Hyo Eun Moon and Sun Ha Paek. Exp Neurobiol 2015;24:103-16. https://doi.org/10.5607/en.2015.24.2.103

Hye Ran Park, Jae Meen Lee, Hyo Eun Moon, Dong Soo Lee, Bung-Nyun Kim, Jinhyun Kim, Dong Gyu Kim and Sun Ha Paek. Exp Neurobiol 2016;25:1-13. https://doi.org/10.5607/en.2016.25.1.1

Dopaminergic degeneration is a pathologic hallmark of Parkinson's disease (PD), which can be assessed by dopamine transporter imaging such as FP-CIT SPECT. Until now, imaging has been routinely interpreted by human though it can show interobserver variability and result in inconsistent diagnosis. In this study, we developed a deep learning-based FP-CIT SPECT interpretation system to refine the imaging diagnosis of Parkinson's disease. This system trained by SPECT images of PD patients and normal controls shows high classification accuracy comparable with the experts' evaluation referring quantification results. Its high accuracy was validated in an independent cohort composed of patients with PD and nonparkinsonian tremor. In addition, we showed that some patients clinically diagnosed as PD who have scans without evidence of dopaminergic deficit (SWEDD), an atypical subgroup of PD, could be reclassified by our automated system. Our results suggested that the deep learning-based model could accurately interpret FP-CIT SPECT and overcome variability of human evaluation. It could help imaging diagnosis of patients with uncertain Parkinsonism and provide objective patient group classification, particularly for SWEDD, in further clinical studies.

Despite aggressive multimodal treatment, glioblastoma (GBM), a grade IV primary brain tumor, still portends a poor prognosis with a median overall survival of 12–16 months. The complexity of GBM treatment mainly lies in the inter- and intra-tumoral heterogeneity, which largely contributes to the treatment-refractory and recurrent nature of GBM. By paving the road towards the development of personalized medicine for GBM patients, the cancer genome atlas classification scheme of GBM into distinct transcriptional subtypes has been considered an invaluable approach to overcoming this heterogeneity. Among the identified transcriptional subtypes, the mesenchymal subtype has been found associated with more aggressive, invasive, angiogenic, hypoxic, necrotic, inflammatory, and multitherapy-resistant features than other transcriptional subtypes. Accordingly, mesenchymal GBM patients were found to exhibit worse prognosis than other subtypes when patients with high transcriptional heterogeneity were excluded. Furthermore, identification of the master mesenchymal regulators and their downstream signaling pathways has not only increased our understanding of the complex regulatory transcriptional networks of mesenchymal GBM, but also has generated a list of potent inhibitors for clinical trials. Importantly, the mesenchymal transition of GBM has been found to be tightly associated with treatment-induced phenotypic changes in recurrence. Together, these findings indicate that elucidating the governing and plastic transcriptomic natures of mesenchymal GBM is critical in order to develop novel and selective therapeutic strategies that can improve both patient care and clinical outcomes. Thus, the focus of our review will be on the recent advances in the understanding of the transcriptome of mesenchymal GBM and discuss microenvironmental, metabolic, and treatment-related factors as critical components through which the mesenchymal signature may be acquired. We also take into consideration the transcriptomic plasticity of GBM to discuss the future perspectives in employing selective therapeutic strategies against mesenchymal GBM.