Jeremiah W Sanders¹, Jason M Johnson², Henry Szu-Meng Chen¹, Melissa Chen², Maria Gule-Monroe², Zijian Zhou¹, Tina M Briere³, Jing Li⁴, Jingfei Ma¹, and Ho-Ling Liu^1
1Imaging Physics, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ²Neuroradiology, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ³Radiation Physics, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ⁴Radiation Oncology, University of Texas MD Anderson Cancer Center, Houston, TX, United States

Metastasis is regarded as a highly inefficient process in that less than 0.01% of circulating tumor cells eventually succeeds in forming secondary tumor growths. The seed and soil theory suggests that certain tumor cells have a specific affinity for the milieu of certain organs. Numerous studies have attempted to better understand the biology behind the distribution of brain metastasis and theories include both variables involving the metastasis and the target brain parenchyma. This work establishes probabilistic maps of locations where 13 primary cancer subtypes metastasize to the brain from T1-weighted MRIs of 955 patients undergoing stereotactic radiosurgery.

Thomas Reith¹, Eun-Jung Choi², Jongho Lee², Melissa Prah¹, Robert Wujek^1,3, Mona Al-Gizawiy¹, Christopher Chitambar⁴, and Kathleen Schmainda^1
1Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States, ²Electrical and Computer Engineering, Seoul National University, Seoul, Korea, Republic of, ³Biomedical Engineering, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Hematology & Oncology, Medical College of Wisconsin, Milwaukee, WI, United States

Basal ganglia iron levels were assessed in 55 patients with gliomas by using a deep neural network-powered quantitative susceptibility mapping method (QSMnet+). Basal ganglia QSM values increased with higher tumor grade, suggesting a correlation between BG iron content and glioma severity.

Shanshan Jiang¹, Pengfei Guo², Hye-Young Heo¹, John Laterra³, Charles Eberhart⁴, Michael Lim⁵, Peter C.M. van Zijl^1,6, and Jinyuan Zhou^1
1Radiology, Johns Hopkins University, Baltimore, MD, United States, ²Computer Science, Johns Hopkins University, Baltimore, MD, United States, ³Neurology, Johns Hopkins University, Baltimore, MD, United States, ⁴Pathology, Johns Hopkins University, Baltimore, MD, United States, ⁵Neurosurgery, Johns Hopkins University, Baltimore, MD, United States, ⁶F.M. Kirby Research Center, Kennedy Krieger Institute, Baltimore, MD, United States

Assessment of glioma treatment is based on pathological evaluation via biopsies or radiological criteria using follow-up MRI, which is either invasive or time consuming. Amide protein transfer weighted (APTw) MRI has been validated to accurately detect recurrent malignant gliomas by multiple studies. The cutting edge methodology of radiomics provides quantitative measurements for imaging diagnosis. Here, we develop an automated framework that integrates APTw MRI radiomic features with a machine learning model to evaluate treatment response for gliomas. Our results suggest that the use of APTw features enabled the radiomic model to reach a more accurate assessment of the treatment effect.

Rupsa Bhattacharjee^1,2, Rakesh Kumar Gupta³, Rana Patir⁴, Sandeep Vaishya⁴, Sunita Ahlawat⁴, and Anup Singh^1,5
1Center for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi, India, ²Philips Health Systems, Philips India Limited, Gurugram, India, ³Department of Radiology, Fortis Memorial Research Institute, Gurugram, India, ⁴Department of Neurosurgery, Fortis Memorial research Institute, Gurugram, India, ⁵Department of Biomedical Engineering, All India Institute of Medical Sciences, New Delhi, India

No standalone method is yet reported to differentiate PCNSL and grade-IV glioma with highest accuracy and diagnostic confidence. Imaging based differentiation between these two types has been a challenging problem. Objective of this study is to evaluate performance of texture-based features of SWI in improved differentiation of PCNSL from grade-IV gliomas. Proposed approach based upon texture feature-extraction from segmented SWI lesion, enabled automatic classification of tumors into primary central nervous system lymphoma and grade-IV glioma cases. One of the texture feature, Contrast, provided highest AUC along with high sensitivity and specificity. This classification might improve diagnosis and grading of tumors.  

Rushi Chen¹, Yan Bai², Mathias Nittka³, Gregor Koerzdoerfer³, Xianchang Zhang⁴, and Meiyun Wang^2
1Henan provincial people's hospital& Zhengzhou University People’s Hospital, Zhengzhou, China, ²Henan provincial people's hospital, Zhengzhou, China, ³MR Pre-development, Siemens Healthcare, Erlangen, Germany, ⁴MR Collaboration, Siemens Healthcare Ltd, Beijing, China

Currently, there is no effective pharmacological treatment for non-functioning pituitary adenoma, including gonadotroph adenoma (GPA). Recent findings showed that an alternative pharmacological treatment targeted at somatostatin receptor 3 may be promising, which required accurate diagnosis of GPA before surgery. This study aimed to evaluate the utility of magnetic resonance fingerprinting (MRF) in the pre-surgical differentiation of GPA from non-functioning, non-gonadotropin adenoma (NGPA). The results showed that GPAs have significantly higher T1 and T2 values in the solid tumor than NGPA, suggesting that MRF may have potential for diagnosing GPA and benefit its treatment plan.

Reneira Seeamber¹, Katherine L Ordidge^2,3, Felice d’Arco⁴, Kshitij Mankad⁴, Tara D Barwick^2,3, Adam D Waldman^5,6, Patrick W Hales⁷, and Matthew Grech-Sollars^2,3
1Department of Computing, Imperial College London, London, United Kingdom, ²Department of Surgery and Cancer, Imperial College London, London, United Kingdom, ³Department of Imaging, Imperial College Healthcare NHS Trust UK, London, United Kingdom, ⁴Great Ormond Street Children’s Hospital, London, United Kingdom, ⁵Department of Medicine, Imperial College London, London, United Kingdom, ⁶Centre for Clinical Brain Sciences, The University of Edinburgh, Edinburgh, United Kingdom, ⁷Department of Developmental Neurosciences, UCL Great Ormond Street Institute of Child Health, London, United Kingdom

Radiological diagnosis of certain brain tumours remains poor. The development of automated tools that can extract valuable, yet currently underused, information from routine MRI acquisitions have been explored. The present study aimed to evaluate the accuracy of three automatic tumour segmentation programmes using the BRATS validation 2018 glioma dataset (n=66). The performance of BraTumIA was assessed across an adult glioma dataset from Imperial College Healthcare NHS Trust (n=13) and a paediatric brain tumour dataset from Great Ormond Street Hospital NHS Foundation Trust. DeepMedic segments adult gliomas more accurately than BraTumIA and ONCOhabitats. However, BraTumIA provides a more user-friendly segmentation tool.

