<![CDATA[Pulmonary nodules are lumps in the lungs measuring approximately 3 cm that are used as a marker or early symptom for certain lung diseases. This study aims to analyze and compare the accuracy of the Pseudo Nearest Neighbor Rule (PNNR) algorithm using the Euclidean Distance and Manhattan Distance functions in detecting pulmonary nodules in CT scan images. The PNNR classification algorithm is used to reduce the influence of noise or outliers in the classification process so that false positives (predictions of non-nodule objects as nodules) can be reduced. The dataset used is 200 patient data which are then used as training data and test data. In this study, the results of the PNNR classification using the Euclidean Distance method succeeded in obtaining fewer false positives compared to the Manhattan method, namely 284 (5.68 FP/C) while in Manhattan it was 318 (6.36 FP/C). However, the PNNR algorithm using the Euclidean method obtained fewer true positives compared to the Manhattan method, with 32 true positives, while the Manhattan method had 33 true positives. This indicates that the PNNR algorithm using the Euclidean method is good at overcoming false positives but with a lower level of sensitivity or recognition of true positives compared to the Manhattan distance. In further research, classification optimization can be carried out by selecting the feature set used.]]>
Copyrights © 2026
