Related Experiment Video
Updated: Oct 1, 2025
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
An Efficient Stacked Deep Transfer Learning Model for Automated Diagnosis of Lyme Disease
Ahmad Ali AlZubi1, Shailendra Tiwari2, Kuldeep Walia3
1Computer Science Department, Community College, King Saud University, Riyadh, Saudi Arabia.
Abstract:
Lyme disease is one of the most common vector-borne infections. It typically causes cardiac illnesses, neurologic illnesses, musculoskeletal disorders, and dermatologic conditions. However, most of the time, it is poorly diagnosed due to many similarities with other diseases such as drug rash. Given the potentially serious consequences of unnecessary antimicrobial treatments, it is essential to understand frequent and uncommon diagnoses that explain symptoms in this population. Recently, deep learning models have been used for the diagnosis of various rash-related diseases. However, these models suffer from overfitting and color variation problems. To overcome these problems, an efficient stacked deep transfer learning model is proposed that can efficiently distinguish between patients infected with Lyme (+) or infected with other infections. 2nd order edge-based color constancy is used as a preprocessing approach to reduce the impact of multisource light from images acquired under different setups. The AlexNet pretrained learning model is used for building the Lyme disease diagnosis model. To prevent overfitting, data augmentation techniques are also used to augment the dataset. In addition, 5-fold cross-validation is also used. Comparative analysis indicates that the proposed model outperforms the existing models in terms of accuracy, f-measure, sensitivity, specificity, and area under the curve.
Insights
This study introduces an improved deep learning model for diagnosing Lyme disease from skin images, overcoming common challenges like overfitting and color variation. The new model accurately distinguishes Lyme disease from other conditions, aiding in better diagnosis.
Area of Science:
- Medical informatics
- Artificial intelligence in healthcare
- Dermatology
Background:
- Lyme disease is a common vector-borne illness with diverse symptoms, often leading to misdiagnosis due to similarities with other conditions like drug rash.
- Accurate diagnosis is crucial to prevent unnecessary antimicrobial treatments and manage potential serious health consequences.
- Current deep learning models for rash diagnosis face challenges such as overfitting and color variations.
Purpose of the Study:
- To develop an efficient stacked deep transfer learning model for accurate Lyme disease diagnosis.
- To overcome limitations of existing deep learning models, specifically overfitting and color variations in medical images.
- To differentiate between patients with Lyme disease and those with other infections based on dermatologic presentations.
Main Methods:
- Utilized 2nd order edge-based color constancy for image preprocessing to standardize lighting conditions.
- Employed the AlexNet pre-trained model as the foundation for the Lyme disease diagnostic model.
- Implemented data augmentation and 5-fold cross-validation to enhance model robustness and prevent overfitting.
Main Results:
- The proposed stacked deep transfer learning model demonstrated superior performance compared to existing models.
- Achieved higher accuracy, f-measure, sensitivity, specificity, and area under the curve in distinguishing Lyme disease.
- Effectively addressed challenges of overfitting and color variations in diagnostic imaging.
Conclusions:
- The developed deep learning model offers an efficient and accurate method for Lyme disease diagnosis from dermatologic images.
- This approach can aid clinicians in differentiating Lyme disease from other conditions, improving patient management.
- The model's performance suggests a promising advancement in AI-driven medical diagnostics for vector-borne diseases.
