Radiomics-Based Machine Learning to Predict Recurrence in Glioma Patients Using Magnetic Resonance Imaging.

OBJECTIVE: Recurrence is a major factor in the poor prognosis of patients with glioma. The aim of this study was to predict glioma recurrence using machine learning based on radiomic features. METHODS: We recruited 77 glioma patients, consisting of 57 newly diagnosed patients and 20 patients with recurrence. After extracting the radiomic features from T2-weighted images, the data set was randomly divided into training (58 patients) and testing (19 patients) cohorts. An automated machine learning method (the Tree-based Pipeline Optimization Tool) was applied to generate 10 independent recurrence prediction models. The final model was determined based on the area under the curve (AUC) and average specificity. Moreover, an independent validation set of 20 patients with glioma was used to verify the model performance. RESULTS: Recurrence in glioma patients was successfully predicting by machine learning using radiomic features. Among the 10 recurrence prediction models, the best model achieved an accuracy of 0.81, an AUC value of 0.85, and a specificity of 0.69 in the testing cohort, but an accuracy of 0.75 and an AUC value of 0.87 in the independent validation set. CONCLUSIONS: Our algorithm that is generated by machine learning exhibits promising power and may predict recurrence noninvasively, thereby offering potential value for the early development of interventions to delay or prevent recurrence in glioma patients.

Does epilepsy always indicate worse outcomes? A longitudinal follow-up analysis of 485 glioma patients.

BACKGROUND: Epilepsy is one of the most common glioma complications, and the two may be connected in more ways than we understand. We aimed to investigate the clinical features of glioma-associated epilepsy and explore the risk factors associated with it. METHODS: We collected clinical information from 485 glioma patients in the Nanjing Brain Hospital and conducted 4 periodic follow-up visits. Based on the collected data, we analyzed the clinical characteristics of glioma patients with or without epilepsy and their relationship with survival. RESULTS: Among glioma patients, younger people were more likely to have epilepsy. However, epilepsy incidence was independent of gender. Patients with grade II gliomas were most likely to develop epilepsy, while those with grade IV gliomas were least likely. There was no difference in Karnofsky Performance Status scores between patients with glioma-associated epilepsy and those without epilepsy. Additionally, epilepsy was independently associated with longer survival in the World Health Organization grade IV glioma patients. For grades II, III, and IV tumors, the 1-year survival rate of the epilepsy group was higher than that of the non-epilepsy group. CONCLUSIONS: Epilepsy did not lead to worse admission performance and correlated with a better prognosis for patients with grade IV glioma.

Development of a loop-mediated isothermal amplification assay for the rapid detection of Russula subnigricans and Russula japonica.

Russula subnigricans is the only deadly species in the genus Russula with a mortality rate of more than 50%, and Russula japonica is the most common poisonous species, making rapid species identification in mushroom poisoning incidents extremely important. The main objective of this study was to develop a rapid, specific, sensitive, and simple loop-mediated isothermal amplification (LAMP) assay for the detection of R. subnigricans and R. japonica. Two sets of species-specific LAMP primers targeting internal transcribed spacer (ITS) regions were designed to identify R. subnigricans and R. japonica. The results demonstrated that while LAMP could specifically detect R. subnigricans and R. japonica, the polymerase chain reaction (PCR) could not distinguish R. subnigricans from Russula nigricans. In addition, the results demonstrated that, compared to electrophoresis-LAMP and real-time quantitative LAMP (RT-qLAMP), the detection sensitivity of HNB-LAMP (a mixture of LAMP with hydroxy naphthol blue (HNB) dye) for R. subnigricans could reach 0.5 pg/µl and was 100-fold higher than that of PCR. The LAMP reaction could be completed in 45 min, which is much faster than the conventional PCR. In the future, LAMP can be used a quick, specific, and sensitive detection tool in various fields.

Effectiveness of Home Care Interventions in Patients with Moderate to Severe Craniocerebral Injury Combined with Epilepsy.

Objective: To investigate the effect of a home care model on coping style and quality of life of patients with postcranial surgery complicated by epilepsy. Methods: One hundred and forty-four patients with postcranial surgery complicated by epilepsy admitted to our hospital from July 2017 to April 2018 were selected as study subjects and randomly divided into 63 cases each in the observation group and the control group. The control group was given nursing interventions including health education, psychological intervention, medication guidance, complication prevention, and follow-up management, while the observation group was jointly given collaborative family nursing interventions. At a follow-up of 6 months, indicators such as coping style, treatment compliance, and quality of life were compared between the two groups. Results: Patients in the observation group had significantly higher problem solving, help seeking, and rationalization scores and significantly lower self-blame scores than the control group (P < 0.01); significantly higher treatment compliance than the control group (P < 0.01); and significantly higher social functioning, emotional well-being, and energy/fatigue scores than the control group (P < 0.01). Conclusion: The home care model helps to promote the development of positive coping styles, improve treatment compliance, and improve the quality of life of patients after cranial surgery.

Altered Static and Dynamic Voxel-mirrored Homotopic Connectivity in Patients with Frontal Glioma.

