Noninvasive ROS imaging and drug delivery monitoring in the tumor microenvironment.

Reactive oxygen species (ROS) that are overproduced in certain tumors can be considered an indicator of oxidative stress levels in the tissue. Here, we report a magnetic resonance imaging (MRI)-based probe capable of detecting ROS levels in the tumor microenvironment (TME) using ROS-responsive manganese ion (Mn2+)-chelated, biotinylated bilirubin nanoparticles (Mn@bt-BRNPs). These nanoparticles are disrupted in the presence of ROS, resulting in the release of free Mn2+, which induces T1-weighted MRI signal enhancement. Mn@BRNPs show more rapid and greater MRI signal enhancement in high ROS-producing A549 lung carcinoma cells compared with low ROS-producing DU145 prostate cancer cells. A pseudo three-compartment model devised for the ROS-reactive MRI probe enables mapping of the distribution and concentration of ROS within the tumor. Furthermore, doxorubicin-loaded, cancer-targeting ligand biotin-conjugated Dox/Mn@bt-BRNPs show considerable accumulation in A549 tumors and also effectively inhibit tumor growth without causing body weight loss, suggesting their usefulness as a new theranostic agent. Collectively, these findings suggest that Mn@bt-BRNPs could be used as an imaging probe capable of detecting ROS levels and monitoring drug delivery in the TME with potential applicability to other inflammatory diseases.

CycleSeg: Simultaneous synthetic CT generation and unsupervised segmentation for MR-only radiotherapy treatment planning of prostate cancer.

BACKGROUND: MR-only radiotherapy treatment planning is an attractive alternative to conventional workflow, reducing scan time and ionizing radiation. It is crucial to derive the electron density map or synthetic CT (sCT) from MR data to perform dose calculations to enable MR-only treatment planning. Automatic segmentation of relevant organs in MR images can accelerate the process by preventing the time-consuming manual contouring step. However, the segmentation label is available only for CT data in many cases. PURPOSE: We propose CycleSeg, a unified framework that generates sCT and corresponding segmentation from MR images without access to MR segmentation labels METHODS: CycleSeg utilizes the CycleGAN formulation to perform unpaired synthesis of sCT and image alignment. To enable MR (sCT) segmentation, CycleSeg incorporates unsupervised domain adaptation by using a pseudo-labeling approach with feature alignment in semantic segmentation space. In contrast to previous approaches that perform segmentation on MR data, CycleSeg could perform segmentation on both MR and sCT. Experiments were performed with data from prostate cancer patients, with 78/7/10 subjects in the training/validation/test sets, respectively. RESULTS: CycleSeg showed the best sCT generation results, with the lowest mean absolute error of 102.2 and the lowest Fréchet inception distance of 13.0. CycleSeg also performed best on MR segmentation, with the highest average dice score of 81.0 and 81.1 for MR and sCT segmentation, respectively. Ablation experiments confirmed the contribution of the proposed components of CycleSeg. CONCLUSION: CycleSeg effectively synthesized CT and performed segmentation on MR images of prostate cancer patients. Thus, CycleSeg has the potential to expedite MR-only radiotherapy treatment planning, reducing the prescribed scans and manual segmentation effort, and increasing throughput.

Deep learning-based k-space-to-image reconstruction and super resolution for diffusion-weighted imaging in whole-spine MRI.