Yuki Kanazawa¹, Masafumi Harada¹, Mitsuharu Miyoshi², Takashi Abe¹, Yuki Matsumoto¹, and Yasushi Takagi^3
1Tokushima University, Tokushima, Japan, ²Global MR Applications and Workflow, GE Healthcare Japan, Hino, Japan, ³Department of Neurosurgery, Tokushima University, Tokushima, Japan

To achieve detailed characterization of brain tumors, we demonstrated CEST imaging with multi-pool model [bulk water, magnetization transfer (MT), amide proton transfer (APT), and nuclear Overhauser effect (NOE)] derive from the Bloch equation. This study was performed using five patients with brain tumors. As a result, there was high linearity between MTR_asym and T₂/T₁ in each tumor (R² = 0.91, P<0.05). Moreover, there were significantly different NOE signals for tumors and normal white matter (P<0.05). Detection using APT and NOE make it possible to provide more detailed information of brain tumors than MTR_asym analysis.

Hatef Mehrabian¹, Mark Ruschin², Sten Myrehaug^3,4, Hany Soliman^1,3,4, Arjun Sahgal^1,3,4, and Greg J Stanisz^1,5
1Physical Sciences, Sunnybrook Research Institute, Toronto, ON, Canada, ²Medical Physics, Sunnybrook Health Sciences Centre, Toronto, ON, Canada, ³Radiation Oncology, Sunnybrook Health Sciences Centre, Toronto, ON, Canada, ⁴Radiation Oncology, University of Toronto, Toronto, ON, Canada, ⁵Medical Biophysics, University of Toronto, Toronto, ON, Canada

Chemo-radiation treatment for GBM damages normal brain tissues and leads to significant deterioration in patients’ quality of life months or years after treatment. Magnetization transfer (MT) is sensitive to white matter damage and may detect it early. We observed significant reduction in MT exchange rate (RM_0B) that quantifies white matters integrity as early as one month after 6-week chemo-radiation. The treatment impact on normal brain tissue was observed even in regions that received low radiation dose. More importantly, not all patients experienced MT reduction showing the heterogeneity of response to treatment and the potential of MT in identifying vulnerable patients.

Tatsuya Yamamoto¹, Kuniyoshi Hayashi², and Hirohiko Kimura^3
1Department of Diagnostic Radiology, The Cancer Institute Hospital of the Japanese Foundation for Cancer Research, Tokyo, Japan, ²Division of Biostatistics and Bioinformatics, Graduate School of Public Health, St. Luke’s International University, Tokyo, Japan, ³Department of Radiology, Faculty of Medical Sciences, University of Fukui, Fukui, Japan

Primary central nervous system lymphomas (PCNSLs) are sometimes difficult to distinguish from glioblastomas (GBMs) based on routine magnetic resonance examination. This study assessed the utility of histogram analysis of intratumoral contrast-enhanced region using contrast-enhanced T1WI (CET1WI) with a 3D-spoiled gradient recalled acquisition in the steady state sequence for PCNSLs and GBMs to determine whether histogram statistics differed between the two tumors using the commercial software. There were significant differences in skewness, entropy, and angular second moment between PCNSLs and GBMs. This suggests the possibility of the differential diagnosis of PCNSLs and GBMs using histogram analysis of CET1WI.

Rui Zhang¹, Yan Bai¹, Xianchang Zhang², Qin Feng¹, Mengke Wang¹, Menghuan Zhang¹, Mathias Nittka³, Gregor Koerzdoerfer³, Yusong Lin⁴, and Meiyun Wang^1
1Department of Medical Imaging, Henan Provincial People’s Hospital & Zhengzhou University People’s Hospital;Henan Key Laboratory of Neurological Imaging, Zhengzhou, China, ²MR Collaboration, Siemens Healthcare Ltd., Beijing, China, ³MR Pre-development, Siemens Healthcare, Erlangen, Germany, ⁴Cooperative Innovation Center of Internet Healthcare & School of Software and Applied Technology, Zhengzhou University, Zhengzhou, China

Tumor consistency is an important factor in determining the resectability and surgical outcome of meningiomas. This study investigated the use of magnetic resonance fingerprinting (MRF) to distinguish between meningiomas characterized as soft or hard based on intraoperative patient records. Mann-Whitney U tests were used to compare longitudinal relaxation time (T1) and transverse relaxation time (T2) values derived from MRF. Soft meningiomas had significantly higher T1 and T2 values than hard meningiomas. Both T1 and T2 showed good diagnostic performance, suggesting that MRF may provide valuable information for preoperative differentiation of meningioma consistency and help guide the treatment plan.

Teng Jin¹, Jing Wang¹, and Zhenwei Yao^2
1Radiology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China, ²Department of Radiology, Huashan Hospital, Fudan University, Shanghai, China

For reducing the gadolinium dose of patients with metastasis, we seek a valuable tool to lower contrast agent doses but keep MR image quality, which is important for patients’ healthcare. We analyzed the optimum scanning time for 1/2 dose CE-T2FLAIR, and compared the CR value of metastases and enhancement degree among the 1/2 dose CE-T2 FLAIR, 1/2 dose CE-T1WI and full dose CE-BRAVO sequence. Our team further analyzed the relationship between enhancement pattern, lesion size and image quality. Our study not only reduced the harm from GBCA administrations but also improved detection rate of brain metastases.

Kamlesh Jobanputra¹, Bhaswati Roy², Karuna Raj¹, Amit Agarwal¹, and Frank Yu^1
1Radiology, University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Anesthesiology, University of California at Los Angeles, Los Angeles, CA, United States

An investigation that stemmed from observation that there were dentate volume changes on SWI in low grade tumor of the brainstem showed significant decrease in signal intensity in the ipsilateral dentate nucleus in high grade neoplasms and decrease in ipsilateral dentate volumes in low grade neoplasms. We think that dentate iron deposition and its imaging is yet an unexplored field, especially in brain tumor imaging. Further evaluation with QSM and STI with histopathological correlation may also help explore pathways of iron homeostasis.

Koung Mi Kang¹, Seung Hong Choi¹, Inpyeong Hwang¹, Roh-Eul Yoo¹, Tae Jin Yun¹, Ji-hoon Kim¹, and Chul-Ho Sohn^1
1Seoul National University Hospital, Seoul, Republic of Korea

We investigated that whether arterial input functions (AIFs) obtained from Dynamic susceptibility-contrast (DSC)-MRI, or AIF_DSC values, improved reliability and diagnostic accuracy of DCE parameters, for differentiating glioblastoma (GBM) from primary CNS Lymphoma (PCNSL), compared with AIFs derived from DCE-MRI (AIF_DCE). This study demonstrated that AIF_DSC –driven DCE parameters had better diagnostic accuracy and reliability for the differential diagnosis than those from AIF_DCE. In addition, the AIF_DSC-driven mean K_trans showed better accuracy than the other DCE parameters and rCBV in the discrimination of GBM without marked vascularity from PCNSL.

