Artificial intelligence (AI) transforms intestinal parasite diagnosis, particularly through deep learning models like convolutional neural networks (CNNs). This paper reviews the application of AI, especially CNNs, in automating parasite detection and classification from microscopic images. Integrating AI into parasitology diagnostics speeds up the process, reduces human error, and enhances treatment and patient outcomes. However, there is a need for more datasets reflecting the African context to ensure accurate ground-truthing, particularly in low- and middle-income countries (LMICs) across Africa. Most AI models for medical diagnosis are trained on datasets from high-income countries, which may not capture the unique epidemiological, genetic, and environmental factors prevalent in African populations. This can lead to less accurate diagnoses and treatment recommendations in African LMICs. For example, intestinal parasitic infections are common in many African regions, yet the datasets used to train AI models often lack sufficient representation from these areas. Developing datasets that reflect the diverse African context is crucial for improving the accuracy and reliability of AI-based diagnostic tools. Issues like overfitting, data privacy, and cost also require attention. Collaboration between researchers, healthcare professionals, and technologists is essential to address these challenges. Standardized protocols for data collection, model training, and validation are necessary for reliable AI systems. Combining AI with traditional techniques holds promise for better parasite diagnosis, ultimately improving African healthcare outcomes
BT - East African Journal of Information Technology DO - 10.37284/eajit.7.1.2282 IS - 1 LA - ENG M3 - Article N2 -Artificial intelligence (AI) transforms intestinal parasite diagnosis, particularly through deep learning models like convolutional neural networks (CNNs). This paper reviews the application of AI, especially CNNs, in automating parasite detection and classification from microscopic images. Integrating AI into parasitology diagnostics speeds up the process, reduces human error, and enhances treatment and patient outcomes. However, there is a need for more datasets reflecting the African context to ensure accurate ground-truthing, particularly in low- and middle-income countries (LMICs) across Africa. Most AI models for medical diagnosis are trained on datasets from high-income countries, which may not capture the unique epidemiological, genetic, and environmental factors prevalent in African populations. This can lead to less accurate diagnoses and treatment recommendations in African LMICs. For example, intestinal parasitic infections are common in many African regions, yet the datasets used to train AI models often lack sufficient representation from these areas. Developing datasets that reflect the diverse African context is crucial for improving the accuracy and reliability of AI-based diagnostic tools. Issues like overfitting, data privacy, and cost also require attention. Collaboration between researchers, healthcare professionals, and technologists is essential to address these challenges. Standardized protocols for data collection, model training, and validation are necessary for reliable AI systems. Combining AI with traditional techniques holds promise for better parasite diagnosis, ultimately improving African healthcare outcomes
PB - East African Nature and Science Organization PY - 2024 SP - 366 EP - 379 T2 - East African Journal of Information Technology TI - AI Methods and Algorithms for Diagnosis of Intestinal Parasites: Applications, Challenges and Future Opportunities UR - https://journals.eanso.org/index.php/eajit/article/view/2282/2954 VL - 7 SN - 2707-5354, 2707-5346 ER -