PURPOSE: To assess the feasibility of deep learning (DL)-based k-space-to-image reconstruction and super resolution for whole-spine diffusion-weighted imaging (DWI). METHOD: This retrospective study included 97 consecutive patients with hematologic and/or oncologic diseases who underwent DL-processed whole-spine MRI from July 2022 to March 2023. For each patient, conventional (CONV) axial single-shot echo-planar DWI (b = 50, 800 s/mm2) was performed, followed by DL reconstruction and super resolution processing. The presence of malignant lesions and qualitative (overall image quality and diagnostic confidence) and quantitative (nonuniformity [NU], lesion contrast, signal-to-noise ratio [SNR], contrast-to-noise ratio [CNR], and ADC values) parameters were assessed for DL and CONV DWI. RESULTS: Ultimately, 67 patients (mean age, 63.0 years; 35 females) were analyzed. The proportions of vertebrae with malignant lesions for both protocols were not significantly different (P: [0.55-0.99]). The overall image quality and diagnostic confidence scores were higher for DL DWI (all P ≤ 0.002) than CONV DWI. The NU, lesion contrast, SNR, and CNR of each vertebral segment (P ≤ 0.04) but not the NU of the sacral segment (P = 0.51) showed significant differences between protocols. For DL DWI, the NU was lower, and lesion contrast, SNR, and CNR were higher than those of CONV DWI (median values of all segments; 19.8 vs. 22.2, 5.4 vs. 4.3, 7.3 vs. 5.5, and 0.8 vs. 0.7). Mean ADC values of the lesions did not significantly differ between the protocols (P: [0.16-0.89]). CONCLUSIONS: DL reconstruction can improve the image quality of whole-spine diffusion imaging.

Attention fusion network with self-supervised learning for staging of osteonecrosis of the femoral head (ONFH) using multiple MR protocols.

BACKGROUND: Osteonecrosis of the femoral head (ONFH) is characterized as bone cell death in the hip joint, involving a severe pain in the groin. The staging of ONFH is commonly based on Magnetic resonance imaging and computed tomography (CT), which are important for establishing effective treatment plans. There have been some attempts to automate ONFH staging using deep learning, but few of them used only MR images. PURPOSE: To propose a deep learning model for MR-only ONFH staging, which can reduce additional cost and radiation exposure from the acquisition of CT images. METHODS: We integrated information from the MR images of five different imaging protocols by a newly proposed attention fusion method, which was composed of intra-modality attention and inter-modality attention. In addition, a self-supervised learning was used to learn deep representations from a large amount of paired MR-CT dataset. The encoder part of the MR-CT translation network was used as a pretraining network for the staging, which aimed to overcome the lack of annotated data for staging. Ablation studies were performed to investigate the contributions of each proposed method. The area under the receiver operating characteristic curve (AUROC) was used to evaluate the performance of the networks. RESULTS: Our model improved the performance of the four-way classification of the association research circulation osseous (ARCO) stage using MR images of the multiple protocols by 6.8%p in AUROC over a plain VGG network. Each proposed method increased the performance by 4.7%p (self-supervised learning) and 2.6%p (attention fusion) in AUROC, which was demonstrated by the ablation experiments. CONCLUSIONS: We have shown the feasibility of the MR-only ONFH staging by using self-supervised learning and attention fusion. A large amount of paired MR-CT data in hospitals can be used to further improve the performance of the staging, and the proposed method has potential to be used in the diagnosis of various diseases that require staging from multiple MR protocols.

An agonistic anti-Tie2 antibody suppresses the normal-to-tumor vascular transition in the glioblastoma invasion zone.

Tumor progression is intimately associated with the vasculature, as tumor proliferation induces angiogenesis and tumor cells metastasize to distant organs via blood vessels. However, whether tumor invasion is associated with blood vessels remains unknown. As glioblastoma (GBM) is featured by aggressive invasion and vascular abnormalities, we characterized the onset of vascular remodeling in the diffuse tumor infiltrating zone by establishing new spontaneous GBM models with robust invasion capacity. Normal brain vessels underwent a gradual transition to severely impaired tumor vessels at the GBM periphery over several days. Increasing vasodilation from the tumor periphery to the tumor core was also found in human GBM. The levels of vascular endothelial growth factor (VEGF) and VEGF receptor 2 (VEGFR2) showed a spatial correlation with the extent of vascular abnormalities spanning the tumor-invading zone. Blockade of VEGFR2 suppressed vascular remodeling at the tumor periphery, confirming the role of VEGF-VEGFR2 signaling in the invasion-associated vascular transition. As angiopoietin-2 (ANGPT2) was expressed in only a portion of the central tumor vessels, we developed a ligand-independent tunica interna endothelial cell kinase 2 (Tie2)-activating antibody that can result in Tie2 phosphorylation in vivo. This agonistic anti-Tie2 antibody effectively normalized the vasculature in both the tumor periphery and tumor center, similar to the effects of VEGFR2 blockade. Mechanistically, this antibody-based Tie2 activation induced VE-PTP-mediated VEGFR2 dephosphorylation in vivo. Thus, our study reveals that the normal-to-tumor vascular transition is spatiotemporally associated with GBM invasion and may be controlled by Tie2 activation via a novel mechanism of action.

