PapersFlow Research Brief
AI in cancer detection
Research Guide
What is AI in cancer detection?
AI in cancer detection is the application of deep learning and machine learning techniques, such as convolutional neural networks, to analyze medical images including histopathology, digital pathology, and whole slide imaging for computer-aided detection, diagnosis, prognosis, and prediction of cancers like breast and skin cancer.
This field encompasses 76,184 works focused on medical image analysis using deep learning for cancer detection. Studies emphasize convolutional neural networks and whole slide imaging to enhance accuracy in breast cancer diagnosis and cancer prognosis. Key advancements include tools like QuPath for digital pathology image analysis.
Topic Hierarchy
Research Sub-Topics
Deep Learning for Histopathology Image Analysis
Researchers apply CNNs and transformers to whole slide images for tumor detection, grading, and subtype classification in histopathology. Studies emphasize data augmentation, transfer learning, and annotation efficiency.
Radiomics in Cancer Prognosis
This sub-topic extracts high-dimensional features from medical images to predict tumor outcomes, recurrence, and treatment response using machine learning. Validation involves multi-center cohorts and feature selection methods.
Computer-Aided Detection in Breast Cancer
Developments in CAD systems for mammography, ultrasound, and MRI using DL for lesion detection and false positive reduction. Performance is benchmarked against radiologists in clinical trials.
Digital Pathology and Whole Slide Imaging
This area advances software tools like QuPath for WSI processing, AI integration, and telepathology workflows. Research addresses scanning artifacts, standardization, and computational scalability.
Data Augmentation Techniques for Medical Imaging
Investigations into GANs, mixup, and elastic deformations to augment limited cancer imaging datasets for robust DL model training. Evaluations measure generalization on unseen pathologies.
Why It Matters
AI in cancer detection supports computer-aided detection systems that match dermatologist performance in classifying skin cancer, as Esteva et al. (2017) showed deep neural networks achieving dermatologist-level accuracy on 129,450 skin lesion images in "Dermatologist-level classification of skin cancer with deep neural networks". Litjens et al. (2017) reviewed deep learning applications across medical imaging tasks, including cancer detection in histopathology and radiology, in "A survey on deep learning in medical image analysis" with 13,164 citations. Bankhead et al. (2017) introduced QuPath, enabling high-throughput tumor identification and biomarker evaluation on whole slide images, as detailed in "QuPath: Open source software for digital pathology image analysis" with 7,826 citations. These contributions improve diagnostic efficiency in histopathology for breast cancer and digital pathology workflows.
Reading Guide
Where to Start
"A survey on deep learning in medical image analysis" by Litjens et al. (2017) provides a foundational overview of deep learning applications in medical imaging including cancer detection tasks, making it ideal for beginners to grasp core methods and challenges.
Key Papers Explained
Litjens et al. (2017) in "A survey on deep learning in medical image analysis" reviews convolutional neural networks for histopathology and cancer detection, setting context for Esteva et al. (2017)'s "Dermatologist-level classification of skin cancer with deep neural networks" which applies these networks to achieve expert-level skin lesion classification. Bankhead et al. (2017)'s "QuPath: Open source software for digital pathology image analysis" builds on imaging needs by offering tools for whole slide analysis, complementing Shorten and Khoshgoftaar (2019)'s "A survey on Image Data Augmentation for Deep Learning" that addresses data limitations in such analyses. Gillies et al. (2015)'s "Radiomics: Images Are More than Pictures, They Are Data" introduces quantitative feature extraction foundational to AI enhancements.
Paper Timeline
Most-cited paper highlighted in red. Papers ordered chronologically.
Advanced Directions
Research continues to emphasize convolutional neural networks for histopathology and whole slide imaging in breast cancer diagnosis. Focus persists on improving computer-aided detection accuracy and cancer prognosis via digital pathology tools.