Jian Wang¹, Xin Gao¹, Chunhui Jiang¹, and Jundi Liu^1
1The First Affiliated Hospital of Xinjiang Medical University, Urumqi, China, Xinjiang, China

Due to similar appearance in conventional MR images, it is often difficult to differentiate cerebral vesicular echinococcosis (CAE) from brain metastases (BM). In this study, user perfusion weighted imaging technique, perfusion characteristics of these 2 diseases were systematically investigated. Our results indicated that significant difference of semi-quantitative parameters relative cerebral blood flow (rCBF) and relative cerebral blood volume (rCBV) from PWI between CAE and BM. Hence PWI can provide valuable additional information in differential disagnosis of alveolar echinococcosis and brain metastases.

Inpyeong Hwang¹, Seung Hong Choi¹, Jin Wook Kim², Roh-Eul Yoo¹, Koung Mi Kang¹, Tae Jin Yun¹, Ji-hoon Kim¹, and Chul-ho Sohn^1
1Department of Radiology, Seoul National University Hospital, Seoul, Korea, Republic of, ²Department of Neurosurgery, Seoul National University Hospital, Seoul, Korea, Republic of

There are few known predictive factors for response to gamma-knife radiosurgery (GKRS) in vestibular schwannoma (VS). This study investigated the relationship between DCE-MR derived pharmacokinetic parameters and tumor volume change at one year after GKRS follow-up in VS. This prospective study performed a volumetric measurement of DCE-MR pharmacokinetic parameters before GKRS. The patients underwent followed-up MR at post one year, tumor volume was measured, and pharmacokinetic parameters were compared between tumor growth group and stable group. Only the K^trans value showed statistical differences. Our results suggest that K^trans value has the potential to predict tumor response in VS after GKRS.

Laura C. Bell¹, Leland S. Hu², Yuxiang Zhou², Kathleen M. Schmainda³, Jerrold L. Boxerman⁴, and C. Chad Quarles^1
1Barrow Neurological Institute, Phoenix, AZ, United States, ²Mayo Clinic, Scottsdale, AZ, United States, ³Medical College of Wisconsin53226, Milwaukee, WI, United States, ⁴Rhode Island Hospital and Alpert Medical School of Brown University, Providence, RI, United States

Current recommendations for brain tumor dynamic susceptibility contrast MRI include the use of a moderate flip angle (MFA) and a full standard preload dose.¹ Recently, simulations² identified an equally accurate protocol that requires no preload and uses a low flip angle (LFA) and optimized echo time.  This was validated in vivo³ by comparing LFA- and MFA-based mean tumor cerebral blood volume. Here we aim to systematically characterize agreement by evaluating the pixel-wise agreement in enhancing tumor, edema, white matter, and gray matter regions. We found high agreement in all regions further strengthening the clinical utility of the LFA protocol.

Natenael B Semmineh¹, Usha Bhalla², Laura C Bell¹, Ashley M Stokes¹, Matthew D Lee³, Leland Hu⁴, Jerrold L Boxerman^3,5, and C Chad Quarles^1
1Division of Neuroimaging Research, Barrow Neurological Institute, Phoenix, AZ, United States, ²Brown University, Providence, RI, United States, ³Alpert Medical School - Brown University, Providence, RI, United States, ⁴Department of Radiology, Mayo Clinic Arizona, phoenix, AZ, United States, ⁵Department of Diagnostic Imaging, Rhode Island Hospital, Providence, RI, United States

Numerous methods have been proposed to overcome the confounding influence of contrast agent leakage on DSC-MRI derived CBV maps in brain tumors. Recently, we leveraged a digital reference object to identify the optimal acquisition and analysis methods for GBM patients, a protocol that is now being recommended as the standardized approach for all clinical scans. It is unknown if the biological differences between GBM and metastases would confound the use of this protocol in patients with brain metastases. In this study we characterize the influence of this GBM-centric protocol resulted in lower metastases CBV accuracy and precision.

Mamta Gupta¹, Abhinav Gupta¹, Anup Singh², Jitender Saini³, Rana Patir⁴, Sunita Ahlawat⁵, Vani Santosh³, Neha Vats², Manish Awasthi², Suhail Parvaze⁶, and Rakesh Kumar Gupta^1
1Department of Radiology, Fortis Memorial Research Institute, Gurgaon, India, ²Centre for Biomedical Engineering, Indian Institute of Technology, New Delhi, India, ³National Institute of Mental, Health and Neurosciences, Bangalore, India, ⁴Department of Neurosurgery, Fortis Memorial Research Institute, Gurgaon, India, ⁵SRL Diagnostics, Fortis Remorial Research Institute, Gurgaon, India, ⁶Philips Innovation Campus, Bangalore, India

The differentiation of oligodendroglioma and astrocytoma has become increasingly important due to the distinct sensitivity of oligodendroglioma (OD) to chemotherapy and prolonged survival compared to astrocytoma (AT) of similar grades. The purpose of the present study was to determine whether unique characteristics of OD and AT are visible on MR imaging and to assess the added value of perfusion imaging in differentiating OD from AT tumors across similar grades in large study population. Our results demonstrated no characteristic/specific conventional MRI features and perfusion parameters those could clearly differentiate similar grades of OD’s and AT tumors.

Takashi Abe¹, Maki Otomo¹, Mihoko Kondo¹, Yuta Arai¹, Yoichi Otomi¹, Yuki Matsumoto², Yuki Kanazawa², Kohei Nakajima³, Yoshifumi Mizobuchi³, Yasushi Takagi³, and Masafumi Harada^2
1Radiology and Radiation Oncology, Tokushima University Hospital, Tokushima, Japan, ²Tokushima University, Tokushima, Japan, ³Neurosergery, Tokushima University Hospital, Tokushima, Japan

One hundred one cases with brain tumor were included (62: HGG, 10: LGG, 18: metastasis, 11: lymphoma), and divided into two groups by the presence or absence of ASL hyperintensity outside of CE area (ASL-CE mismatch). Only the cases with glioma or lymphoma (23 HGG, 7 LGG, 3 lymphoma) revealed ASL-CE mismatch, and the cases with metastasis didn't (p = 0.001). Similar results were obtained in a sub-analysis where data were divided by MRI vendor. This novel finding may reflect tumor invasion and/or vascular growth in the surrounding brain parenchyma, and is useful for identifying metastasis from the other tumors.

