Partial Volume Effects Research Articles

Semi-automated segmentation methods of SSTR PET for dosimetry prediction in refractory meningioma patients treated by SSTR-targeted peptide receptor radionuclide therapy.

Tumor dosimetry with somatostatin receptor-targeted peptide receptor radionuclide therapy (SSTR-targeted PRRT) by 177Lu-DOTATATE may contribute to improved treatment monitoring of refractory meningioma. Accurate dosimetry requires reliable and reproducible pretherapeutic PET tumor segmentation which is not currently available. This study aims to propose semi-automated segmentation methods to determine metabolic tumor volume with pretherapeutic 68Ga-DOTATOC PET and evaluate SUVmean-derived values as predictive factors for tumor-absorbed dose. Thirty-nine meningioma lesions from twenty patients were analyzed. The ground truth PET and SPECT volumes (VolGT-PET and VolGT-SPECT) were computed from manual segmentations by five experienced nuclear physicians. SUV-related indexes were extracted from VolGT-PET and the semi-automated PET volumes providing the best Dice index with VolGT-PET (Volopt) across several methods: SUV absolute-value (2.3)-threshold, adaptative methods (Jentzen, Otsu, Contrast-based method), advanced gradient-based technique, and multiple relative thresholds (% of tumor SUVmax, hypophysis SUVmean, and meninges SUVpeak) with optimal threshold optimized. Tumor-absorbed doses were obtained from the VolGT-SPECT, corrected for partial volume effect, performed on a 360° whole-body CZT-camera at 24, 96, and 168h after administration of 177Lu-DOTATATE. Volopt was obtained from 1.7-fold meninges SUVpeak (Dice index 0.85 ± 0.07). SUVmean and total lesion uptake (SUVmeanxlesion volume) showed better correlations with tumor-absorbed doses than SUVmax when determined with the VolGT (respective Pearson correlation coefficients of 0.78, 0.67, and 0.56) or Volopt (0.64, 0.66, and 0.56). Accurate definition of pretherapeutic PET volumes is justified since SUVmean-derived values provide the best tumor-absorbed dose predictions in refractory meningioma patients treated by 177Lu-DOTATATE. This study provides a semi-automated segmentation method of pretherapeutic 68Ga-DOTATOC PET volumes to achieve good reproducibility between physicians. SUVmean-derived values from pretherapeutic 68Ga-DOTATOC PET are predictive of tumor-absorbed doses in refractory meningiomas treated by 177Lu-DOTATATE, justifying to accurately define pretherapeutic PET volumes. This study provides a semi-automated segmentation of 68Ga-DOTATOC PET images easily applicable in routine. • SUVmean-derived values from pretherapeutic 68Ga-DOTATOC PET images provide the best predictive factors of tumor-absorbed doses related to 177Lu-DOTATATE PRRT in refractory meningioma. • A 1.7-fold meninges SUVpeak segmentation method used to determine metabolic tumor volume on pretherapeutic 68Ga-DOTATOC PET images of refractory meningioma treated by 177Lu-DOTATATE is as efficient as the currently routine manual segmentation method and limits inter- and intra-observer variabilities. • This semi-automated method for segmentation of refractory meningioma is easily applicable to routine practice and transferrable across PET centers.

Segmentation-Free PVC for Cardiac SPECT Using a Densely-Connected Multi-Dimensional Dynamic Network.