Tumor hypoxia represses γδ T cell-mediated antitumor immunity against brain tumors.

The anatomic location and immunologic characteristics of brain tumors result in strong lymphocyte suppression. Consequently, conventional immunotherapies targeting CD8 T cells are ineffective against brain tumors. Tumor cells escape immunosurveillance by various mechanisms and tumor cell metabolism can affect the metabolic states and functions of tumor-infiltrating lymphocytes. Here, we discovered that brain tumor cells had a particularly high demand for oxygen, which affected γδ T cell-mediated antitumor immune responses but not those of conventional T cells. Specifically, tumor hypoxia activated the γδ T cell protein kinase A pathway at a transcriptional level, resulting in repression of the activatory receptor NKG2D. Alleviating tumor hypoxia reinvigorated NKG2D expression and the antitumor function of γδ T cells. These results reveal a hypoxia-mediated mechanism through which brain tumors and γδ T cells interact and emphasize the importance of γδ T cells for antitumor immunity against brain tumors.

Feasibility of Synthetic Computed Tomography Images Generated from Magnetic Resonance Imaging Scans Using Various Deep Learning Methods in the Planning of Radiation Therapy for Prostate Cancer.

We aimed to evaluate and compare the qualities of synthetic computed tomography (sCT) generated by various deep-learning methods in volumetric modulated arc therapy (VMAT) planning for prostate cancer. Simulation computed tomography (CT) and T2-weighted simulation magnetic resonance image from 113 patients were used in the sCT generation by three deep-learning approaches: generative adversarial network (GAN), cycle-consistent GAN (CycGAN), and reference-guided CycGAN (RgGAN), a new model which performed further adjustment of sCTs generated by CycGAN with available paired images. VMAT plans on the original simulation CT images were recalculated on the sCTs and the dosimetric differences were evaluated. For soft tissue, a significant difference in the mean Hounsfield unites (HUs) was observed between the original CT images and only sCTs from GAN (p = 0.03). The mean relative dose differences for planning target volumes or organs at risk were within 2% among the sCTs from the three deep-learning approaches. The differences in dosimetric parameters for D98% and D95% from original CT were lowest in sCT from RgGAN. In conclusion, HU conservation for soft tissue was poorest for GAN. There was the trend that sCT generated from the RgGAN showed best performance in dosimetric conservation D98% and D95% than sCTs from other methodologies.

Artificial neural network for Slice Encoding for Metal Artifact Correction (SEMAC) MRI.

PURPOSE: To develop new artificial neural networks (ANNs) to accelerate slice encoding for metal artifact correction (SEMAC) MRI. METHODS: Eight titanium phantoms and 77 patients after brain tumor surgery involving metallic neuro-plating instruments were scanned using SEMAC at a 3T Skyra scanner. For the phantoms, proton-density, T1-, and T2-weighted images were acquired for developing both multilayer perceptron (MLP) and convolutional neural network (CNN). For the patients, T2-weighted images were acquired for developing CNN. All networks were trained with the SEMAC factor 4 or 6 as input and the factor 12 as label, yielding an acceleration factor of 3 or 2. Performance of the CNN model was compared against parallel imaging and compressed sensing on the phantom datasets. Two extra T1-weighted in vivo sets were acquired to investigate generalizability of the models to different contrasts. RESULTS: Both multilayer perceptron and CNN provided artifact-suppressed images better than the input images and comparable to the label images visually and quantitatively, a trend observable regardless of input SEMAC factor and image type (P < .01). CNN suppressed the artifacts better than multilayer perceptron, parallel imaging, and compressed sensing (P < .01). Tests on the patient datasets demonstrated clear metal artifact suppression visually and quantitatively (P < .01). Tests on T1 datasets also demonstrated clear visual metal artifact suppression. CONCLUSION: Our study introduced a new effective way of artificial neural networks to accelerate SEMAC MRI while maintaining the comparable quality of metal artifact suppression. Application on the preliminary patient datasets proved the feasibility in clinical usage, which warrants further investigation.