Papers at a Glance
| # | Paper | Year | Venue | Citations | Open Access |
|---|---|---|---|---|---|
| 1 | A survey on deep learning in medical image analysis | 2017 | Medical Image Analysis | 13.2K | ✓ |
| 2 | Dermatologist-level classification of skin cancer with deep ne... | 2017 | Nature | 12.8K | ✓ |
| 3 | A survey on Image Data Augmentation for Deep Learning | 2019 | Journal Of Big Data | 11.4K | ✓ |
| 4 | Radiomics: Images Are More than Pictures, They Are Data | 2015 | Radiology | 7.8K | ✕ |
| 5 | QuPath: Open source software for digital pathology image analysis | 2017 | Scientific Reports | 7.8K | ✓ |
| 6 | Unified segmentation | 2005 | NeuroImage | 7.4K | ✕ |
| 7 | A method of comparing the areas under receiver operating chara... | 1983 | Radiology | 7.0K | ✕ |
| 8 | The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) | 2014 | IEEE Transactions on M... | 6.1K | ✓ |
| 9 | ImageJ2: ImageJ for the next generation of scientific image data | 2017 | BMC Bioinformatics | 6.0K | ✓ |
| 10 | Radiomics: Extracting more information from medical images usi... | 2012 | European Journal of Ca... | 5.6K | ✓ |
Frequently Asked Questions
What role do convolutional neural networks play in AI for cancer detection?
Convolutional neural networks analyze histopathology images and whole slide imaging for cancer detection. Litjens et al. (2017) surveyed their use in medical image analysis tasks including cancer diagnosis in "A survey on deep learning in medical image analysis". They improve accuracy in computer-aided detection for breast cancer diagnosis.
How does AI assist in skin cancer classification?
Deep neural networks classify skin cancer at dermatologist-level accuracy. Esteva et al. (2017) trained networks on 129,450 images to distinguish keratinocyte carcinomas from benign seborrheic keratoses in "Dermatologist-level classification of skin cancer with deep neural networks". This supports clinical decision-making in dermatology.
What is QuPath used for in digital pathology?
QuPath is open source software for whole slide image analysis in digital pathology. Bankhead et al. (2017) developed it for tumor identification and biomarker evaluation in "QuPath: Open source software for digital pathology image analysis". It handles high-throughput analysis of histopathology images.
What are radiomics in cancer detection?
Radiomics extracts quantitative features from medical images for cancer analysis. Gillies et al. (2015) described radiomics as processes for high-throughput feature extraction in "Radiomics: Images Are More than Pictures, They Are Data". Lambin et al. (2012) detailed advanced feature analysis for tumor characterization in "Radiomics: Extracting more information from medical images using advanced feature analysis".
How is image data augmentation applied in deep learning for cancer detection?
Image data augmentation prevents overfitting in convolutional neural networks reliant on large datasets. Shorten and Khoshgoftaar (2019) surveyed techniques to expand training data in "A survey on Image Data Augmentation for Deep Learning". It enhances model performance in medical image analysis for cancer tasks.
What benchmarks exist for brain tumor segmentation?
The BRATS benchmark evaluates multimodal brain tumor image segmentation algorithms. Menze et al. (2014) reported results from 20 algorithms on 65 multi-contrast MR scans in "The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)". It standardizes performance assessment in glioma detection.
Open Research Questions
- ? How can deep learning models generalize across diverse histopathology datasets for breast cancer detection beyond current benchmarks?
- ? What integration strategies between radiomics features and convolutional neural networks optimize cancer prognosis predictions?
- ? How do whole slide imaging analysis tools like QuPath scale to real-time clinical workflows for multiple cancer types?
- ? Which data augmentation methods most effectively mitigate overfitting in small-sample cancer imaging studies?
- ? What multimodal fusion techniques improve segmentation accuracy in brain tumor detection as in BRATS challenges?
Recent Trends
The field maintains 76,184 works with prominent use of convolutional neural networks in histopathology for breast cancer diagnosis.
Surveys like Litjens et al. with 13,164 citations and Esteva et al. (2017) with 12,836 citations underscore sustained emphasis on deep learning in medical image analysis and skin cancer classification.
2017Research AI in cancer detection with AI
PapersFlow provides specialized AI tools for Computer Science researchers. Here are the most relevant for this topic:
AI Literature Review
Automate paper discovery and synthesis across 474M+ papers
Code & Data Discovery
Find datasets, code repositories, and computational tools
Deep Research Reports
Multi-source evidence synthesis with counter-evidence
AI Academic Writing
Write research papers with AI assistance and LaTeX support
See how researchers in Computer Science & AI use PapersFlow
Field-specific workflows, example queries, and use cases.
Start Researching AI in cancer detection with AI
Search 474M+ papers, run AI-powered literature reviews, and write with integrated citations — all in one workspace.
See how PapersFlow works for Computer Science researchers