Deep learning-based classification of lymphedema and other lower limb edema diseases using clinical images

Abstract

Lymphedema is a chronic condition characterized by lymphatic fluid accumulation, primarily affecting the limbs. Its diagnosis is challenging due to symptom overlap with conditions like chronic venous insufficiency (CVI), deep vein thrombosis (DVT), and systemic diseases, often leading to diagnostic delays that can extend up to ten years. These delays negatively impact patient outcomes and burden healthcare systems. Conventional diagnostic methods rely heavily on clinical expertise, which may fail to distinguish subtle variations between these conditions. This study investigates the application of artificial intelligence (AI), specifically deep learning, to improve diagnostic accuracy for lower limb edema. A dataset of 1622 clinical images was used to train sixteen convolutional neural networks (CNNs) and transformer-based models, including EfficientNetV2, which achieved the highest accuracy of 78.6%. Grad-CAM analyses enhanced model interpretability, highlighting clinically relevant features such as swelling and hyperpigmentation. The AI system consistently outperformed human evaluators, whose diagnostic accuracy plateaued at 62.7%. The findings underscore the transformative potential of AI as a diagnostic tool, particularly in distinguishing conditions with overlapping clinical presentations. By integrating AI with clinical workflows, healthcare systems can reduce diagnostic delays, enhance accuracy, and alleviate the burden on medical professionals. While promising, the study acknowledges limitations, such as dataset diversity and the controlled evaluation environment, which necessitate further validation in real-world settings. This research highlights the potential of AI-driven diagnostics to revolutionize lymphedema care, bridging gaps in conventional methods and supporting healthcare professionals in delivering more precise and timely interventions. Future work should focus on external validation and hybrid systems integrating AI and clinical expertise for comprehensive diagnostic solutions.

Fig. 1

Workflow diagram illustrating the steps involved in developing and validating the deep-learning for lower limb edema, from data collection to statistical evaluation.

Fig. 2

Example photographs illustrating each classification category .

Fig. 3

A controlled picture acquisition setup demonstrating standardized lighting, fixed distances, a consistent background, and calibrated camera settings.

Fig. 4

Workflow of lower leg image preprocessing: cropped, padded, resized.

Fig. 5

Illustration of the Architecture of EfficientNetV2, highlighting the use of Fused-MBConv layers for early-stage efficiency and MBConv layers for complex feature extraction. The comparison structure shows how Fused-MBConv replaces depthwise convolutions with standard convolutions, enhancing computational efficiency, while MBConv retains depthwise separable convolutions for capturing more detailed features.

Fig. 6

Grad-CAM visualizations illustrating the regions of focus for CNN and transformer-based models when classifying normal, systemic disease, CVI&DVT, and lymphedema conditions.

Fig. 7

Receiver Operating Characteristic (ROC) curves of CNN and transformer-based models, illustrating their performance in classifying CVI&DVT, lymphedema, systemic disease, and normal conditions using the validation dataset.

Fig. 8

Confusion matrices showing the validation results for eight different deep learning models in classifying CVI & DVT, lymphedema, normal, and systemic disease conditions.

Fig. 9

Receiver Operating Characteristic (ROC) curves of CNN and transformer-based models, illustrating their performance in classifying CVI&DVT, lymphedema, systemic disease, and normal conditions using the 28-question spot-diagnosis quiz.

Fig. 10

Confusion matrices displaying the classification results of eight different deep learning models from a 28-question spot-diagnosis quiz, assessing their ability to classify CVI & DVT, lymphedema, normal, and systemic disease conditions.

Fig. 11

Confusion matrices displaying the diagnostic accuracy results based on professional role and experience level from a 28-question spot-diagnosis quiz, assessing their ability to classify CVI & DVT, lymphedema, normal, and systemic disease conditions.