Paulien H.M. Voorter^1,2, Marieke van den Kerkhof^2,3, Daniëlle B.P. Eekers⁴, Richard A.M. Canters⁴, Wouter van Elmpt⁴, Joost J. de Jong^2,3, Lisanne P.W. Canjels^2,3,5, Jacobus F.A. Jansen^2,3,5, Alida A. Postma², and Walter H. Backes^2,3
1Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ²Department of Radiology and Nuclear Medicine, Maastricht University Medical Center, Maastricht, Netherlands, ³School for Mental Health and Neuroscience, Maastricht University, Maastricht, Netherlands, ⁴Department of Radiation Oncology (MAASTRO), GROW – School for Oncology and Developmental Biology, Maastricht University Medical Center, Maastricht, Netherlands, ⁵Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands

Blood-brain barrier (BBB) disruption underlies the origin of many brain disorders. Measuring subtle BBB leakage of gadolinium-based contrast agents is commonly performed by a continuous and time-consuming dynamic MRI protocol. However, BBB leakage is subtle, for which discrete sampling at strategic time-points might be sufficient. This study explores the feasibility of the time-efficient interleaved protocol by applying it to brain tumor patients. We were able to measure high permeability in a craniopharyngioma and low permeability in a low-grade tumor and in healthy tissue. The time-efficient protocol is promising and will be further evaluated in more patients.

Ingomar Gutmann¹, Julia Furtner², Gregor Kasprian², Gerhard Krexner³, Daniela Prayer², and Patrick Thurner^2
1University of Vienna, Vienna, Austria, ²Medical University Vienna, Vienna, Austria, ³University of Vienna, Wien, Austria

We present DCE acquisition with high spatial resolution that excels the guidelines of the ACR for spatial resolution, maintaining whole brain coverage at all times, while at the same time providing a temporal sampling rate of fast enough to make accurate flow and permeability measurements from T1-weighted DCE MRI in a clinical neuro-oncology setting possible.Furthermore we demonstrate the use of radial spoke aggregation in GRASP reconstruction for DCE perfusion imaging of the brain and show preliminary results on the impact of variable temporal resolution on estimating pharmacokinetic models.

Ivar Thokle Hovden¹, Oliver M. Geier¹, Ingrid Digernes¹, Elies Fuster-Garcia¹, Grethe Løvland², Einar Vik-Mo³, Torstein R. Meling^3,4, and Kyrre E. Emblem^1
1Department of Diagnostic Physics, Oslo University Hospital, Oslo, Norway, ²The Intervention Centre, Oslo University Hospital, Oslo, Norway, ³Department of Neurosurgery, Oslo University Hospital, Oslo, Norway, ⁴Department of Neurosurgery, Geneva University Hospitals, Geneva, Switzerland

We studied the changes in rCBV introduced by echo-planar imaging (EPI) distortion correction methods when applied to EPI dynamic susceptibility contrast (DSC)-MRI. The results obtained using FSL TOPUP and EPIC distortion correction methods indicate that both subcortical and cortical regions are affected and returning an overall increase in rCBV. In the context of longitudinal EPI-based analysis in glioblastoma patients, EPI distortions and subsequent corrections are important determiners for assessing robust responses of rCBV change.
 

Yangyingqiu Liu¹, Yanwei Miao¹, Ailian Liu¹, Qingwei Song¹, and Lizhi Xie^2
1Department of Radiology, First Affiliated Hospital of Dalian Medical University, Dalian, China, ²GE healthcare, Beijing, China

Proliferative states of pituitary macroadenomas is assessed by radiomics analysis on Dynamic Contrast Enhanced MRI(DCE-MRI). The results show that DCE-MRI could be used as a new noninvasive method for identification the proliferative states of pituitary macroadenomas.

zongqiong sun¹, weiqiang dou², and shudong hu^1
1MR, Affiliated hospital of Jiangnan university, Wuxi, China, ²MR research, GE healthcare,MR research China, Beijing, China

CBF values of PCNSL are significant lower than those of MBT, and this is helpful to differentiate the two tumors by using 3D ASL MR perfusion imaging from quantitative parameters analyses.

Li Xiang¹, Lihua Sun¹, Longsheng Wang¹, and Yaqiong Ge^2
1Radiology, The second affiliated hospital of Anhui Medical University, Hefei, China, ²GE Healthcare, Shanghai, China

DCE-MRI was used to conduct quantitative analysis of microvascular permeability in three different areas of meningioma: Lesion, Edge and Normal. 37 patients with pathologically confirmed meningioma were enrolled, while quantitative parameters were measured. The parameters of k^trans, TTP, MAX. Conc, iAUC in Lesion group and Edge group were statistically significant, and the parameters of K^trans, k_ep, V_e, TTP, Max.Conc as well as AUC in Lesion group and Normal group, were statistically significance. The ktrans parameter has the highest AUC value, which means the better performance for permeability study of WHO with different grades of meningiomas and invasive meningiomas. 

David Lu¹, Jay J. Pillai^1,2, Hanzhang Lu¹, and Yang Li^1
1Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ²Department of Neurosurgery, Johns Hopkins University School of Medicine, Baltimore, MD, United States

We developed a histogram-based CBF analysis method for the prediction of tumor histopathologic grade and molecular marker (including IDH1 mutation, ATRX loss, p53 mutation, MGMT methylation, and 1p/19q co-deletion) in a group of de novo brain glioma patients. ASL MRI may be a non-invasive and cost-effective approach to assist in brain tumor diagnosis and prognosis.

Arian Lasocki^1,2 and Grant McArthur^2,3
1Cancer Imaging, Peter MacCallum Cancer Centre, Melbourne, Australia, ²The Sir Peter MacCallum Department of Oncology, The University of Melbourne, Parkville, Australia, ³Medical Oncology, Peter MacCallum Cancer Centre, Melbourne, Australia

In small intracranial melanoma metastases, MRI evidence of melanin was more common in BRAF^V600-mutant patients than BRAF-wildtype, supporting that melanin content is greater in not just the primary, but also in the distant metastases. Haemorrhage and central necrosis were more common in larger metastases. Central necrosis was also more common in BRAF-mutant patients who had not previously received anti-BRAF therapy, suggesting that metastases in BRAF-mutant patients have a tendency to central necrosis, though this is modified by anti-BRAF therapy, thus resulting in more solid metastases.

Samuel Bobholz¹, Allison Lowman², Alexander Barrington³, Michael Brehler², Sean McGarry¹, Jennifer Connelly⁴, Elizabeth Cochran⁵, Anjishnu Banerjee⁶, and Peter LaViolette^2,3
1Biophysics, Medical College of Wisconsin, Wauwatosa, WI, United States, ²Radiology, Medical College of Wisconsin, Wauwatosa, WI, United States, ³Biomedical Engineering, Medical College of Wisconsin, Wauwatosa, WI, United States, ⁴Neurology, Medical College of Wisconsin, Wauwatosa, WI, United States, ⁵Pathology, Medical College of Wisconsin, Wauwatosa, WI, United States, ⁶Biostatistics, Medical College of Wisconsin, Wauwatosa, WI, United States

This study sought to assess the ability for voxel-wise MRI intensity values to distinguish between co-registered pathological annotations of autopsy tissue samples from brain cancer patients.  Though single image and pairwise image assessments did not reveal separable intensity distributions for the pathological annotation classes, ensemble-based predictive modelling using multiparametric MRI intensities proved able to predict pathological annotations with modest accuracy.  These results suggest a complex relationship between MRI values and pathological features that are most accurately assessed in terms of multiple MR imaging modalities.

