Fetal Congenital Heart Disease Detection Using Firefly Segmentation and Decision Tree Classification Technique

Identifying congenital heart disease (CHD) during the fetal stage is challenging, making it difficult to detect in its early phases. CHD can be slightly visualized in ultrasound scan during pregnancy which will make some serious effects to infants after child birth. CHD detection using machine learning algorithm is proposed in this work which will identify birth defects in early stages and thus the medical experts can take necessary decision in right time. Firefly algorithm for image segmentation and decision tree classifier for CHD detection is used in this work. After conducting the validation and testing of the proposed methodology, several measures are available to evaluate the effectiveness of the classifier, such as the precision of the input data, precision of given value, recall of data, and F1 score value. These measures are important in assessing the effectiveness of the classifier's ability to make accurate classifications. These metrics offer insightful data about how effectively the classifier perform in terms of accurately classifying instances and handling various error types. An F1 score of 87.88%, accuracy of 88%, precision of 88.78%, and recall of 87% are obtained using the suggested approach. Collectively, these measures evaluate how well the classifier performed in correctly identifying the data. When compared to the existing methodology, the proposed method yields improved classification results.