Hasil Pencarian  ::  Simpan CSV :: Kembali

"Penelitian ini mengembangkan Uji Korelasi Computer Aided Diagnosis CAD radiografi paru anak dengan menggunakan metode segmentasi Markov Random Field MRF untuk membantu mendeteksi abnormalitas paru dengan kecenderungan infeksi. Metode MRF mencari abnormalitas berdasarkan nilai piksel citra. Metode MRF dikerjakan dengan empat variasi yaitu MRF tanpa filter, median filter MRF, wiener filter MRF dan adapthisteq MRF. ROC hasil segmentasi wiener filter relatif lebih tinggi dari tanpa filter. Hasil ROC wiener filter menunjukkan nilai akurasi akurasi akurasi 81,4 , sensitivitas 82,0 , spesifitas 80,0 , presisi 91,1 dan overall error 18,6 . Sedangkan ROC untuk tanpa filter maupun filter yang lain menunjukkan lebih rendah dari nilai ROC wiener filter. Namun perbedaan ROC untuk setiap jenis tingkat keberhasilan tidak lebih dari 5 , artinya keempat metode MRF masih dapat diimplementasikan. Nilai piksel paru abnormal dengan metode MRF tanpa filter, median filter MRF, dan adapthisteq MRF sama yaitu 205-255. Nilai piksel paru abnormal dengan metode wiener filter MRF yaitu 197-255. Citra paru belum dapat menentukan secara definitif penyakit infeksi paru pada anak.

---

This study developed a correlation test Computer Aided Diagnosis CAD radiographic of children pulmonary using segmentation Markov Random Field MRF method to detect lung abnormalities with infection trends. MRF method searched abnormalities by value of the image pixel. MRF method used four variations, namely MRF without a filter, median filter MRF, wiener filter MRF, and adapthisteq MRF. ROC segmentation results wiener filter is relatively higher than without a filter. ROC wiener filter results show the value of accuracy 81.4 , sensitivity 82.0 , specificity 80.0 , precision 91.1 and overall error of 18.6 . While the ROC for unfiltered and filter others show lower than the value of ROC wiener filter. However, differences in ROC for any kind of success rate is not more than 5 , meaning that all four methods MRF can still be implemented. Abnormal lung pixel value with MRF method without filter, median filter MRF, and adapthisteq MRF same namely 205 255. Abnormal lung pixel values by the method of wiener filter MRF is 197 255. Radiographic of children pulmonary can not definitively determine lung infections in children."

"This book provides cutting-edge insights into autonomous vehicles and road terrain classification, and introduces a more rational and practical method for identifying road terrain. It presents the MRF algorithm, which combines the various sensors classification results to improve the forward LRF for predicting upcoming road terrain types. The comparison between the predicting LRF and its corresponding MRF show that the MRF multiple-sensor fusion method is extremely robust and effective in terms of classifying road terrain. The book also demonstrates numerous applications of road terrain classification for various environments and types of autonomous vehicle, and includes abundant illustrations and models to make the comparison tables and figures more accessible. "