Imaging of laryngeal tumors from macro to micro

Abstract

This thesis takes a closer look at the radiological and histopathological characteristics of laryngeal and hypopharyngeal tumors. In the first part of this thesis, we looked at the prognostic value of imaging biomarkers on the treatment outcome after radiotherapy. This was done in a review of the literature in Chapter 2 and in a patient cohort in Chapter 3. We found that tumor volume, as well as little tumor perfusion, were independent prognostic factors of worse treatment outcome. The second part of this thesis dove into computational pathology methods. In Chapter 4, we validated a method that detected positively stained cells in tumor tissue. In Chapter 5, the heterogeneity of pathological biomarkers was quantified using the Haralick features. Even though it is well-known that heterogeneous tumors are more aggressive, tumor heterogeneity is rarely considered when making treatment decisions. This study shoes a method of quantifying the heterogeneity of pathological biomarkers. In the third part of this thesis, radiology and pathology were combined. We made digital reconstructions of surgically removed tumors and registered them to in vivo imaging. This enabled us to compare pathological and radiological imaging. In Chapter 6, we did this on a voxel-by-voxel level. In Chapter 7, tumor delineations made for radiotherapeutic purposes were compared to the pathological tumor volume. This showed that diffusion-weighted MRI led to better tumor delineations on in vivo imaging, making it an excellent addition to the radiotherapy treatment planning in head and neck patients. Pathology and radiology are two similar disciplines with complementary information. Pathology can see a lot of a little, whereas radiology sees a little of a lot. A more intensive collaboration between the two can help us better understand the complex tumor environment and benefit both clinical and research efforts.