Meningeal lymphatic vessels at the skull base drain cerebrospinal fluid.

Recent work has shown that meningeal lymphatic vessels (mLVs), mainly in the dorsal part of the skull, are involved in the clearance of cerebrospinal fluid (CSF), but the precise route of CSF drainage is still unknown. Here we reveal the importance of mLVs in the basal part of the skull for this process by visualizing their distinct anatomical location and characterizing their specialized morphological features, which facilitate the uptake and drainage of CSF. Unlike dorsal mLVs, basal mLVs have lymphatic valves and capillaries located adjacent to the subarachnoid space in mice. We also show that basal mLVs are hotspots for the clearance of CSF macromolecules and that both mLV integrity and CSF drainage are impaired with ageing. Our findings should increase the understanding of how mLVs contribute to the neuropathophysiological processes that are associated with ageing.

Evaluation of Tumor Blood Flow Using Alternate Ascending/Descending Directional Navigation in Primary Brain Tumors: A Comparison Study with Dynamic Susceptibility Contrast Magnetic Resonance Imaging.

OBJECTIVE: Alternate ascending/descending directional navigation (ALADDIN) is a novel arterial spin labeling technique that does not require a separate spin preparation pulse. We sought to compare the normalized cerebral blood flow (nCBF) values obtained by ALADDIN and dynamic susceptibility contrast (DSC) perfusion magnetic resonance imaging (MRI) in patients with primary brain tumors. MATERIALS AND METHODS: Sixteen patients with primary brain tumors underwent MRI scans including contrast-enhanced T1-weighted imaging, DSC perfusion MRI, and ALADDIN. The nCBF values of normal gray matter (GM) and tumor areas were measured by both DSC perfusion MRI and ALADDIN, which were compared by the Wilcoxon signed rank test. Subgroup analyses according to pathology were performed with the Wilcoxon signed rank test. RESULTS: Higher mean nCBF values of GM regions in the bilateral frontal lobe, temporal lobe, and caudate were detected by ALADDIN than by DSC perfusion MRI (p <0.05). In terms of the mean or median nCBF values and the mean of the top 10% nCBF values from tumors, DSC perfusion MRI and ALADDIN did not statistically significantly differ either overall or in each tumor group. CONCLUSION: ALADDIN tended to detect higher nCBF values in normal GM, as well as higher perfusion portions of primary brain tumors, than did DSC perfusion MRI. We believe that the high perfusion signal on ALADDIN can be beneficial in lesion detection and characterization.

Human glioblastoma arises from subventricular zone cells with low-level driver mutations.

Glioblastoma (GBM) is a devastating and incurable brain tumour, with a median overall survival of fifteen months1,2. Identifying the cell of origin that harbours mutations that drive GBM could provide a fundamental basis for understanding disease progression and developing new treatments. Given that the accumulation of somatic mutations has been implicated in gliomagenesis, studies have suggested that neural stem cells (NSCs), with their self-renewal and proliferative capacities, in the subventricular zone (SVZ) of the adult human brain may be the cells from which GBM originates3-5. However, there is a lack of direct genetic evidence from human patients with GBM4,6-10. Here we describe direct molecular genetic evidence from patient brain tissue and genome-edited mouse models that show astrocyte-like NSCs in the SVZ to be the cell of origin that contains the driver mutations of human GBM. First, we performed deep sequencing of triple-matched tissues, consisting of (i) normal SVZ tissue away from the tumour mass, (ii) tumour tissue, and (iii) normal cortical tissue (or blood), from 28 patients with isocitrate dehydrogenase (IDH) wild-type GBM or other types of brain tumour. We found that normal SVZ tissue away from the tumour in 56.3% of patients with wild-type IDH GBM contained low-level GBM driver mutations (down to approximately 1% of the mutational burden) that were observed at high levels in their matching tumours. Moreover, by single-cell sequencing and laser microdissection analysis of patient brain tissue and genome editing of a mouse model, we found that astrocyte-like NSCs that carry driver mutations migrate from the SVZ and lead to the development of high-grade malignant gliomas in distant brain regions. Together, our results show that NSCs in human SVZ tissue are the cells of origin that contain the driver mutations of GBM.

