Are Deep Learning Classification Results Obtained on CT Scans Fair and Interpretable?