Towards Automated Melanoma Detection with Deep Learning: Data - - PowerPoint PPT Presentation

Motivation Related work Proposed approach Empirical results Conclusion

Towards Automated Melanoma Detection with Deep Learning: Data Purification and Augmentation

Devansh Bisla, Anna Choromanska, Russell S. Berman, Jennifer A. Stein, David Polsky

New York University, New York, NY, USA

Code: https://bit.ly/2KFRp5e Paper: https://bit.ly/2FBgOZP

Motivation Related work Proposed approach Empirical results Conclusion

Motivation

Figure 1: (Left) Data Imbalancedness (Right) Data Impurities

Motivation Related work Proposed approach Empirical results Conclusion

Existing computational techniques

Traditional machine learning

Hand-crafted extraction of features from the data such as

Lesion Symmetry/Asymmetry. Irregular borders. Non-Uniform pigmentation. Lesion size.

Problem: not scalable to large data sets.

Deep Learning

Automatically extract features from large sized data. Problem: Needs large, balanced, and unbiased data.

Motivation Related work Proposed approach Empirical results Conclusion

Traditional training

Visualization results for the conventionally-trained model (Top): Original

• image. (Bottom): Visualization mask overlaid on the original image.

The model overfits to image occlusions such as hairs, rulers and ink marks.

Motivation Related work Proposed approach Empirical results Conclusion

Proposed approach

Data Impurities:

Removal of unwanted objects such as hair, rulers etc.

Data Imbalancedness

Synthetic data generation. Data augmentation.

Motivation Related work Proposed approach Empirical results Conclusion

Data purification

Thresholding in the LUV color space combined with morphological operations. Note that this may also remove dark regions belonging to the lesion itself.[Philippe Schmid-Saugeon et al] Overlay the processed image with the segmented lesion

• btained from our segmentation algorithm.

Motivation Related work Proposed approach Empirical results Conclusion

Data purification - results

(a) (b) (c) (d) (e)

Figure 2: Top: Original images. Bottom: Images obtained after a,b) scales, c) hairs and scales, and d,e) hairs removal.

Motivation Related work Proposed approach Empirical results Conclusion

Data generation

Figure 3: Architecture of Generative Adversarial Network

Main idea: Train a generator network to generate images which have similar distribution to the one followed by the training data, but do not appear in the training data set. The discriminator provides a feedback on similarity between the two distributions. We generated 350 images of melanoma and 750 images of seborrheic keratosis.

Motivation Related work Proposed approach Empirical results Conclusion

Data generation - results

0.02 0.04 0.06 0.08 0.10 Mean Squared Error 10 20 30 40 50 60 Frequency Histogram for Seborrheic Keratosis 0.0 0.1 0.2 0.3 Mean Squared Error 10 20 30 40 50 60 70 Frequency Histogram for Melanoma

Figure 4: Histograms of the MSE values for (left) seborrheic keratosis and (right) melanoma.

Motivation Related work Proposed approach Empirical results Conclusion

Data generation - results

Seborrheic Keratosis 0.02 0.04 0.059 Melanoma 0.02 0.09 0.18

Motivation Related work Proposed approach Empirical results Conclusion

Classification results: confusion matrix

M N S K Predicted label M N SK True label 80 19 18 89 269 35 12 6 72 Confusion matrix M N S K Predicted label M N SK True label 83 23 11 38 338 17 6 15 69 Confusion matrix

Figure 5: Confusion matrix obtained by traditional baseline (left) and proposed model (right).

Motivation Related work Proposed approach Empirical results Conclusion

Classification results: ROC-AUC

Mean Value ROC-AUC Our Approach 0.915 Kazuhisa Matsunaga[K. Matsunaga et al.] 0.911 RECOD Titans[A. Menegola et al.] 0.908

Table 1: Leader-board for melanoma and seborrheic keratosis combined.

Method 82% 89% 95% Top AVG[K. Matsunaga et al.] 0.729 0.588 0.366 Top SK [I. Gonzalez Diaz et al.] 0.727 0.555 0.404 Top M [A. Menegola et al.] 0.747 0.590 0.395 Our Approach 0.697 0.648 0.492

Table 2: Specificity values at sensitivity levels of 82%/89%/95% for melanoma classification. Top AVG, Top SK, and Top M denote the winning approaches of the ISIC 2017 challenge.

Motivation Related work Proposed approach Empirical results Conclusion

Classification results visualized

TP FP FN TN

Figure 6: Visualization results for seborrheic keratosis. Top: Original

• image. Bottom: Visualization result.

Motivation Related work Proposed approach Empirical results Conclusion

Classification results visualized

TP FP FN TN

Figure 7: Visualization results for Nevus. Top: Original image. Bottom: Visualization result.

Motivation Related work Proposed approach Empirical results Conclusion

Conclusion

Deep learning based methods are the most accurate and scalable, but they require large, pure and balanced training data sets. We presented solutions to improve effectiveness of classification systems by data purification (removal of unwanted objects) and data augmentation (synthetic data generation).