In nuclear imaging, limited resolution causes partial volume effects (PVEs) that affect image sharpness and quantitative accuracy. Partial volume correction (PVC) methods incorporating high-resolution anatomical information from CT or MRI have been demonstrated to be effective. However, such anatomical-guided methods typically require tedious image registration and segmentation steps. Accurately segmented organ templates are also hard to obtain, particularly in cardiac SPECT imaging, due to the lack of hybrid SPECT/CT scanners with high-end CT and associated motion artifacts. Slight mis-registration/mis-segmentation would result in severe degradation in image quality after PVC. In this work, we develop a deep-learning-based method for fast cardiac SPECT PVC without anatomical information and associated organ segmentation. The proposed network involves a densely-connected multi-dimensional dynamic mechanism, allowing the convolutional kernels to be adapted based on the input images, even after the network is fully trained. Intramyocardial blood volume (IMBV) is introduced as an additional clinical-relevant loss function for network optimization. The proposed network demonstrated promising performance on 28 canine studies acquired on a GE Discovery NM/CT 570c dedicated cardiac SPECT scanner with a 64-slice CT using Technetium-99m-labeled red blood cells. This work showed that the proposed network with densely-connected dynamic mechanism produced superior results compared with the same network without such mechanism. Results also showed that the proposed network without anatomical information could produce images with statistically comparable IMBV measurements to the images generated by anatomical-guided PVC methods, which could be helpful in clinical translation.

Energy minimization segmentation model based on MRI images.

Medical image segmentation is an important tool for doctors to accurately analyze the volume of brain tissue and lesions, which is important for the correct diagnosis of brain diseases. However, manual image segmentation methods are time-consuming, subjective and lack of repeatability, it needs to develop automatic and reliable methods for image segmentation. Magnetic Resonance Imaging (MRI), a non-invasive imaging technique, is commonly used to detect, characterize and quantify tissues and lesions in the brain. Partial volume effect, gray scale in homogeneity, and lesions presents a great challenge for automatic medical image segmentation methods. So, the paper is dedicated to address the impact of partial volume effect and multiple sclerosis lesions on the segmentation accuracy in MRI. The objective function of the improved model and the post-processing method of lesion filling are researched based on the fuzzy clustering space and energy model. In particular, an energy-minimized segmentation algorithm is proposed. Through experimental verification, the AR-FCM algorithm can better overcome the problem of low segmentation accuracy of the RFCM algorithm for tissue boundary voxels and improve the segmentation accuracy of this algorithm. Meanwhile, a multi-channel input energy-minimization segmentation method with lesion filling and anatomical mapping is further proposed. The feasibility of the lesion filling strategy using post-processing can be confirmed and the segmentation accuracy is increased by comparison experiments.

SEGMENTATION OF MRI IMAGES USING A COMBINATION OF ACTIVE CONTOUR MODELING AND MORPHOLOGICAL PROCESSING

Image segmentation in brain magnetic resonance imaging (MRI) largely relates to dividing brain tissue into components like white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF). Using the segmentation outputs, medical images can be 3D reconstructed and visualized efficiently. It is common for MRI pictures to have issues such as partial volume effects, asymmetrical grayscale, and noise. As a result, high accuracy in brain MRI picture segmentation is challenging to achieve in practical applications. In this paper, we developed an effective algorithm for brain MRI image segmentation utilizing a combination of statistical and partial differential equation-based approaches, based on a neuro-mechanical model. The findings of this work demonstrate that by combining various segmentation approaches, it is possible to quickly segment brain MRI data at a degree of precision necessary for different applications. Here, we show that when we use nonlinear filtering, [Formula: see text]-means clustering, and active contour modeling, we can get very good results when we segment brain MRI images. It is clear that the proposed approach has higher segmentation performance and can properly separate brain tissue from a large number of MRI images.

Abstract 2767: Simultaneous 18F-FDG PET/MR imaging for metastases identification in a disseminated human preclinical melanoma model

