A new method for discriminating cerebellar tumors from each other: algorithmic approach

Abstract

Objective: To evaluate the MR imaging findings of the cerebellar tumors (CT) in adults and to make an algorithm to discriminate these tumors from each other.Materials and Methods: Adult patients with CT who were treated surgically were included. Brain MRs and patient records were evaluated for pathological diagnosis, imaging findings and Ki-67 labeling index. The apparent diffusion coefficient (ADC) values of the tumors and centrum semiovale (CS) and the ratio of each other were noted. Kruskal-Wallis test was used for comparison.Results: 100 adults (45 women, 55 men with a mean age of 46.6±15.6 years) with pathologically proven CT were included. Tumors were divided into five groups as: a. primitive neuro-ectodermal tumor (PNET) + medulloblastoma (MB), b. grade 3+4 gliomas, c. grade 1+2 gliomas, d. metastasis (MET) and e. lymphoma. The ratio of the mean ADC (x10-3 mm2/s) of the tumor to those of CS was 1.21±0.22, 1.38±0.93, 1.98±0.69, 1.48±0.45 and 1.25±0.11 for PNET amp;MBs, grade 3+4 gliomas, grade 1+2 gliomas, METs and lymphomas, respectively. Significant difference was detected between grade 1+2 gliomas and PNET amp;MB (P=0.018), MET (P=0.004) and lymphoma (P=0.002). Conclusion: CT in adults might be differentiated from each other by using some MR imaging clues and patient characteristics.

Keywords

• apparent diffusion coefficient
• cerebellum
• cerebellar tumors
• görünür difüzyon katsayısı
• serebellum
• serebellar tümörler