Combining State-of-the-Art Pre-Trained Deep Learning Models: A Noble Approach for Skin Cancer Detection Using Max Voting Ensemble

Abstract

Skin cancer poses a significant healthcare challenge, requiring precise and prompt diagnosis for effective treatment. While recent advances in deep learning have dramatically improved medical image analysis, including skin cancer classification, ensemble methods offer a pathway for further enhancing diagnostic accuracy. This study introduces a cutting-edge approach employing the Max Voting Ensemble Technique for robust skin cancer classification on ISIC 2018: Task 1-2 dataset. We incorporate a range of cutting-edge, pre-trained deep neural networks, including MobileNetV2, AlexNet, VGG16, ResNet50, DenseNet201, DenseNet121, InceptionV3, ResNet50V2, InceptionResNetV2, and Xception. These models have been extensively trained on skin cancer datasets, achieving individual accuracies ranging from 77.20% to 91.90%. Our method leverages the synergistic capabilities of these models by combining their complementary features to elevate classification performance further. In our approach, input images undergo preprocessing for model compatibility. The ensemble integrates the pre-trained models with their architectures and weights preserved. For each skin lesion image under examination, every model produces a prediction. These are subsequently aggregated using the max voting ensemble technique to yield the final classification, with the majority-voted class serving as the conclusive prediction. Through comprehensive testing on a diverse dataset, our ensemble outperformed individual models, attaining an accuracy of 93.18% and an AUC score of 0.9320, thus demonstrating superior diagnostic reliability and accuracy. We evaluated the effectiveness of our proposed method on the HAM10000 dataset to ensure its generalizability. Our ensemble method delivers a robust, reliable, and effective tool for the classification of skin cancer. By utilizing the power of advanced deep neural networks, we aim to assist healthcare professionals in achieving timely and accurate diagnoses, ultimately reducing mortality rates and enhancing patient outcomes.

Keywords: classification; deep learning; max voting; medical imaging; skin cancer; transfer learning.

Figure 1

Skin Cancer—Malignant and benign sample images [3].

Figure 2

Abstract view of proposed Max Voting-based skin cancer classification.

Figure 3

Overall workflow architecture of the proposed max voting-based ensemble technique.

Figure 4

Skin Cancer—Malignant and benign sample images.

Figure 5

Flowchart depicting the steps involved in individual training and testing the models (Models: MobileNetV2, AlexNet, VGG16, ResNet50, DenseNet201, DenseNet121, InceptionV3, ResNet50V2, InceptionResNetV2, and Xception).

Figure 6

Training Batch size and Learning Rate versus Accuracy.

Figure 7

Image size vs. accuracy and preprocessing batch size vs. accuracy.

Figure 8

Dropout vs. accuracy and layers vs. accuracy.

Figure 9

Accuracy and Loss Curve of Xception Model for 30 Epochs.

Figure 10

Accuracy and Loss Curve of Xception Model for 200 Epochs.

Figure 11

ROC Curve and AUC Scores for the MobileNetV2, Xception, and Max Voting.

Figure 12

Confusion Matrix: Individual Models (MobileNetV2, Xception) Versus Proposed (Max Voting-based Ensemble) Model.

Figure 13

ROC Curve and AUC Scores for the models (Models: MobileNetV2, AlexNet, VGG16, ResNet50, DenseNet201, DenseNet201, InceptionV3, ResNet50V2, InceptionResNetV2, Xception and Max Voting).

Figure 14

ROC Curve and AUC Scores for HAM10000 Dataset using proposed Max Voting Ensemble Technique.

Figure 15

ROC Curve and Confusion Matrix using Max Voting Technique for HAM10000 Dataset.