.805 during validation. In scenarios with limited data, using Breast Imaging-Reporting and Data System category three (BI-RADS 3) cases, AUC improves from 0.91 to 0.96 on the official CBIS-DDSM test set and from 0.79 to 0.83 on a challenging validation set. This study underscores the benefits of vision-language models in jointly training diverse image-clinical datasets from multiple healthcare institutions, effectively addressing challenges related to non-aligned tabular features. Combining training data enhances breast cancer prediction on the EMBED dataset, outperforming all other experiments. In summary, our research emphasizes the efficacy of frozen large-scale pretrained vision-language models in multimodal breast cancer prediction, offering superior performance and stability over conventional methods, reinforcing their potential for breast cancer risk assessment.

show abstract