Source-linked AI summary
A Multi-Scale CNN and Curriculum Learning Strategy for Mammogram Classification
William Lotter, Greg Sorensen, David Cox
TL;DR
Full-image mammogram classification is difficult because small, subtle lesions create a “needle in a haystack” problem, while existing approaches often target pipeline subtasks. The paper uses lesion-specific, multi-scale patch classifiers to initialize a scanning-window CNN trained end-to-end, achieving 0.92 AUROC on DDSM. The results support this architecture and curriculum strategy as a promising approach for mammogram classification, while future work seeks to reduce reliance on hand-drawn segmentation masks.
Problem
Method
Results
0.92 AUROC was achieved on the DDSM dataset against final pathology outcomes.
Source: §1 · p. 2Takeaways & Limitations
The multi-scale CNN and lesion-specific curriculum strategy provide a promising approach for full-image mammogram classification and outperform standard out-of-the-box CNN models.
Source: §4 · p. 7
Abstract
from arXiv · showhide
Screening mammography is an important front-line tool for the early detection of breast cancer, and some 39 million exams are conducted each year in the United States alone. Here, we describe a multi-scale convolutional neural network (CNN) trained with a curriculum learning strategy that achieves high levels of accuracy in classifying mammograms. Specifically, we first train CNN-based patch classifiers on segmentation masks of lesions in mammograms, and then use the learned features to initialize a scanning-based model that renders a decision on the whole image, trained end-to-end on outcome data. We demonstrate that our approach effectively handles the "needle in a haystack" nature of full-image mammogram classification, achieving 0.92 AUROC on the DDSM dataset.
1 Introduction
Screening mammography supports early breast-cancer detection but remains difficult to interpret reliably because lesions are small and subtle, readings are tedious, and false positives impose costs. The paper addresses whole-image cancer-status classification, a challenging “needle in a haystack” problem, with a two-stage deep-learning approach achieving 0.92 AUROC on DDSM.
- Screening mammography is important for early breast-cancer detection, but interpretation is tedious, error-prone, and associated with false positives, cost, and patient stress.Source: §1 · p. 1
- Many prior deep-learning approaches address individual mammography subtasks rather than binary cancer-status classification from an entire mammogram.Source: §1 · p. 1
- The proposed curriculum strategy trains multi-scale patch classifiers before using their features to initialize an end-to-end image-level model.Source: §1 · p. 2
- 0.92 AUROC was achieved on DDSM against final pathology outcomes.Source: §1 · p. 2
2 Multi-Scale CNN with Curriculum Learning Strategy
The method uses lesion-specific, multi-scale patch learning to initialize a scanning-window model for full-image classification. It combines patch features across scales through pooling and concatenation, then fine-tunes the complete system end-to-end.
Motivation
- Lesions occupy small image regions, with median masks covering about 0.5% of calcification images and 1.2% of mass images.The high-resolution DDSM images require detecting fine detail in small regions.Source: §2 · p. 2
Patch-level curriculum
- The first stage trains separate Wide ResNet patch classifiers for calcifications and masses using segmentation-derived labels.Training proceeds from abnormality detection to pathology-determined malignancy.Source: §2 · p. 3
Image-level training
- The second stage tiles each mammogram with minimally overlapping patches and extracts features using the trained patch models.Source: §2 · p. 3
- Global average pooling aggregates patch features at each scale, followed by concatenation and a softmax classifier.Source: Figure 2 · p. 3, §2 · p. 4
- The image-level model is fine-tuned end-to-end using binary image-level labels.Source: §2 · pp. 3–4
3 Experiments
Experiments evaluate the model on patient-separated DDSM data using breast-level predictions and compare it with InceptionV3 and ablated training choices. The proposed model reaches 0.92 ± 0.02 AUC, while removing size augmentation or lesion pre-training reduces performance.
Dataset and evaluation
- DDSM contains 10,480 images from 2,620 cases, split into 87% training, 5% validation, and 8% testing by patient.Each case includes CC and MLO views for both breasts.Source: §3 · p. 4
Training setup
- The image-level model uses initialized patch weights, multi-scale 256x256 patches, overlap normalization, and malignant-lesion labels aggregated across breast views.Source: §3 · p. 5
Baseline
- InceptionV3 cannot directly use random input-size augmentation because it assumes a fixed input size.Source: §3 · p. 6, Figure 4 · p. 6
Ablations
- 0.89 ± 0.02 AUC resulted without size-augmentation training, compared with 0.92 ± 0.02 AUC for the full model.Source: §3 · pp. 6–7
Baselines and ablations
- 0.65 ± 0.04 AUC resulted without DDSM lesion pre-training, while InceptionV3 achieved 0.77 ± 0.03 with and 0.59 ± 0.04 without ImageNet pre-training.Source: §3 · p. 7
4 Conclusions
The paper presents a multi-scale CNN with lesion-specific curriculum learning for full-image mammogram classification. It reports promising performance relative to standard CNNs while identifying architecture and training strategy as important components and proposing less reliance on hand-drawn masks as future work.
- The proposed multi-scale scanning-window CNN with lesion-specific curriculum learning achieves promising full-image mammogram classification results.Source: §4 · p. 7
- The approach performs significantly better than standard out-of-the-box CNN models.Source: §4 · p. 7
- Experiments indicate that both architecture choice and training scheme are important for performance.Source: §4 · p. 7
- Future work aims to learn interest points more unsupervisedly to reduce reliance on hand-drawn segmentation masks.Source: §4 · p. 7