Contralateral regions play critical role in functional compensation in glioma patients. Voxel-mirrored homotopic connectivity (VMHC) characterizes the intrinsic functional connectivity (FC) of the brain, considered to have a regional functional basis. We aimed to investigate the alterations of brain regional function and VMHC in patients with frontal glioma, and further investigated the correlation between these alterations and cognition. We enrolled patients with frontal glioma and matched healthy controls (HC). We chose degree centrality (DC), regional homogeneity (ReHo), and VMHC to investigate the alterations of regional function and intrinsic FC in patients. Furthermore, partial correlation analyses were conducted to explore the relationship between imaging functional indicators and cognitions. Compared with HC, patients showed decreased static VMHC within right and left middle frontal gyrus (MFG.R, MFG.L), left superior frontal gyrus (SFG.L), right precuneus (PCUN.R), and left precuneus (PCUN.L), decreased static DC within left cingulate gyrus (CG.L), right superior frontal gyrus (SFG.R), and right postcentral gyrus (POCG.R), decreased static ReHo within CG.L, decreased dynamic ReHo within right inferior parietal lobule (IPL.R), but increased dynamic VMHC (dVMHC) within PCUN.R and PCUN.L. Furthermore, values of decreased VMHC within MFG.R, decreased DC within CG.L, decreased ReHo within CG.L, and increased dVMHC within PCUN.R were significantly positively correlated with cognitive functions. We preliminarily confirmed glioma causes regional dysfunction and disturbs long-distance FC, and long-distance FC showed strong instability in patients with frontal glioma. Meanwhile, the correlation analyses indicated directions for cognitive protection in patients with frontal glioma.

Sleep quality of patients with postoperative glioma at home.

BACKGROUND: The sleep quality of patients with postoperative glioma at home has not been reported worldwide. In this study, the sleep quality of postoperative glioma patients at home was explored through telephone interviews. AIM: To investigate the sleep quality of patients with postoperative glioma. METHODS: Glioma patients who were treated and operated on at Nanjing Brain Hospital from May 2012 to August 2019 were selected as participants via convenience sampling. A total of 549 patients were recruited by telephone, and 358 of those patients responded to the Pittsburgh Sleep Quality Index (PSQI). RESULTS: The total PSQI score was 5.19 ± 3.39 in all patients with postoperative glioma, which was higher than that of the normal group. There were 135 patients with poor sleep quality (PSQI > 7), which accounted for 37.7% of all respondents. The scores of the seven dimensions of the PSQI were sleep latency (1.22 ± 1.05), sleep disturbance (1.12 ± 0.98), subjective sleep quality (1.03 ± 0.92), habitual sleep efficiency (1.02 ± 0.91), daytime dysfunction (0.95 ± 0.43), sleep persistence (0.83 ± 0.76), and use of sleeping medication (0.25 ± 0.24). The results showed that the subjective sleep quality of the patients who underwent an operation 1-2 years ago was worse than that of the patients who underwent an operation more than 2 years ago (P < 0.05). Older patients had worse sleep latency and sleep duration scores than younger patients (P < 0.05). Sleep disturbance and daytime dysfunction were worse in patients who received postoperative radiotherapy/ chemotherapy (P < 0.05). CONCLUSION: The sleep quality of patients with postoperative glioma at home is worse than that of normal adults. Moreover, difficulty falling asleep and sleep disorders are common complications among these patients, and age, postoperative duration, and postoperative chemoradiotherapy could affect sleep quality.

Structural and Functional Reorganization Within Cognitive Control Network Associated With Protection of Executive Function in Patients With Unilateral Frontal Gliomas.

Background: The cognitive control network (CCN) is widely considered to be a frontoparietal circuit that is involved in executive function. This study aimed to investigate the structural and functional plasticity within the CCN in unilateral frontal gliomas, which are associated with the protection of executive functions. Methods: To detect structural and functional changes within the CCN, we measured gray matter (GM) volume, regional homogeneity, the amplitude of low-frequency fluctuation (ALFF), degree centrality, and functional connectivity within the CCN in 37 patients with gliomas invading the left frontal lobe (n = 16) or the right frontal lobe (n = 21) and 40 healthy controls (CNs). Partial correlation analysis was performed to assess the association between the altered structural and functional indices and executive function. Results: When the tumor invaded the left frontal lobe, the patients showed reduced ALFF in the dorsal medial prefrontal cortex (dmPFC) within the CCN and increased ALFF in the right superior parietal cortex (rSP) within the CCN compared to the CNs. When the tumor invaded the right frontal lobe, the patients showed significantly increased GM volume and ALFF in the left superior parietal cortex (lSP) within the CCN compared to the CNs. Furthermore, the patients showed significantly increased functional connectivities between the lSP and the dmPFC and between the lSP and the rSP within the CCN compared to the CNs. Increased ALFF in the lSP within the CCN was positively correlated with executive function. Conclusions: Tumors invading the frontal lobe induced contralesional structural and functional reorganization within the posterior CCN in patients with unilateral frontal gliomas. This further suggests that the contralesional superior parietal cortex acts as a functional compensation hub within the CCN, which may protect it against the detrimental effects of tumor invasion on executive functions.