Deep learning-based risk stratification of preoperative breast biopsies using digital whole slide images DOI Creative Commons
Constance Boissin, Yinxi Wang, Abhinav Sharma

et al.

Breast Cancer Research, Journal Year: 2024, Volume and Issue: 26(1)

Published: June 3, 2024

Abstract Background Nottingham histological grade (NHG) is a well established prognostic factor in breast cancer histopathology but has high inter-assessor variability with many tumours being classified as intermediate grade, NHG2. Here, we evaluate if DeepGrade, previously developed model for risk stratification of resected tumour specimens, could be applied to risk-stratify biopsy specimens. Methods A total 11,955,755 tiles from 1169 whole slide images preoperative biopsies 896 patients diagnosed Stockholm, Sweden, were included. deep convolutional neural network model, was the prediction low- and high-risk tumours. It evaluated against clinically assigned grades NHG1 NHG3 on specimen also corresponding resection using area under operating curve (AUC). The value DeepGrade setting time-to-event analysis. Results Based images, predicted cases clinical an AUC 0.908 (95% CI: 0.88; 0.93). Furthermore, out 432 clinically-assigned NHG2 tumours, 281 (65%) DeepGrade-low 151 (35%) DeepGrade-high. Using multivariable Cox proportional hazards hazard ratio between groups estimated 2.01 1.06; 3.79). Conclusions provided only specimen. results demonstrate that can provide decision support identify based biopsies, thus improving early treatment decisions.

Language: Английский

Wrapper-based deep feature optimization for activity recognition in the wearable sensor networks of healthcare systems DOI Creative Commons
Karam Kumar Sahoo,

Raghunath Ghosh,

Saurav Mallik

et al.

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Jan. 18, 2023

Abstract The Human Activity Recognition (HAR) problem leverages pattern recognition to classify physical human activities as they are captured by several sensor modalities. Remote monitoring of an individual’s has gained importance due the reduction in travel and during pandemic. Research on HAR enables one person either remotely monitor or recognize another person’s activity via ubiquitous mobile device using sensor-based Internet Things (IoT). Our proposed work focuses accurate classification daily from both accelerometer gyroscope data after converting into spectrogram images. feature extraction process follows leveraging pre-trained weights two popular efficient transfer learning convolutional neural network models. Finally, a wrapper-based selection method been employed for selecting optimal subset that reduces training time improves final performance. model tested three benchmark datasets namely, HARTH, KU-HAR HuGaDB achieved 88.89%, 97.97% 93.82% respectively these datasets. It is be noted achieves improvement about 21%, 20% 6% overall accuracies while utilizing only 52%, 45% 60% original set HuGaDB, HARTH respectively. This proves effectiveness our methodology.

Language: Английский

Citations

28
Accuracy Analysis of Deep Learning Methods in Breast Cancer Classification: A Structured Review DOI Creative Commons
Marina Yusoff, Toto Haryanto, Heru Suhartanto

et al.

Diagnostics, Journal Year: 2023, Volume and Issue: 13(4), P. 683 - 683

Published: Feb. 11, 2023

Breast cancer is diagnosed using histopathological imaging. This task extremely time-consuming due to high image complexity and volume. However, it important facilitate the early detection of breast for medical intervention. Deep learning (DL) has become popular in imaging solutions demonstrated various levels performance diagnosing cancerous images. Nonetheless, achieving precision while minimizing overfitting remains a significant challenge classification solutions. The handling imbalanced data incorrect labeling further concern. Additional methods, such as pre-processing, ensemble, normalization techniques, have been established enhance characteristics. These methods could influence be used overcome balancing issues. Hence, developing more sophisticated DL variant improve accuracy reducing overfitting. Technological advancements fueled automated diagnosis growth recent years. paper reviewed studies on capability classify images, objective this study was systematically review analyze current research Additionally, literature from Scopus Web Science (WOS) indexes reviewed. assessed approaches applications papers published up until November 2022. findings suggest that especially convolution neural networks their hybrids, are most cutting-edge currently use. To find new technique, necessary first survey landscape existing hybrid conduct comparisons case studies.

Language: Английский

Citations

23
A hybrid lightweight breast cancer classification framework using the histopathological images DOI
Daniel Addo, Shijie Zhou, Kwabena Sarpong

et al.

Journal of Applied Biomedicine, Journal Year: 2023, Volume and Issue: 44(1), P. 31 - 54

Published: Dec. 22, 2023

Language: Английский

Citations

23
Deep learning approaches to detect breast cancer: a comprehensive review DOI

Amir Mohammad Sharafaddini,

Kiana Kouhpah Esfahani,

N. Mansouri

et al.