Abstract Melanoma is a skin cancer at high risk of metastatic progression, hence associated with poorer prognosis. Imaging represents an essential tool to assess tumor stage, but also to monitor disease progression and therapy response. However, a single imaging technique cannot efficiently detect and characterize all metastases. Therefore, we evaluated the potential of simultaneous 18F-FDG PET/MR imaging in preclinical models of disseminated melanoma at different stages. Female Nude and SDRG rats were intravenously injected with CMEL 5 cells, originating from brain metastases of a human melanoma. On day (D) 60, 70, and 80 post-injection, Nude rats were imaged using brain T2-weighted (T2w) 7T MRI and whole-body simultaneous 18F-FDG (10-15 MBq) PET/MR. Furthermore, whole-body and brain T2w MR images were acquired using a 4.7T MRI on D20 and D34 post-injection in SDRG rats. Visual detection and quantitative analysis of lesions were performed on MRI or PET/MR images, and results were confronted with necropsy (melanin pigmented lesions) and gamma counting data. From D60 post-CMEL-5 cell injection, all Nude animals exhibited metastases in one to four different locations. Moreover, most of metastases locations found in clinical setting were detected in our experimental model using simultaneous 18F-FDG PET/MR imaging, even allowing the observation of brain lesions on T2w MR images. MRI further allowed precise contouring of each organ of interest to effectively identify bone, lung, adrenal and spleen metastases with enhanced 18F-FDG uptake. Additionally, PET image quantification in adrenals was consistent with gamma counting results. As anticipated, small lesions below PET spatial resolution limit (i.e. <1 mm) were not detected by imaging. SDRG rats displayed a faster and more homogenous tumor spread in brain, lungs and liver on MR images, while more metastases were seen during autopsy (additional spreading to pancreas, adrenals, reproductive system, bones, lymph nodes and skin). The experimental model of disseminated CMEL-5 melanoma in Nude rats demonstrates its ability to mimic the human pathology. Hence, it appears to be a promising tool to monitor therapy response and potentially recognize progression from pseudo-progression. Simultaneous PET/MR imaging has shown to be more effective than stand-alone techniques in detecting melanoma dissemination, except for brain and small lesions for which the PET data are altered by partial volume effect. Thus, by complementing previously acquired MR images from the novel SDRG rat model with simultaneously acquired PET images, we could also provide insights on this promising metastatic model. Citation Format: Alexandra Oudot, Peggy Provent, Alan Courteau, Nicolas Hoffmann, Melanie Guillemin, Jordan Longin, Sarah Belderbos, Bertrand Collin, Marc Hillairet de Boisferon, Alexandre Cochet, Cyril Berthet. Simultaneous 18F-FDG PET/MR imaging for metastases identification in a disseminated human preclinical melanoma model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2767.

Deep Learning-Based Algorithm for Automatic Detection of Pulmonary Embolism in Chest CT Angiograms

Since the prompt recognition of acute pulmonary embolism (PE) and the immediate initiation of treatment can significantly reduce the risk of death, we developed a deep learning (DL)-based application aimed to automatically detect PEs on chest computed tomography angiograms (CTAs) and alert radiologists for an urgent interpretation. Convolutional neural networks (CNNs) were used to design the application. The associated algorithm used a hybrid 3D/2D UNet topology. The training phase was performed on datasets adequately distributed in terms of vendors, patient age, slice thickness, and kVp. The objective of this study was to validate the performance of the algorithm in detecting suspected PEs on CTAs. The validation dataset included 387 anonymized real-world chest CTAs from multiple clinical sites (228 U.S. cities). The data were acquired on 41 different scanner models from five different scanner makers. The ground truth (presence or absence of PE on CTA images) was established by three independent U.S. board-certified radiologists. The algorithm correctly identified 170 of 186 exams positive for PE (sensitivity 91.4% [95% CI: 86.4-95.0%]) and 184 of 201 exams negative for PE (specificity 91.5% [95% CI: 86.8-95.0%]), leading to an accuracy of 91.5%. False negative cases were either chronic PEs or PEs at the limit of subsegmental arteries and close to partial volume effect artifacts. Most of the false positive findings were due to contrast agent-related fluid artifacts, pulmonary veins, and lymph nodes. The DL-based algorithm has a high degree of diagnostic accuracy with balanced sensitivity and specificity for the detection of PE on CTAs.

