Klasifikasi Penyakit Ginjal Berdasarkan Gambar CT Scan Menggunakan Deep Learning

Abstract

Kidney is a vital organ in the human excretory system that has an important role in filtering blood and removing waste. Kidney is a serious problem for humans if not treated or treated in the wrong way, misinterpretation of the disease will also have a fatal impact on the patient, therefore, a system is needed that can help health professionals who can classify the right type of kidney condition. This research implements You Only Look Once version 9 (YOLOv9) to classify 4 types of kidney conditions namely kidney stones, cysts, normal, and tumors with the amount of data used as much as 3600 data consisting of 2880 data as training data, 360 data as validation data, and 360 data as testing data. The results show that the YOLOv9 algorithm is less able to classify kidney conditions with an accuracy value of 89.90%, precision 88.67%, recall 86.30%, and f1-score 87.42%. This model is able to classify the type of kidney condition simultaneously with as many as 40 images. These results show that the system created using the YOLOv9 algorithm has successfully classified the types of conditions in the kidneys.