Timing calibration comparison research of integrated TOF-PET/MR / 生物医学工程学杂志

Integrated TOF-PET/MR is a multimodal imaging system which can acquire high-quality magnetic resonance (MR) and positron emission tomography (PET) images at the same time, and it has time of flight (TOF) function. The TOF-PET system usually features better image quality compared to traditional PET because it is capable of localizing the lesion on the line of response where annihilation takes place. TOF technology measures the time difference between the detectors on which the two 180-degrees-seperated photons generated from positron annihilation are received. Since every individual crystal might be prone to its timing bias, timing calibration is needed for a TOF-PET system to work properly. Three approaches of timing calibration are introduced in this article. The first one named as fan-beam method is an iterative method that measures the bias of the Gaussian distribution of timing offset created from a fan-beam area constructed using geometric techniques. The second one is to find solutions of the overdetermination equations set using L1 norm minimization and is called L1-norm method. The last one called L2-norm method is to build histogram of the TOF and find the peak, and uses L2 norm minimization to get the result. This article focuses on the comparison of the amount of the data and the calculation time needed by each of the three methods. To avoid location error of the cylinder radioactive source during data collection, we developed a location calibration algorithm which could calculate accurate position of the source and reduce image artifacts. The experiment results indicate that the three approaches introduced in this article could enhance the qualities of PET images and standardized uptake values of cancer regions, so the timing calibration of integrated TOF-PET/MR system was realized. The fan-beam method has the best image quality, especially in small lesions. In integrated TOF-PET/MR timing calibration, we recommend using fan-beam method.

Clonality of multiple uterine leiomyomas / 中华病理学杂志

<p><b>OBJECTIVE</b>To study the clonality of uterine leiomyomas, especially the relationship between different nodules in multinodular cases.</p><p><b>METHODS</b>Genomic DNA was extracted from fresh tissue samples and digested through incubation with Hpa II and amplified through nested polymerase chain reaction for phosphoglycerate kinase (PGK) gene. The products were treated with Bst XI and resolved on agarose gels.</p><p><b>RESULTS</b>Among the 103 cases examined, 32 (31%) carried the polymorphic Bst XI site at the PGK locus. Eighty-nine tumors from the 29 cases were subjected to the cloning assay. Loss of polymorphism at the PGK locus was found in all tumor nodules, indicating the monoclonality of the tumor. The relationship between multiple tumors was also assessed by comparing their inactivated alleles. Seven nodules from a leiomyosarcoma were found to have originated from a single cell. However, the relationship was found to be more complicated, as demonstrated in 15 cases of multiple leiomyomas. The same inactivated allele was found in all nodules of 8 cases and in most nodules in 2 cases, while totally different inactivation patterns were observed in 5 cases. The difference was not associated with cell proliferation.</p><p><b>CONCLUSIONS</b>Clonality analysis can be applied to define the clonality of focal or nodular lesions. Uterine leiomyomas are of clonal origin. Multiple uterine leiomyomas may be subtyped into fully independent and aggressive types as well as a mixed type of both.</p>