ResAttn-recon: Residual self-attention based cortical surface reconstruction

Introduction: The accurate cerebral cortex surface reconstruction is crucial for the study of neurodegenerative diseases. Existing voxelwise segmentation-based approaches like FreeSurfer and FastSurfer are limited by the partial volume effect, meaning that reconstruction details highly rely on the resolution of the input volume. In the computer version area, the signed distance function has become an efficient method for 3D shape representation, the inherent continuous nature makes it easy to capture the fine details of the target object at an arbitrary resolution. Additionally, as one of the most valuable breakthroughs in deep learning research, attention is a powerful mechanism developed to enhance the performance of the encoder-decoder architecture.Methods: To further improve the reconstruction accuracy of the cortical surface, we proposed ResAttn-Recon, a residual self-attention based encoder-decoder framework. In this framework, we also developed a lightweight decoder network with skip connections. Furthermore, a truncated and weighted L1 loss function are proposed to accelerate network convergence, compared to simply applying the L1 loss function.Results: The intersection over union curve in the training process achieved a steeper slope and a higher peak (0.948 vs. 0.920) with a truncated L1 loss. Thus, the average symmetric surface distance (AD) for the inner and outer surfaces is 0.253 ± 0.051 and the average Hausdorff distance (HD) is 0.629 ± 0.186, which is lower than that of DeepCSR, whose absolute distance equals 0.283 ± 0.059 and Hausdorff distance equals 0.746 ± 0.245.Discussion: In conclusion, the proposed residual self-attention-based framework can be a promising approach for improving the cortical surface reconstruction performance.

Optimization of the Algorithm for the Implementation of Point Spread Function in the 3D-OSEM Reconstruction Algorithm Based on the List-Mode Micro PET Data

Positron emission tomography (PET) is a popular research topic. People are becoming more interested in PET images as they become more widely available. However, the partial volume effect (PVE) in PET images remains one of the most influential factors causing the resolution of PET images to degrade. It is possible to reduce this PVE and achieve better image quality by measuring and modeling the point spread function (PSF) and then accounting for it inside the reconstruction algorithm. In this work, we examined the response characteristics of the MetisTM PET/CT system by acquiring 22Na point source at different locations in the field of view (FOV) of the scanner and reconstructing with small pixel size for images to obtain their radial, tangential, and axial full-width half maximum (FWHM). An image-based model of the PSF model was then obtained by fitting asymmetric two-dimensional Gaussians on the 22Na images. This PSF model determined by FWHM in three directions was integrated into a three-dimensional ordered subsets expectation maximization (3D-OSEM) algorithm based on a list-mode format to form a new PSF-OSEM algorithm. We used both algorithms to reconstruct point source, Derenzo phantom, and mouse PET images and performed qualitative and quantitative analyses. In the point source study, the PSF-OSEM algorithm reduced the FWHM of the point source PET image in three directions to about 0.67 mm, and in the phantom study, the PET image reconstructed by the PSF-OSEM algorithm had better visual effects. At the same time, the quantitative analysis results of the Derenzo phantom were better than the original 3D-OSEM algorithm. In the mouse experiment, the results of qualitative and quantitative analyses showed that the imaging quality of PSF-OSEM algorithm was better than that of 3D-OSEM algorithm. Our results show that adding the PSF model to the 3D-OSEM algorithm in the MetisTM PET/CT system helps to improve the resolution of the image and satisfy the qualitative and quantitative analysis criteria.

Blood-brain barrier water exchange measurements using FEXI: Impact of modeling paradigm and relaxation timeeffects.

To evaluate potential modeling paradigms and the impact of relaxation time effects on human blood-brain barrier (BBB) water exchange measurements using FEXI (BBB-FEXI), and to quantify the accuracy, precision, and repeatability of BBB-FEXI exchange rate estimates at 3 . Three modeling paradigms were evaluated: (i) the apparent exchange rate (AXR) model; (ii) a two-compartment model ( ) explicitly representing intra- and extravascular signal components, and (iii) a two-compartmentmodeladditionally accounting for finite compartmental and relaxation times ( ). Each model had three free parameters. Simulations quantified biases introduced by the assumption of infinite relaxation times in the AXR and models, as well as the accuracy and precision of all three models. The scan-rescan repeatability of all paradigms was quantified for the first time in vivo in 10 healthy volunteers (age range 23-52 years; five female). The assumption of infinite relaxation times yielded exchange rate errors in simulations up to 42%/14% in the AXR/ models, respectively. Accuracy was highest in the compartmental models; precision was best in the AXR model. Scan-rescan repeatability in vivo was good for all models, with negligible bias and repeatability coefficients in grey matter of , , and . Compartmental modelling of BBB-FEXI signals can provide accurate and repeatable measurements of BBB water exchange; however, relaxation time and partial volume effects may cause model-dependent biases.

