Human papilomaviru-related oropharyngeal squamous cell carcinoma and radiomics: A new era?

BACKGROUND: The increase of the incidence of human papillomavirus dependent oropharyngeal squamous cell carcinoma is alarming, although we have greatly progressed in the classification and staging of this disease. We now know that human papillomavirus related oropharyngeal squamous cell carcinoma is a sub-type of head and neck squamous cell carcinoma with favourable prognosis and good response to therapy that needs a proper system of classification and staging. Thus, in routine practice it is essential to test patients for the presence of human papillomavirus. The most popular technique to assess human papillomavirus status is immunohistochemistry on biopsy samples with p16, which is an excellent surrogate for high-risk human papillomavirus infection. Another highly sensitive and specific tissue-based technique for the detection of human papillomavirus is RNAscope In situ hybridization that has a prohibitive cost, limiting its use in routine practice. Radiomics is an artificial intelligence based non-invasive method of computational analysis of computed tomography, magnetic resonance imaging, positron emission tomography, and ultrasound images. METHODS: In this review, we summarise the last findings of radiomics applied to human papillomavirus associated oropharyngeal squamous cell carcinoma. RESULTS: A growing body of evidence suggest that radiomics is able to characterise and detect early relapse after treatment, and enable development of tailored therapy of human papillomavirus positive oropharyngeal squamous cell carcinoma.

A three-dimensional morphological reconstruction of uterine leiomyoma pseudocapsule vasculature by the Allen-Cahn mathematical model.

OBJECTIVE: To allow a morphological evaluation of pseudocapsule vasculature of uterine leiomyomas (ULs). DESIGN: Preliminary prospective study. SETTING: University-affiliated hospitals. PATIENTS: Ten women with symptomatic myomas warranting laparotomic hysterectomies. METHODS: Pseudocapsules of ULs were isolated from the surrounding myometrium, for a digital reconstruction of pseudocapsule vessels and analysis by three-dimensional structure applying the Allen-Cahn mathematical model. The images were converted in a matrix and an appropriate created software elaborated the images. The "nonlocal" Allen-Cahn mathematical model was applied to reduce the image noise and to regulate vessel outlines obtaining a "clean" image. MAIN OUTCOME MEASURES: The geometrical characteristics and morphology of vessel pseudocapsule network of ULs. RESULTS: The disarray in vascular architecture was evident from the absence of vessel parallelism and variable intervascular distances. An abnormal vascular branching of pseudocapsule was indicated by the different density of vessels per space and some vascular walls without interruption indicated vessel tortuosity. There were vascular spaces, which did not communicate with other vessels ("cul-de-sac" vessels). CONCLUSIONS: Pseudocapsule vasculature showed increase of tortuosity, disarray, abnormal branching and the presence of "cul-de-sac" vessels. Three-dimensional reconstruction of leiomyoma pseudocapsule vasculature network, despite the benign nature of ULs, showed geometrical characteristics of malignant neoplasm vessels.