Samuel Bobholz¹, Allison Lowman², Alexander Barrington³, Michael Brehler², Jennifer Connelly⁴, Elizabeth Cochran⁵, Anjishnu Banerjee⁶, and Peter LaViolette^2,3
1Biophysics, Medical College of Wisconsin, Wauwatosa, WI, United States, ²Radiology, Medical College of Wisconsin, Wauwatosa, WI, United States, ³Biomedical Engineering, Medical College of Wisconsin, Wauwatosa, WI, United States, ⁴Neurology, Medical College of Wisconsin, Wauwatosa, WI, United States, ⁵Pathology, Medical College of Wisconsin, Wauwatosa, WI, United States, ⁶Biostatistics, Medical College of Wisconsin, Wauwatosa, WI, United States

This study sought to compare localized predictions of cellular density via radiomics-based and neural network-based modeling, using co-registered autopsy tissue samples from 16 brain cancer patients as ground truth.  We found that radiomics models tended to slightly outperform the neural networks, despite evidence of overfitting in all radiomics-based models and an Alexnet-based transfer learning model.  These results suggest that radiomics models tend to perform at least as well as neural network when applied to this dataset, but the propensity of these models for overfitting highlights further needs to be addressed with modelling on larger data sets.

jing Zhang¹, kuan Yao², Zhenyu Liu³, Junlin Zhou¹, Guojing Zhang⁴, and Yuntai Cao^1
1Radiology, Lanzhou University Second Hospital, Lanzhou, China, ²School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, Shanghai, China, ³CAS Key Laboratory of Molecular Imaging, Institute of Automation, Beijing, Beijing, China, ⁴Lanzhou University Second Hospital, Lanzhou, China

Objectives Using a radiomics method to predict brain invasion by meningioma. Methods 1595 quantitative imaging features were extracted. LASSO was performed to select features. SVM was used to fit the predictive model. Furthermore, a nomogram was constructed, and validated using decision curve analysis (DCA). Results 8 features were significantly correlated with brain invasion. The radiomics model derived from the fusing MRI sequences resulted in the best discrimination ability, with AUC of0.855(95%CI, 0.829-0.882), sensitivity of 80.32% (95%CI, 75.56%-85.25%). Conclusions The radiomics model developed in this study provided a new non-invasive way to facilitate the preoperative prediction of brain invasion in meningioma.

Jing Zhang¹, Jianqing Sun², Guojing Zhang¹, Yuntai Cao¹, and Junlin Zhou^1
1Radiology, Lanzhou University Second Hospital, Lanzhou, China, ²Philips Healthcare, Shanghai, Shanghai, China

Objectives Radiomics method was used to predict bone invasion. Methods 1227 quantitative imaging features were extracted. Recursive Feature Elimination (RFE) was performed to select the most informative features. Ridge Classifier was chosen to predict model. Results The AUCof the radiomics model derived from CET1WI and T2WI sequence were0.72,0.72and 0.72,0.64 in the training and test datasets , respectively, and combined CET1WI and T2WI sequences were 0.73and 0.72 when predict bone invasion. Conclusions The radiomics model developed in this study may aid neurosurgeons in the pre-operative prediction of bone invasion by meningiomas ,which can contribute to make clinical strategies and predict prognosis.

Jing Zhang¹, Yanghong Ou¹, and Shuangfeng Dai^2
1Lanzhou University Second Hospital, Lanzhou, China, ²Huiying Medical Technology Co., Ltd., Beijing, China

Radiomics provides a tool for comprehensive quantification and visualization of intra-tumor heterogeneity at the radiological level. Several radiomics studies have been reported in prediction of the survival and treatment response of glioma. And most researches had focused on binary classification of the Low-grade glioma (grade II) and high-grade glioma (grade III and grade IV). However, the influence of different MRI scan plane on the radiomic features of glioma has not been investigated. The purpose of this retrospective study was to demonstrate the feasibility of radiomics methods to determine the three subtypes (grade II, III, and IV) of glioma based on multi-sequences and different scan plane of magnetic resonance imaging (MRI).

Bárbara Schmitz Abecassis^1,2, Koji Sakai², Tomonori Toyotsuji², Yoshiaki Ota^2,3, and Kei Yamada^2
1Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, ²Department of Radiology, Kyoto Prefectural University of Medicine, Kyoto, Japan, ³Department of Radiology, University of Michigan, Ann Arbor, MI, United States

Glioma delineation is complex and time-consuming. We evaluated whether a technically less challenging method could potentially bypass this tedious process, when using a simple manual circular-ROI by less trained personnel. We assessed whether applying a circular-ROI to clinical 3T-MRIs from 12 glioma patients, would extract significantly different GLI features. We compared it to the gold-standard manual delineation approach, having 3 observers with different levels of expertise placing the ROIs. Experience matters to extract consistent GLI features across delineation runs. The different ROI methods extracted significantly different GLI features, emphasizing the importance of delineation strategies when analyzing heterogeneous tumors like gliomas.

Rushi Chen¹, Yan Bai², Mengke Wang², Feng Qin², Menghuan Zhang², Xueping Zhang², Shang Wang², Yongchao Zhao², Ting Fang², Xinhui Wang², Huan Zhao², Li Mao³, Xiuli Li³, and Meiyun Wang^2
1Henan provincial people's hospital& Zhengzhou University People’s Hospital, Zhengzhou, China, ²Henan provincial people's hospital, Zhengzhou, China, ³Deepwise AI Lab, Beijing, China

The primary treatment for prolactinomas is dopamine agonist, while transsphenoidal surgery is recommended as primary therapy for the non-prolactinomas. This study aimed to evaluate the radiomic model in the differentiation of prolactinoma from non-prolactinomas before surgery. Our results demonstrated that the eXtreme Gradient Boosting model based on 2264 radiomic features extracted from the original and preprocessed sagittal contrast-enhanced T1WI and coronal T2WI yielded the AUC value of 1.000 and 0.828 on the train set and test set, respectively. The radiomic model may have potential for differentiating the prolactinomas from non-prolactinomas, which may be beneficial to guide the treatment plan.

Yangyingqiu Liu¹, Yanwei Miao¹, Ailian Liu¹, Qingwei Song¹, and Bing Wu^2
1Department of Radiology, First Affiliated Hospital of Dalian Medical University, Dalian, China, ²GE healthcare, Beijing, China

Hormone immunohistochemistry typing of pituitary macroadenomas is assessed by radiomics analysis on Dynamic Contrast Enhanced MRI(DCE-MRI). The results show that DCE-MRI could be used as a new noninvasive method to predictive the hormone immunohistochemistry typingof pituitary macroadenomas.