Estimation of free water-corrected microscopic fractional anisotropy.

Water diffusion anisotropy MRI is sensitive to microstructural changes in the brain that are hallmarks of various neurological conditions. However, conventional metrics like fractional anisotropy are confounded by neuron fiber orientation dispersion, and the relatively low resolution of diffusion-weighted MRI gives rise to significant free water partial volume effects in many brain regions that are adjacent to cerebrospinal fluid. Microscopic fractional anisotropy is a recent metric that can report water diffusion anisotropy independent of neuron fiber orientation dispersion but is still susceptible to free water contamination. In this paper, we present a free water elimination (FWE) technique to estimate microscopic fractional anisotropy and other related diffusion indices by implementing a signal representation in which the MRI signal within a voxel is assumed to come from two distinct sources: a tissue compartment and a free water compartment. A two-part algorithm is proposed to rapidly fit a set of diffusion-weighted MRI volumes containing both linear- and spherical-tensor encoding acquisitions to the representation. Simulations and in vivo acquisitions with four healthy volunteers indicated that the FWE method may be a feasible technique for measuring microscopic fractional anisotropy and other indices with greater specificity to neural tissue characteristics than conventional methods.

A PET-CT study on neuroinflammation in Huntington's disease patients participating in a randomized trial with laquinimod.

Microglia activation, an indicator of central nervous system inflammation, is believed to contribute to the pathology of Huntington's disease. Laquinimod is capable of regulating microglia. By targeting the translocator protein, 11C-PBR28 PET-CT imaging can be used to assess the state of regional gliosis in vivo and explore the effects of laquinimod treatment. This study relates to the LEGATO-HD, multi-centre, double-blinded, Phase 2 clinical trial with laquinimod (US National Registration: NCT02215616). Fifteen patients of the UK LEGATO-HD cohort (mean age: 45.2 ± 7.4 years; disease duration: 5.6 ± 3.0 years) were treated with laquinimod (0.5 mg, N = 4; 1.0 mg, N = 6) or placebo (N = 5) daily. All participants had one 11C-PBR28 PET-CT and one brain MRI scan before laquinimod (or placebo) and at the end of treatment (12 months apart). PET imaging data were quantified to produce 11C-PBR28 distribution volume ratios. These ratios were calculated for the caudate and putamen using the reference Logan plot with the corpus callosum as the reference region. Partial volume effect corrections (Müller-Gartner algorithm) were applied. Differences were sought in Unified Huntington's Disease Rating Scale scores and regional distribution volume ratios between baseline and follow-up and between the two treatment groups (laquinimod versus placebo). No significant change in 11C-PBR28 distribution volume ratios was found post treatment in the caudate and putamen for both those treated with laquinimod (N = 10) and those treated with placebo (N = 5). Over time, the patients treated with laquinimod did not show a significant clinical improvement. Data from the 11C-PBR28 PET-CT study indicate that laquinimod may not have affected regional translocator protein expression and clinical performance over the studied period.