Improving resolution of MR images with an adversarial network incorporating images with different contrast.

PURPOSE: The routine MRI scan protocol consists of multiple pulse sequences that acquire images of varying contrast. Since high frequency contents such as edges are not significantly affected by image contrast, down-sampled images in one contrast may be improved by high resolution (HR) images acquired in another contrast, reducing the total scan time. In this study, we propose a new deep learning framework that uses HR MR images in one contrast to generate HR MR images from highly down-sampled MR images in another contrast. MATERIALS AND METHODS: The proposed convolutional neural network (CNN) framework consists of two CNNs: (a) a reconstruction CNN for generating HR images from the down-sampled images using HR images acquired with a different MRI sequence and (b) a discriminator CNN for improving the perceptual quality of the generated HR images. The proposed method was evaluated using a public brain tumor database and in vivo datasets. The performance of the proposed method was assessed in tumor and no-tumor cases separately, with perceptual image quality being judged by a radiologist. To overcome the challenge of training the network with a small number of available in vivo datasets, the network was pretrained using the public database and then fine-tuned using the small number of in vivo datasets. The performance of the proposed method was also compared to that of several compressed sensing (CS) algorithms. RESULTS: Incorporating HR images of another contrast improved the quantitative assessments of the generated HR image in reference to ground truth. Also, incorporating a discriminator CNN yielded perceptually higher image quality. These results were verified in regions of normal tissue as well as tumors for various MRI sequences from pseudo k-space data generated from the public database. The combination of pretraining with the public database and fine-tuning with the small number of real k-space datasets enhanced the performance of CNNs in in vivo application compared to training CNNs from scratch. The proposed method outperformed the compressed sensing methods. CONCLUSIONS: The proposed method can be a good strategy for accelerating routine MRI scanning.

Improving Arterial Spin Labeling by Using Deep Learning.

Purpose To develop a deep learning algorithm that generates arterial spin labeling (ASL) perfusion images with higher accuracy and robustness by using a smaller number of subtraction images. Materials and Methods For ASL image generation from pair-wise subtraction, we used a convolutional neural network (CNN) as a deep learning algorithm. The ground truth perfusion images were generated by averaging six or seven pairwise subtraction images acquired with (a) conventional pseudocontinuous arterial spin labeling from seven healthy subjects or (b) Hadamard-encoded pseudocontinuous ASL from 114 patients with various diseases. CNNs were trained to generate perfusion images from a smaller number (two or three) of subtraction images and evaluated by means of cross-validation. CNNs from the patient data sets were also tested on 26 separate stroke data sets. CNNs were compared with the conventional averaging method in terms of mean square error and radiologic score by using a paired t test and/or Wilcoxon signed-rank test. Results Mean square errors were approximately 40% lower than those of the conventional averaging method for the cross-validation with the healthy subjects and patients and the separate test with the patients who had experienced a stroke (P < .001). Region-of-interest analysis in stroke regions showed that cerebral blood flow maps from CNN (mean ± standard deviation, 19.7 mL per 100 g/min ± 9.7) had smaller mean square errors than those determined with the conventional averaging method (43.2 ± 29.8) (P < .001). Radiologic scoring demonstrated that CNNs suppressed noise and motion and/or segmentation artifacts better than the conventional averaging method did (P < .001). Conclusion CNNs provided superior perfusion image quality and more accurate perfusion measurement compared with those of the conventional averaging method for generation of ASL images from pair-wise subtraction images. © RSNA, 2017.