Amirah Faisal Alsaedi^1,2, Jasmina Panovska-Griffiths³, Xavier Golay², and Sotirios Bisdas^2,4
1Department of Radiology Technology, Taibah University, Medina, Saudi Arabia, ²Department of Brain Repair & Rehabilitation, UCL, Queen Square, Institute of Neurology, London, United Kingdom, ³Department of Applied Health Research, UCL, London, United Kingdom, ⁴Lysholm Department of Neuroradiology, The National Hospital for Neurology and Neurosurgery, University College Hospitals NHS Trust, London, United Kingdom

This study aimed to assess the diagnostic performance of multiparametric MRI radiomics for glioma class prediction according to the WHO 2016 classification. Histogram features were extracted from prospectively acquired multiparametric MRI (pCASL, DSC-MRI, DCE-MRI, and DWI) in 32 patients with primary gliomas. The uncombined significant features of ASL, ADC, DSC, and DCE, revealed diagnostic performances varying from low (44% ) to fair (86%) and unable to predict all the histomolecular classes. However, combining them for each MRI method, independently, enhanced the diagnostic accuracy up to 100% and predict all the classes. This alludes the use of multimodal radiomics for glioma classification.

George Zenzerovich¹ and Tim Q Duong^2
1Radiology, Stony Brook University, Stony Brook, NY, United States, ²Stony Brook University, Stony Brook, NY, United States

Texture features obtained from peritumoral area of diagnostic MR image of Glioblastoma Multiforme were used to improve texture based prediction of tumor progression. The peritumoral area was drawn and examined then derived texture features were added to conventional model to improve performance.

Shingo Kihira¹, Nadejda Tsankova², Adam Bauer³, Yu Sakai¹, Nicole Zubizarreta⁴, Jane Houldsworth², Fahad Khan², Adilia Hormigo⁵, Constantinos Hadjipanayis⁶, and Kambiz Nael^1
1Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, ²Pathology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, ³Radiology, Kaiser Permanente Fontana Medical Center, Fontana, CA, United States, ⁴Institute for Health Care Delivery Sciences, Icahn School of Medicine at Mount Sinai, New York, NY, United States, ⁵Neurology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, ⁶Neurosurgery, Icahn School of Medicine at Mount Sinai, New York, NY, United States

In this retrospective study, we used a commercially available texture analysis software (Olea Medical) to construct a multiparametric MRI radiomic model that can be used to predict several important prognostic biomarkers in a cohort of patients with brain glioma. A total of 92 texture features were calculated from both FLAIR and T1C+ images using a volume-of-interest analysis encompassing the entire FLAIR hyperintense tumor.  Radiomic features obtained from our multiparametric MR texture model were able to predict genetic biomarkers of brain glioma with predictive accuracies ranging from modest (62.4%) for MGMT to nearing 90% for IDH-1 and ARTX.

YiXin HU¹, Lan Li¹, JiuQuan Zhang¹, and Jianqing SUN^2
1Chongqing University Cancer Hospital & Chongqing Cancer Institute & Chongqing Cancer Hospital, ChongQing, China, ²Philips Healthcare, Shanghai, China

To build a radiomics signature from MR images to predict high grade glioma (HGG, WHO grade III and IV) patients with different prognosis. Then, to explore the predictive value based on the edge of the tumor area and PBZ, respectively.

Junyi Dong¹, Yangyingqiu Liu², Yanwei Miao¹, Huicong Shen³, and Yan Guo^4
1First Affiliated Hospital of Dalian Medical University, Dalian,116011,China, Dalian, China, ²Department of Radiology, First Affiliated Hospital of Dalian Medical University, Dalian,116011,China, Dalian, China, ³Beijing Tiantan hospital, Beijing, China, ⁴Life science, Shenyang, China

Intracranial solitary fibrous tumor(SFT)/hemangiopericytoma (HPC) is a rare malignant tumor originating from the intracranial vasculature, while Angiomatous meningioma (AM) is also a rare benign one as a histological subtype of meningioma with World Health Origination (WHO) grade I. The two tumors have similar location and conventional MRI features, but the treatment and prognosis are quite different. Texture analysis can get more information that can't be seen in conventional MRI images. It is very necessary to establish the differentiation model of texture analysis between the two kinds of tumors. 

Zezhong Ye¹, Sam E. Gary², Jeffrey D. Viox³, Anthony T. Wu⁴, Joshua Lin⁵, Peng Sun¹, Joshua B. Rubin⁶, Sonika Dahiya⁷, and Sheng-Kwei Song^1
1Radiology, Washington University School of Medicine, St. Louis, MO, United States, ²Medical Scientist Training Program, The University of Alabama at Birmingham, Birmingham, AL, United States, ³School of Medicine, University of Missouri - Kansas City, Kansas City, MO, United States, ⁴Biomedical Engineering, Washington University, St. Louis, MO, United States, ⁵Keck School of Medicine, The University of Southern California, Los Angeles, CA, United States, ⁶Pediatrics, Washington University School of Medicine, St. Louis, MO, United States, ⁷Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, United States

Current clinical diagnosis, surgical resection, and assessment of treatment response for high-grade brain tumor patients relies heavily on gadolinium-enhanced T1-weighted MRI, although such imaging is non-specific for tumor and merely reflects a disrupted blood-brain barrier. The complex tumor microenvironment and spatial heterogeneity make high-grade brain tumor very difficult to characterize using current clinical imaging modalities. We developed a novel imaging strategy to characterize key tumor histological features and demonstrated its capability to accurately predict these histology. Extending this approach to larger cohorts of both tumor specimens and patients could provide further validation and facilitate its clinical translation for patient management.

Qing Ji¹, Matthew Scoggins¹, Angela Edwards¹, John O. Glass¹, Tara Brinkman^2,3, Zoltan Patay¹, and Wilburn E. Reddick^1
1Diagnostic Imaging, St.Jude Children's Research Hospital, Memphis, TN, United States, ²Psychology, St.Jude Children's Research Hospital, Memphis, TN, United States, ³Epidemiology and Cancer Control, St.Jude Children's Research Hospital, Memphis, TN, United States

A CTC fiber pathway template was built in MNI space from the CTC data of 10 healthy controls. The template was aligned with the principle diffusion directions of each individual DTI data of additional 10 healthy controls and 11 post-surgical medulloblastoma patients using a linear algorithm developed in this study. The aligned template can accurately mimic the real CTC pathways in an individual subject in both healthy and post-surgical subjects. The DTI parameter values in post-surgical medulloblastoma patients can be accessed using the transformed CTC pathway. 

