Mieke Bus

94 Chapter 6 Table 3: Results OCT vs ELUS Results of OCT vs ELUS in the datasets. OCT ELUS Urothelium 80% (27/34) 0% (0/21) Lamina propria 80% (27/34) 81% (17/21) Muscularis propria 82% (28/34) 81% (17/21) Peri ureteral fat 50% (17/34) 81% (17/21) Maximal imaging depth in lesions 2.0mm 3.5mm* *Maximal imaging depth in lesions of ELUS is approximately 3.5mm due to field-of-view settings. The maximal imaging depth of the ELUS imaging probe used in this study is 20mm. In the OCT imaging datasets the urothelial layer is seen as a low scattering layer and can be differentiated from the lamina propria, which is seen as a high scattering layer. The muscula- ris layer is seen as a low scattering layer. In the first case (figure 4) von Brunn’s nest could be seen in the lamina propria as a low scattering spots within the high scattering lamina propria layer. In ELUS images the muscularis layer was identified as an hypoechoic layer surrounded by the hyperechoic periureteric fat as demonstrated in the first case (figure 4). In both OCT images and ELUS images of suspected areas, lesions could be identified. The OCT imaging depth range is 2mm where the ELUS maximum imaging depth in lesions is 3.5 mm. This maximum imaging depth of ELUS images was limited by the field of view standard and not by loss of imaging signal in the lesions. In specimens 2 and 3 tumor growth exceeded the OCT 2mm imaging depth range and assessment of tumor invasive- ness was hampered (figure 4). In the OCT image of case 2 a papillary structure is recognized. However, the lesion transcends the OCT imaging depth making accurate staging unreliable. Histology diagnosed a TaG2 urothelial carcinoma. In the OCT images of case 3, the tumor is seen as a high scattering lesion that transcends the imaging depth. For this reason, no relia- ble staging can be assessed, although the image is suspicious for invasion. In the fourth case, high resolution OCT images depicted clearly an invasive tumor. Invasion can be recognized as loss of architecture of the underlying layers. In the fifth case, invasive carcinoma was seen as a lesion with loss of architecture of the underlying layers. In this case, invasive carcinoma with a circular growth pattern was found in almost the complete resected specimen. For this reason, only in two of the eight OCT datasets anatomical layers were recognized. In ELUS imaging in case 2, 3, 4, and 5 tumors are recognized as papillary (case 2) or solid lesions (case 3, 4 and 5). Architecture loss is seen in invasive tumors (case 3, 4 and 5) but exact staging and differentiation between non-invasive (Ta) and invasive (≥T1) lesions was not possible due to low-resolution images (figure 4). In the OCT dataset of the pyelum, a nephrostomy tube was recognized.