Mapping blood flow directionality in the human brain.

Diffusion properties of tissue are often expressed on the basis of directional variance, i.e., diffusion tensor imaging. In comparison, common perfusion-weighted imaging such as arterial spin labeling yields perfusion in a scalar quantity. The purpose of this study was to test the feasibility of mapping cerebral blood flow directionality using alternate ascending/descending directional navigation (ALADDIN), a recently-developed arterial spin labeling technique with sensitivity to blood flow directions. ALADDIN was applied along 3 orthogonal directions to assess directional blood flow in a vector form and also along 6 equally-spaced directions to extract blood flow tensor matrix (P) based on a blood flow ellipsoid model. Tensor elements (eigenvalues, eigenvectors, etc) were calculated to investigate characteristics of the blood flow tensor, in comparison with time-of-flight MR angiogram. While the directions of the main eigenvectors were heterogeneous throughout the brain, regional clusters of blood flow directionality were reproducible across subjects. The technique could show heterogeneous blood flow directionality within and around brain tumor, which was different from that of the contralateral normal side. The proposed method is deemed to provide information of blood flow directionality, which has not been demonstrated before. The results warrant further studies to assess changes in the directionality map as a function of scan parameters, to understand the signal sources, to investigate the possibility of mapping local blood perfusion directionality, and to evaluate its usefulness for clinical diagnosis.

Inter-Slice Blood Flow and Magnetization Transfer Effects as A New Simultaneous Imaging Strategy.

The recent blood flow and magnetization transfer (MT) technique termed alternate ascending/descending directional navigation (ALADDIN) achieves the contrast using interslice blood flow and MT effects with no separate preparation RF pulse, thereby potentially overcoming limitations of conventional methods. In this study, we examined the signal characteristics of ALADDIN as a simultaneous blood flow and MT imaging strategy, by comparing it with pseudo-continuous ASL (pCASL) and conventional MT asymmetry (MTA) methods, all of which had the same bSSFP readout. Bloch-equation simulations and experiments showed ALADDIN perfusion signals increased with flip angle, whereas MTA signals peaked at flip angle around 45°-60°. ALADDIN provided signals comparable to those of pCASL and conventional MTA methods emulating the first, second, and third prior slices of ALADDIN under the same scan conditions, suggesting ALADDIN signals to be superposition of signals from multiple labeling planes. The quantitative cerebral blood flow signals from a modified continuous ASL model overestimated the perfusion signals compared to those measured with a pulsed ASL method. Simultaneous mapping of blood flow, MTA, and MT ratio in the whole brain is feasible with ALADDIN within a clinically reasonable time, which can potentially help diagnosis of various diseases.

Investigation of inter-slice magnetization transfer effects as a new method for MTR imaging of the human brain.

We present a new method for magnetization transfer (MT) ratio imaging in the brain that requires no separate saturation pulse. Interslice MT effects that are inherent to multi-slice balanced steady-state free precession (bSSFP) imaging were controlled via an interslice delay time to generate MT-weighted (0 s delay) and reference images (5-8 s delay) for MT ratio (MTR) imaging of the brain. The effects of varying flip angle and phase encoding (PE) order were investigated experimentally in normal, healthy subjects. Values of up to ∼50% and ∼40% were observed for white and gray matter MTR. Centric PE showed larger MTR, higher SNR, and better contrast between white and gray matter than linear PE. Simulations of a two-pool model of MT agreed well with in vivo MTR values. Simulations were also used to investigate the effects of varying acquisition parameters, and the effects of varying flip angle, PE steps, and interslice delay are discussed. Lastly, we demonstrated reduced banding with a non-balanced SSFP-FID sequence and showed preliminary results of interslice MTR imaging of meningioma.