Abstract No. 133 Fabrication and Characterization of Highly Radiopaque Microspheres for Prostate Cancer Chemoembolization

The purpose of this study was to develop and characterize embolic microspheres that are highly radiopaque, such that they could be used for prostate cancer chemo-embolization. The goals are that the spheres can be visualized during injection on fluoroscopic imaging without additional contrast media (with reasonable injection force), and that they can be imaged (with C-arm CT or MSCT) within the prostate and other non-target organs, both during and after the procedure. Finally, these spheres should be loadable with long-release chemotherapeutic agents. A process was developed to manufacture polyethylene glycol (PEG) microspheres where one of two radiopaque agents were added during the manufacturing process. The process was based on droplet generation microfluidics using an olive oil base. The first type had 1 um barium sulfate particles added at 40% wbv. The second type had 100 nm zirconium oxide particles added at 50% wbv. After curing, particles were washed to remove the olive oil. Sphere diameter was measured using light microscopy. The spheres' radiopacity were evaluated individually using microCT, and in aggregate, within a 3D printed vascular phantom, using MSCT. Injection force from a syringe was measured. Finally, the spheres were imaged with x-ray fluoroscopy during injection, using a body-sized, water-filled, vascular phantom. The barium and zirconium spheres measured 111 ± 14 um and 73 ± 18 um respectively. MicroCT measurements of individual sphere radiopacity were 6772 ± 501 HU and 13,486 ± 638 HU. MSCT radiopacity values in a 0.5 mm vessel were 264 HU and 226 HU. The measured injection force from a 3-cc syringe was 20 N, compared with 25 N for commercial microspheres with 50% contrast agent. The detectability on the microspheres on fluoroscopy was very similar to that of commercial microspheres with 50% contrast agent, approximately a 30 pixel-value enhancement over the background value. This project demonstrated the ability to produce highly radiopaque microspheres that have reasonable injection force and are visible both during injection using fluoroscopy, without additional contrast agent, and using MSCT. The HU values in MSCT are lower that microCT due to the lower tube voltage in microCT and also due to the limited spatial resolution (partial volume effect) of MSCT, but these small vessels are nevertheless visible within a 0-100 HU prostate gland. Further work is underway to evaluate drug elution characteristics.

A conformal TOF-DOI Prism-PET prototype scanner for high-resolution quantitative neuroimaging.

Positron emission tomography (PET) has had a transformative impact on oncological and neurological applications. However, still much of PET's potential remains untapped with limitations primarily driven by low spatial resolution, which severely hampers accurate quantitative PET imaging via the partial volume effect (PVE). We present experimental results of a practical and cost-effective ultra-high resolution brain-dedicated PET scanner, using our depth-encoding Prism-PET detectors arranged along a compact and conformal gantry, showing substantial reduction in PVE and accurate radiotracer uptake quantification in small regions. The decagon-shaped prototype scanner has a long diameter of 38.5cm, a short diameter of 29.1 cm, and an axial field-of-view (FOV) of 25.5mm with a single ring of 40 Prism-PET detector modules. Each module comprises a 16 × 16 array of 1.5 × 1.5 × 20-mm3 lutetium yttrium oxyorthosillicate (LYSO) scintillator crystals coupled 4-to-1 to an 8 × 8 array of silicon photomultiplier (SiPM) pixels on one end and to a prismatoid light guide array on the opposite end. The scanner's performance was evaluated by measuring depth-of-interaction (DOI) resolution, energy resolution, timing resolution, spatial resolution, sensitivity, and image quality of ultra-micro Derenzo and three-dimensional (3D) Hoffman brainphantoms. The full width at half maximum (FWHM) DOI, energy, and timing resolutions of the scanner are 2.85 mm, 12.6%, and 271 ps, respectively. Not considering artifacts due to mechanical misalignment of detector blocks, the intrinsic spatial resolution is 0.89-mm FWHM. Point source images reconstructed with 3D filtered back-projection (FBP) show an average spatial resolution of 1.53-mm FWHM across the entire FOV. The peak absolute sensitivity is 1.2% for an energy window of 400-650 keV. The ultra-micro Derenzo phantom study demonstrates the highest reported spatial resolution performance for a human brain PET scanner with perfect reconstruction of 1.00-mm diameter hot-rods. Reconstructed images of customized Hoffman brain phantoms prove that Prism-PET enables accurate radiotracer uptake quantification in small brain regions (2-3 mm). Prism-PET will substantially strengthen the utility of quantitative PET in neurology for early diagnosis of neurodegenerative diseases, and in neuro-oncology for improved management of both primary and metastatic brain tumors.

