What question did this study set out to answer?

The aim is to develop and evaluate a deep learning model to classify pigmented skin lesions and compare its accuracy against dermatologists.

July 29, 2020Open Access

The Development of a Skin Cancer Classification System for Pigmented Skin Lesions Using Deep Learning

Puntos clave

The aim is to develop and evaluate a deep learning model to classify pigmented skin lesions and compare its accuracy against dermatologists.
Extracted 5846 clinical images from 3551 patients, including both malignant and benign lesions.
Created a training dataset with bounding-box annotations for 4732 images and a test dataset with 666 images from 666 patients.
Trained a faster region-based convolutional neural network (FRCNN) and compared its performance with ten board-certified dermatologists and ten dermatologic trainees on clinical image classification.
FRCNN achieved 86.2% accuracy in six-class classification, outperforming board-certified dermatologists (79.5%, p=0.0081) and trainees (75.1%, p<0.00001).
For two-class classification, FRCNN had 91.5% accuracy, 83.3% sensitivity, and 94.5% specificity compared to 86.6%, 86.3%, 86.6% for BCDs and 85.3%, 83.5%, 85.9% for TRNs.
FRCNN demonstrated lower false positive rates (5.5%) and higher positive predictive value (84.7%) than both BCDs and TRNs.

Resumen

Recent studies have demonstrated the usefulness of convolutional neural networks (CNNs) to classify images of melanoma, with accuracies comparable to those achieved by dermatologists. However, the performance of a CNN trained with only clinical images of a pigmented skin lesion in a clinical image classification task, in competition with dermatologists, has not been reported to date. In this study, we extracted 5846 clinical images of pigmented skin lesions from 3551 patients. Pigmented skin lesions included malignant tumors (malignant melanoma and basal cell carcinoma) and benign tumors (nevus, seborrhoeic keratosis, senile lentigo, and hematoma/hemangioma). We created the test dataset by randomly selecting 666 patients out of them and picking one image per patient, and created the training dataset by giving bounding-box annotations to the rest of the images (4732 images, 2885 patients). Subsequently, we trained a faster, region-based CNN (FRCNN) with the training dataset and checked the performance of the model on the test dataset. In addition, ten board-certified dermatologists (BCDs) and ten dermatologic trainees (TRNs) took the same tests, and we compared their diagnostic accuracy with FRCNN. For six-class classification, the accuracy of FRCNN was 86.2%, and that of the BCDs and TRNs was 79.5% (p = 0.0081) and 75.1% (p < 0.00001), respectively. For two-class classification (benign or malignant), the accuracy, sensitivity, and specificity were 91.5%, 83.3%, and 94.5% by FRCNN; 86.6%, 86.3%, and 86.6% by BCD; and 85.3%, 83.5%, and 85.9% by TRN, respectively. False positive rates and positive predictive values were 5.5% and 84.7% by FRCNN, 13.4% and 70.5% by BCD, and 14.1% and 68.5% by TRN, respectively. We compared the classification performance of FRCNN with 20 dermatologists. As a result, the classification accuracy of FRCNN was better than that of the dermatologists. In the future, we plan to implement this system in society and have it used by the general public, in order to improve the prognosis of skin cancer.

Me gusta

Guardar

Ver artículo completo