Abstract:Deep DCIS evaluates the nuclear grade, necrosis, and stromal reactivity of ADH/DCIS on breast biopsies as continuous variables. DCIS outperforms traditional histological grading in predicting upstaging to invasive carcinoma, and the performance improves after incorporating core clinical features. Aims Risk stratification of atypical ductal hyperplasia (ADH) and ductal carcinoma in situ (DCIS), diagnosed using breast biopsy, has great clinical significance. Clinical trials are currently exploring the possibility of active surveillance for low‐risk lesions, whereas axillary lymph node staging may be considered during surgical planning for high‐risk lesions. We aimed to develop a machine‐learning algorithm based on whole‐slide images of breast biopsy specimens and clinical information to predict the risk of upstaging to invasive breast cancer after wide excision. Methods and Results Patients diagnosed with ADH/DCIS on breast biopsy were included in this study, comprising 592 (740 slides) and 141 (198 slides) patients in the development and independent testing cohorts, respectively. Histological grading of the lesions was independently evaluated by two pathologists. Clinical information, including biopsy method, lesion size, and Breast Imaging Reporting and Data System (BI‐RADS) classification of ultrasound and mammograms, were collected. Deep DCIS consisted of three deep neural networks to evaluate nuclear grade, necrosis, and stromal reactivity. Deep DCIS output comprised five parameters: total patches, lesion extent, Deep Grade, Deep Necrosis, and Deep Stroma. Deep DCIS highly correlated with the pathologists' evaluations of both slide‐ and patient‐level labels. All five parameters of Deep DCIS were significantly associated with upstaging to invasive carcinoma in subsequent wide excisional specimens. Using multivariate logistic regression, Deep DCIS predicted upstaging to invasive carcinoma with an area under the curve (AUC) of 0.81, outperforming pathologists' evaluation (AUC, 0.71 and 0.69). After including clinical and hormone receptor status information, performance further improved (AUC, 0.87). This combined model retained its predictive power in two subgroup analyses: the first subgroup included unequivocal DCIS (excluding cases of ADH and DCIS suspicious for microinvasion) (AUC, 0.83), while the second excluded cases of high‐grade DCIS (AUC, 0.81). The model was validated in an independent testing cohort (AUC, 0.81). Conclusion This study demonstrated that deep‐learning models can refine histological evaluation of ADH and DCIS on breast biopsies, which may help guide future treatment planning.

A Domain Adaptive and Label-efficient Deep Learning Method for Ki-67 Scoring in Infiltrative Breast Ductal Carcinoma Cells.

Unsupervised Domain Adaptive Tumor Region Recognition for Ki67 Automated Assisted Quantification.

Predict Ki-67 Positive Cells in H&E-Stained Images Using Deep Learning Independently from IHC-Stained Images

A deep learning pipeline for breast cancer ki-67 proliferation index scoring

Breast MRI Segmentation and Ki-67 High- and Low-Expression Prediction Algorithm Based on Deep Learning

A Hybrid Deep Learning Approach to Predict Malignancy of Breast Lesions Using Mammograms

Deep Learning-Based Grading of Ductal Carcinoma In Situ in Breast Histopathology Images

Transformation from hematoxylin-and-eosin staining to Ki-67 immunohistochemistry digital staining images using deep learning: experimental validation on the labeling index

Automated quantitative analysis of Ki-67 staining and HE images recognition and registration based on whole tissue sections in breast carcinoma

AI-Powered Segmentation of Invasive Carcinoma Regions in Breast Cancer Immunohistochemical Whole-Slide Images

Domain Adaptation using Silver Standard Labels for Ki-67 Scoring in Digital Pathology: A Step Closer to Widescale Deployment

Cytokeratin-Supervised Deep Learning for Automatic Recognition of Epithelial Cells in Breast Cancers Stained for ER, PR, and Ki-67

Two-Stage Deep Learning Method for Breast Cancer Detection Using High-Resolution Mammogram Images

Deep Learning-Based Multi-Class Classification of Breast Digital Pathology Images

Automated Quantification of HER2 Amplification Levels Using Deep Learning

Deep‐learning model to improve histological grading and predict upstaging of atypical ductal hyperplasia / ductal carcinoma in situ on breast biopsy

DeepSDCS: Dissecting cancer proliferation heterogeneity in Ki67 digital whole slide images

Development and Validation of Fully Automatic Deep Learning-Based Algorithms for Immunohistochemistry Reporting of Invasive Breast Ductal Carcinoma

Histopathological Differential Diagnosis and Estrogen Receptor/Progesterone Receptor Immunohistochemical Evaluation of Breast Carcinoma Using a Deep Learning-Based Artificial Intelligence Architecture

Ultrasound-based artificial intelligence model for prediction of Ki-67 proliferation index in soft tissue tumors

Ki‐67 evaluation using deep‐learning model‐assisted digital image analysis in breast cancer