Search

Aim In 2015, the revised International Pediatric Non-Hodgkin Lymphoma Staging System was published. It mentions [18F]-FDG-PET/MRI as the latest method to perform whole-body imaging. However, supporting data are pending. Our aim was to investigate the performance of whole-body [18F]-FDG-PET/MRI in pediatric non-Hodgkin lymphoma patients by using a limited number of MRI sequences. Materials and methods Ten pediatric patients with histologically proven non-Hodgkin lymphoma underwent whole-body [18F]-FDG-PET/MRI at staging. The retrospective analysis included three steps: First, [18F]-FDG-PET and MR scans were evaluated separately by a nuclear medicine physician and a pediatric radiologist. Nineteen nodal and two extranodal regions as well as six organs were checked for involvement. Second, discrepant findings were reviewed together in order to reach consensus. Third, [18F]-FDG-PET/MRI findings were correlated with the results of other clinical investigations. Results Of the 190 lymph node regions evaluated, four were rated controversial. Consensus was reached by considering metabolic, functional and morphologic information combined. Concordantly, [18F]-FDG-PET and MRI detected Waldeyer’s ring involvement in two patients whose Waldeyer’s ring was negative on clinical assessment. In four patients MRI showed pleural effusion. However, in only two of them an increased glucose metabolism as a reliable sign of pleural involvement was detectable. In six patients [18F]-FDG-PET and MRI detected skeletal lesions although bone marrow biopsy was positive in only one of them. Conclusion Despite the small number of cases evaluated, whole-body [18F]-FDG-PET turned out to be a valuable tool for staging of pediatric non-Hodgkin lymphoma.

There are several different algorithms published for the identification of differentially expressed genes in DNA microarray experiments. Such algorithms produce ordered lists of genes. To compare the performance of these algorithms established measurements from Information Retrieval are proposed. A benchmark data set with known properties is generated and published. This benchmark data is used to compare the performance of different algorithms with a new algorithm, called PUL. Surprisingly a clear ordering in performance of the algorithms was observed. PUL outperformed other algorithms by a factor of two. PUL was applied successfully in different practical applications. For these experiments the importance of the genes identified by PUL were independently verified.