I’ve never worked with SPECT data so it’s not clear I can help, but could you elaborate on what you mean by “calculate the lesion on the CT image”? It might be that you can create a segmentation with thresholding on the SPECT and then run segment statistics with the CT as the target volume.