3D Quantitative Mineral Characterization of Particles Using X-ray Computed Tomography

A new method to measure and quantify the 3D mineralogical composition of particulate materials using X-ray computed micro-tomography (CT) is presented. The new method is part of a workflow designed to standardize the analysis of particles based on their microstructures without the need to segment the individual classes or grains. Classification follows a decision tree with criteria derived from particle histogram parameters that are specific to each microstructure, which in turn can be identified by 2D-based automated quantitative mineralogy. The quantification of mineral abundances is implemented at the particle level according to the complexity of the particle by taking into consideration the partial volume effect at interphases. The new method was tested on two samples with different particle size distributions from a carbonate rock containing various microstructures and phases. The method allowed differentiation and quantification of more mineral classes than traditional 3D image segmentation that uses only the grey-scale for mineral classification. Nevertheless, due to lower spatial resolution and lack of chemical information, not all phases identified in 2D could be distinguished. However, quantification of the mineral classes that could be distinguished was more representative than their 2D quantification, especially for coarser particle sizes and for minor phases. Therefore, the new 3D method shows great potential as a complement to 2D-based methods and as an alternative to traditional phase segmentation analysis of 3D images. Particle-based quantification of mineralogical and 3D geometrical properties of particles opens new applications in the raw materials and particle processing industries.

Radiological and audiological predictors of stapes destruction in adherent pars tensa.

Progressive adherent pars tensa occasionally induces ossicular erosion. Specifically, stapes discontinuity adversely affects postoperative hearing. However, this irretrievable sequela is challenging to prove preoperatively, partly because perimatrix inflammation on the pars tensa can obscure the visibility of the ossicles or the partial volume effect of computed tomography (CT) imaging can hamper detailed ossicular visualization. Therefore, there is no consensus regarding the ideal timing for switching from a wait-and-see approach to a surgical one. Herein, we aimed to explore the potential predictors of stapes superstructure destruction in adherent pars tensa. This retrospective cohort study enrolled consecutive patients who underwent primary tympanoplasty for adherent pars tensa categorized as grade IV on Sadé's grading scale between April 2016 and September 2021. The impact of features on otoscopy and CT and air-bone gap (ABG) on stapes superstructure destruction was assessed using uni- and multivariable logistic regression analyses. Sixty-four ears were included. Multivariate analysis revealed the presence of debris on the adherent pars tensa (odds ratio [OR] [95% confidence interval {CI}]): 4.799 [1.063-21.668], p = 0.0415), presence of soft-tissue density occupying the oval window (OR [95% CI]: 13.876 [3.084-62.437], p = 0.0006), and a ≥ 20-dB preoperative ABG at 3kHz (OR [95% CI]: 7.595 [1.596-36.132], p = 0.0108) as independent predictors for stapes superstructure destruction. High preoperative awareness of the possibility of destruction of the stapes superstructure would enable the surgeon to make a timely decision to provide surgical intervention before progression to severe stapes destruction, thereby maintaining long-term satisfactory hearing.

Feasibility of joint mapping of triglyceride saturation and water longitudinal relaxation in a single breath hold applied to high fat-fraction adipose depots in the periclavicular anatomy