Multimedia Tools and Applications, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 20, 2024

Language: Английский

Citations

10
Decoding the black box: Explainable AI (XAI) for cancer diagnosis, prognosis, and treatment planning-A state-of-the art systematic review DOI

Youssef Alaaeldin Ali Mohamed,

Bee Luan Khoo,

Mohd Shahrimie Mohd Asaari

et al.

International Journal of Medical Informatics, Journal Year: 2024, Volume and Issue: 193, P. 105689 - 105689

Published: Nov. 4, 2024

Language: Английский

Citations

9
A patch-based deep learning framework with 5-B network for breast cancer multi-classification using histopathological images DOI
Jehoiada Jackson,

L. Jackson,

Chiagoziem C. Ukwuoma

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 148, P. 110439 - 110439

Published: March 3, 2025

Language: Английский

Citations

1
MDFF-Net: A multi-dimensional feature fusion network for breast histopathology image classification DOI Open Access
Cheng Xu, Ke Yi, Nan Jiang

et al.

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 165, P. 107385 - 107385

Published: Aug. 16, 2023

Language: Английский

Citations

18
Deep learning approaches for breast cancer detection in histopathology images: A review DOI Creative Commons

Lakshmi Priya C V,

Vinai George Biju,

B. Vinod

et al.

Cancer Biomarkers, Journal Year: 2024, Volume and Issue: 40(1), P. 1 - 25

Published: Feb. 1, 2024

BACKGROUND: Breast cancer is one of the leading causes death in women worldwide. Histopathology analysis breast tissue an essential tool for diagnosing and staging cancer. In recent years, there has been a significant increase research exploring use deep-learning approaches detection from histopathology images. OBJECTIVE: To provide overview current state-of-the-art technologies automated images using deep learning techniques. METHODS: This review focuses on algorithms classification We publicly available image datasets detection. also highlight strengths weaknesses these architectures their performance different datasets. Finally, we discuss challenges associated with techniques detection, including need large diverse interpretability models. RESULTS: Deep have shown great promise accurately detecting classifying Although accuracy levels vary depending specific data set, pre-processing techniques, architecture used, results potential improving efficiency CONCLUSION: presented thorough account The integration machine demonstrated promising identifying insights gathered this can act as valuable reference researchers field who are developing diagnostic strategies Overall, objective to spark interest among scholars complex acquaint them cutting-edge

Language: Английский

Citations

8
CCF-GNN: A Unified Model Aggregating Appearance, Microenvironment, and Topology for Pathology Image Classification DOI
Hongxiao Wang, Gang Huang, Zhuo Zhao

et al.

IEEE Transactions on Medical Imaging, Journal Year: 2023, Volume and Issue: 42(11), P. 3179 - 3193

Published: Feb. 27, 2023

Pathology images contain rich information of cell appearance, microenvironment, and topology features for cancer analysis diagnosis. Among such features, becomes increasingly important in immunotherapy. By analyzing geometric hierarchically structured distribution topology, oncologists can identify densely-packed cancer-relevant communities (CCs) making decisions. Compared to commonly-used pixel-level Convolution Neural Network (CNN) cell-instance-level Graph (GNN) CC are at a higher level granularity geometry. However, topological have not been well exploited by recent deep learning (DL) methods pathology image classification due lack effective descriptors gathering patterns. In this paper, inspired clinical practice, we analyze classify comprehensively fine-to-coarse manner. To describe exploit design Cell Community Forest (CCF), novel graph that represents the hierarchical formulation process big-sparse CCs from small-dense CCs. Using CCF as new descriptor tumor cells images, propose CCF-GNN, GNN model successively aggregates heterogeneous (e.g., microenvironment) cell-instance-level, cell-community-level, into image-level classification. Extensive cross-validation experiments show our method significantly outperforms alternative on H&E-stained immunofluorescence disease grading tasks with multiple types. Our proposed CCF-GNN establishes data (TDA) based method, which facilitates integrating multi-level point clouds cells) unified DL framework.

Language: Английский

Citations

14
Exploratory drug discovery in breast cancer patients: A multimodal deep learning approach to identify novel drug candidates targeting RTK signaling DOI

Anush Karampuri,

Sunitha Kundur,

Perugu Shyam

et al.

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 174, P. 108433 - 108433

Published: April 16, 2024

Language: Английский

Citations

6