Sirui Li¹, Wenbo Sun¹, Yuan Zheng², Qing Wei³, Samo Lasic⁴, Shihong Han³, Shuheng Zhang³, Danielle van Westen⁵, Karin Bryskhe⁴, Daniel Topgaard^4,5, and Haibo Xu^1
1Zhongnan Hospital of Wuhan University, Wuhan, China, ²UIH America, Inc., Houston, TX, United States, ³United Imaging Healthcare, Shanghai, China, ⁴Random Walk Imaging, Lund, Sweden, ⁵Lund University, Lund, Sweden

Diffusion kurtosis decomposition (DKD) is a novel advanced diffusion MRI modality relying on customized pulse sequences and high-performance hardware to assess cell shapes and density heterogeneity via the anisotropic and isotropic mean kurtosis parameters, MKa and MKi, which are fundamentally different microstructural properties that are inextricably entangled in conventional diffusion kurtosis imaging (DKI). We have investigated DKD imaging of gliomas in a clinical setting, and for the first time established the correlations between MKa, MKi, and tumor grade. In comparison to conventional diffusion methods, DKD more accurately describes the microstructural changes and provides a useful tool for glioma diagnosis.

Gao Ankang¹, Zhang Huiting², Yang Guang³, Wang Shaoyu², Yan Xu², Bai Jie¹, and Cheng Jingliang^1
1MRI, Dept. of MRI, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, ²MR Scientific Marketing, Siemens Healthcare, Shanghai, China, ³Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China

Diffusion imaging is widely used to noninvasively detect the microscopic diffusion properties of biological tissues in vivo. Advanced diffusion models were recently proposed to provide additional microstructure information. In present work, we applied four diffusion models in glioma grading, including DTI, DKI, MAP-MRI and NODDI model, which could be acquired within a single scan.

Gao Ankang¹, Chen Qianqian¹, Zhu Jinxia², Stefan Huwer³, Bai Jie¹, and Cheng Jingliang^1
1MRI, Dept. of MRI, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, ²MR Collaboration, Siemens Healthcare Ltd., Beijing, China, Beijing, China, ³Siemens Healthcare GmbH, Erlangen, Germany, Erlangen, Germany

By observing the gradually developing characteristics of track density in the area of peritumoral edema, the extent of possible tumor invasion was judged. This may contribute to neurosurgical planning, the postoperative quality of life, and longer survival expectancy. Simultaneously, TDI was able to show the crossing fibers of the tumor part but could not resolve the disturbed diffusion of severe peritumoral edema’s influence on the fibers.

Lia Talozzi¹, Matteo Martinoni², Micaela Mitolo ³, Filippo Badaloni², Sofia Asioli^1,4, Claudia Testa⁵, Raffaele Lodi^1,6, and Caterina Tonon^1,3
1Biomedical and Neuromotor Sciences, University of Bologna, Bologna, Italy, ²IRCCS Istituto delle Scienze Neurologiche di Bologna, Neurosurgery Unit, Bologna, Italy, ³IRCCS Istituto delle Scienze Neurologiche di Bologna, Neuroradiology Functional Diagnostic Unit, Bologna, Italy, ⁴Section of Anatomic Pathology 'M. Malpighi', Bellaria Hospital, Bologna, Italy, ⁵Physics and Astronomy, University of Bologna, Bologna, Italy, ⁶, IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy

Probabilistic arcuate fasciculus (AF) tractography was computed in healthy controls (HC) and for a patient with a diffuse glioma, IDH 1 wild type left frontal lobe.  Bilateral curvature mapping and along-tract DTI measures were evaluated, modelling surface geometry. In HC, hemispheric asymmetries in tract curvature showed a more lateral and inferior trajectory on the left, not measured for the patient’s left AF, medially dislocated. In HC, a left-lateralized asymmetries (higher FA and volume) was measured, and it was preserved for the patient. During pre-surgical evaluation, the proposed developed tractography analyses allowed a quantitative investigation of AF dislocation and microstructural integrity.

Daniel Krahulec¹, Ahmed Radwan², Stefan Sunaert², Maarten Versluis¹, Kim van de Ven¹, and Marcel Breeuwer^1,3
1MR R&D Clinical Science, Philips Healthcare, Best, Netherlands, ²Department of Imaging & Pathology, Translational MRI, KU Leuven, Leuven, Belgium, ³Department of Biomedical Engineering – Medical Image Analysis, Eindhoven University of Technology, Eindhoven, Netherlands

A data analysis pipeline comprising KUL neuroimaging tools was applied to analyze diffusion and anatomical MRI datasets with neoplastic lesions and perform diffusion tractography with anatomical priors. We utilized a novel framework for reducing the amount of false positive streamlines to allow for improved visualization of different fiber bundles. Results illustrate the meaningfulness of this approach in neurosurgery workflow through offering clinicians more information on the quality of perilesional fiber bundles as well as where to resect with care to preserve functionally eloquent areas.

Marco Borri¹, Jose Pedro Lavrador², Irene Brumer^1,3, Enrico De Vita⁴, Jonathan Ashmore^1,5, Francesco Vergani², Ranjeev Bhangoo², Keyoumars Ashkan², and Jozef Jarosz^1
1Neuroradiology, King's College Hospital, London, United Kingdom, ²Neurosurgery, King's College Hospital, London, United Kingdom, ³Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, ⁴Department of Biomedical Engineering, School of Biomedical Engineering and Imaging Sciences, London, United Kingdom, ⁵Department of Medical Physics and Bioengineering, NHS Highland, Inverness, United Kingdom

In this work we have evaluated the feasibility of incorporating advanced fMRI and tractography evaluations into the real-life presurgical management of patients undergoing brain lesion resections. In this pilot cohort of patients, all major preoperative imaging findings were validated by intraoperative measurements. With the inclusion of fMRI end regions, probabilistic tractography allowed a better reconstruction of the corticospinal tract and its branches in regions adjacent or within the tumour with altered or damaged fibre architecture. Robust fMRI-based language lateralization was able to describe likely dominance, in agreement with intraoperative findings and initial postoperative deficit.

Mridula Vij¹, Archana Vadiraj Malagi¹, Esha Badiya Kayal¹, Jitender Saini², and Amit Mehndiratta^1
1Centre for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi, India, ²Department of Neuroimaging & Interventional Radiology, National Institute of Mental Health and Neuro Sciences, Bengaluru, India

The occurrence of brain tumor is increasing; it is believed that early detection will prove to be beneficial for its treatment. IVIM has proved to provide valuable information and has been explored for brain tumor in prior studies. In the following study, IVIM analysis for brain tumor detection was performed using a novel Biexponential model with Total Variation regularization function (BE+TV model), and a comparison with established Biexponential model was performed. Improvement in coefficient of variation by 16.5 to 43.4% was observed across IVIM parameter. Perfusion faction was estimated to be reliably with this novel methodology.  

