Kidney Tumor Detection in Medical Image Processing

Abstract

Accurate classification of medical images, particularly
CT scans, plays a vital role in diagnosing and detecting
renal anomalies, including tumors, stones, and cysts. This work
investigates several cutting-edge deep learning techniques for
the automated identification and categorization of certain renal
conditions. Large datasets comprising thousands of CT images
were used to evaluate different methods according to important
performance metrics such as recall, accuracy, precision, and
F1-score. The proposed approaches demonstrated high classification
accuracy, with several models achieving over 99% in
sensitivity and specificity. Additionally, The study emphasizes the
significance of incorporating explainability techniques to improve
model interpretability, Facilitating clinicians’ comprehension of
the decision-making process. The findings emphasize the potential
of these advanced techniques in enhancing medical diagnostics,
particularly in environments where computational efficiency and
accuracy are critical.