Simran Kukran^1,2, Marianna Inglese², Katherine L Ordidge^2,3, Claire Davies ², Lesley Honeyfield³, Babar L Vaqas⁴, Sophie Camp⁴, David Peterson⁴, Kevin O'Neill⁴, Clara Limback-Stanic⁵, Tara Barwick^2,3, Adam Waldman^6,7, and Matthew Grech-Sollars^2,3
1Department of Radiotherapy and Imaging, Institute of Cancer Research, Sutton, United Kingdom, ²Department of Surgery and Cancer, Imperial College London, London, United Kingdom, ³Department of Imaging, Imperial College London, London, United Kingdom, ⁴Department of Neurosurgery, Imperial College London, London, United Kingdom, ⁵Department of Histopathology, Imperial College London, London, United Kingdom, ⁶Department of Medicine, Imperial College London, London, United Kingdom, ⁷Centre for Clinical Brain Sciences, The University of Edinburgh, Edinburgh, United Kingdom

It is thought that hypercellular regions of glioma have lower ADC values via a restriction in flow path, but there is no consensus regarding the correlation between ADC values and the cellularity of histological biopsies. In this study a slightly stronger negative correlation was observed between sample cellularity and the average ADC across the biopsy region as compared to the average ADC across the whole tumour in patients with glioma. However, neither correlation was found to be significant, which could be due to a small cohort size and the variation in tumour biological factors other than cellularity affecting ADC.

Xiaodan Chen¹, Ying Chen², Lin Lin³, Jie Wu⁴, and Guang Yang^4
1Radiology, Fujian Cancer Hospital, Fuzhou, China, ²Fujian Cancer Hospital, Fuzhou, China, ³Fujian Medical University Union Hospital, Fuzhou, China, ⁴Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China, Shanghai, China

Objectives Presurgical grading is particularly important for selecting the best therapeutic strategy for meningioma patients. Therefore, our study is to investigate the value of histogram analysis of diffusion kurtosis imaging (DKI) maps in the differentiation of grades and histological subtypes of meningiomas.Methods A total of 172 patients with histopathologically proven meningiomas underwent preoperative magnetic resonance imaging (MRI) and were classified into low-grade and high-grade groups. Mean Kurtosis (MK), fractional anisotropy (FA), mean diffusivity (MD) histograms were generated based on solid components of the whole tumour. The following parameters of each histogram were obtained：10th， 25th， 75th, and 90th percentiles, mean，median，maximum，minimum，and kurtosis, skewness, and variance. Comparisons of different grades and subtypes were made by Mann-Whitney U test，Kruskal-Wallis tests, ROC curves analysis, and multiple logistic regression.Results Significantly higher maximum, Skewness, and variance of MD, mean, median, maximum, variance, 10th, 25th, 75th, and 90th percentile of MK were found in high-grade than low-grade meningiomas (all P<0.05). DKI histogram parameters differentiated 7 of 10 pairs of subtype pairs (atypical versus meningothelial/ fibrous/transitional/angiomatous meningiomas; angiomatous versus fibrous/transitional meningioma; fibrous versus meningothelial meningiomas). The 90th percentile of MK yielded the highest AUC of 0.870 and was the only independent indicators for grading meningiomas.Conclusion The histogram analysis of DKI is useful for differentiating meningioma grades and subtypes. The 90th percentile of MK may serve as an optimal parameter for predicting the grade of meningiomas.

shenghui Xie¹, shaoyu Wang², xu Yan², huapeng Zhang², guang Yang³, and yang Gao^1
1Department of Radiology, Affiliated Hospital of Inner Mongolia Medical University, Hohhot, China, ²Siemens Healthineers, Shanghai, China, ³East China Normal University, Shanghai, China

The purpose of this study was to explore the diagnostic value of multimode diffusion imaging in the evaluation of glioma tumor grade and proliferation activity. (DKI) and MAP-MRImodel). The results showed that there were significant differences in MK, RTOP and QIV between high and low grade glioma, and the values of MK and QIV of the tumor parenchyma were positively correlated with ki-67. MK and QIV have great potential in predicting tumor proliferation activity. Multimode diffusion-weighted imaging is valuable for the evaluation of preoperative glioma grade and tumor proliferation activity.

rui hu¹, yi jun wu¹, wen chen¹, ke hua zheng¹, dai zhong wang¹, and lin xu^1
1taihe hospital, shiyan, China

To investigate the value of MRI characteristics in the differential diagnosis of hemangiopericytoma and angiomatous meningioma.15 cases of AM and 11 cases of the HPC which were confirmed by pathology. MRI sequences were routinely available. The mean ADC values of HPC was significantly higher than that of AM (t = 2.613, p =0.02). The mean CBF values (t= -8.99, P<0.01) and FA values (t= -3.66, P<0.01) were lower for HPC compared with AM. MRI imaging characteristics of HPC and AM have significant difference. Combining structural MRI comparative analysis and functional MRI was useful for distinguishing HPC from MA.

Simin Zhang¹, Xiaorui Su¹, Weina Wang¹, Qiang Yue¹, and Qiyong Gong^1
1Huaxi MR Research Center，Department of Radiology, West China Hospital of Sichuan University, Chengdu, China

Multicentric gliomas are rare lesions of central nervous system. Due to the invasive nature of glioma, differentiating multicentric gliomas from diffuse gliomas is difficult. As yet there remains lack of methods to elucidate lesions are totally separated unless surgical biopsy.  We use probabilistic fiber tracking to identify the white matter dissemination routes. Our results showed  the connectivity value and probabilistic value of the lesions in multicentric gliomas were significantly lower than the lesions in diffuse gliomas. The features provide new insight into multicentric gliomas and may aid in accurate classification of  multicentric glioma and  diffuse glioma in vivo.

Andrey Zhylka¹, Nico Sollmann^2,3, Alberto De Luca⁴, Daniel Krahulec⁵, Marcel Breeuwer^1,5, Alexander Leemans⁴, and Josien Pluim^1
1Biomedical Engineering department, Eindhoven University of Technology, Eindhoven, Netherlands, ²Department of Diagnostic and Interventional Neuroradiology, , Klinikum rechts der Isar, Technische Universität München, Munich, Germany, ³TUM-Neuroimaging Center, Klinikum rechts der Isar, Technische Universität München, Munich, Germany, ⁴University Medical Center Utrecht, Utrecht, Netherlands, ⁵MR R&D Clinical Science, Philips Healthcare, Best, Netherlands

Conventional deterministic fiber tractography approaches commonly used in clinical applications are prone to generating false-negative reconstructions, which might influence further decision-making related to treatment and repeated surgery in patients with brain tumors. Surgery-related effects, such as blood inflow into white matter and edema, further distort the diffusion signal, complicating the task of tractography. We evaluated a novel multi-level fiber tractography approach on data of subjects who had undergone tumor resection. A comparison with conventional deterministic approaches is performed. The results were correlated with the reported motor-function deficit grades.