IntroductionSimultaneous mapping of triglyceride (TAG) saturation and tissue water relaxation may improve the characterization of the structure and function of anatomies with significant adipose tissue. While several groups have demonstrated in vivo TAG saturation imaging using MRI, joint mapping of relaxation and TAG saturation is understudied. Such mappings may avoid bias from physiological motion, if they can be done within a single breath-hold, and also account for static and applied magnetic field heterogeneity. MethodsWe propose a transient-state/MR fingerprinting single breath-hold sequence at 3 T, a low-rank reconstruction, and a parameter estimation pipeline that jointly estimates the number of double bonds (NDB), number of methylene interrupted double bonds (NMIDB), and tissue water T1, while accounting for non-ideal radiofrequency transmit scaling and off-resonance effects. We test the proposed method in simulations, in phantom against MR spectroscopy (MRS), and in vivo regions in and around high fat fraction (FF) periclavicular adipose tissue. Partial volume and multi-peak transverse relaxation effects are explored. ResultsThe simulation results demonstrate accurate NDB, NMIDB, and water T1 estimates across a range of NDB, NMIDB, and T1 values. In phantoms, the proposed method's estimates of NDB and NMIDB correlate with those from MR spectroscopy (Pearson correlation ≥0.98), while the water T1 estimates are concordant with a standard phantom. The NDB and NMIDB are sensitive to partial volumes of water, showing increasing bias at FF < 40%. This bias is found to be due to noise and transverse relaxation effects. The in vivo periclavicular adipose tissue has high FF (>90%). The adipose tissue NDB and NMIDB, and muscle T1 estimates are comparable to those reported in the literature. ConclusionRobust estimation of NDB, NMIDB at high FF and water T1 across a broad range of FFs are feasible using the proposed methods. Further reduction of noise and model bias are needed to employ the proposed technique in low FF anatomies and pathologies.

Machine learning predictive performance evaluation of conventional and fuzzy radiomics in clinical cancer imaging cohorts

BackgroundHybrid imaging became an instrumental part of medical imaging, particularly cancer imaging processes in clinical routine. To date, several radiomic and machine learning studies investigated the feasibility of in vivo tumor characterization with variable outcomes. This study aims to investigate the effect of recently proposed fuzzy radiomics and compare its predictive performance to conventional radiomics in cancer imaging cohorts. In addition, lesion vs. lesion+surrounding fuzzy and conventional radiomic analysis was conducted.MethodsPreviously published 11C Methionine (MET) positron emission tomography (PET) glioma, 18F-FDG PET/computed tomography (CT) lung, and 68GA-PSMA-11 PET/magneto-resonance imaging (MRI) prostate cancer retrospective cohorts were included in the analysis to predict their respective clinical endpoints. Four delineation methods including manually defined reference binary (Ref-B), its smoothed, fuzzified version (Ref-F), as well as extended binary (Ext-B) and its fuzzified version (Ext-F) were incorporated to extract imaging biomarker standardization initiative (IBSI)-conform radiomic features from each cohort. Machine learning for the four delineation approaches was performed utilizing a Monte Carlo cross-validation scheme to estimate the predictive performance of the four delineation methods.ResultsReference fuzzy (Ref-F) delineation outperformed its binary delineation (Ref-B) counterpart in all cohorts within a volume range of 938–354987 mm3 with relative cross-validation area under the receiver operator characteristics curve (AUC) of +4.7–10.4. Compared to Ref-B, the highest AUC performance difference was observed by the Ref-F delineation in the glioma cohort (Ref-F: 0.74 vs. Ref-B: 0.70) and in the prostate cohort by Ref-F and Ext-F (Ref-F: 0.84, Ext-F: 0.86 vs. Ref-B: 0.80). In addition, fuzzy radiomics decreased feature redundancy by approx. 20%.ConclusionsFuzzy radiomics has the potential to increase predictive performance particularly in small lesion sizes compared to conventional binary radiomics in PET. We hypothesize that this effect is due to the ability of fuzzy radiomics to model partial volume effects and delineation uncertainties at small lesion boundaries. In addition, we consider that the lower redundancy of fuzzy radiomic features supports the identification of imaging biomarkers in future studies. Future studies shall consider systematically analyzing lesions and their surroundings with fuzzy and binary radiomics.

Spatial normalization discrepancies between native and MNI152 brain template scans in gamma ventral capsulotomy patients

In neurosurgery, spatial normalization emerged as a tool to minimize inter-subject variability and study target point locations based on standard coordinates. The Montreal Neurological Institute's 152 brain template (MNI152) has become the most widely utilized in neuroimaging studies, but has been noted to introduce partial volume effects, distortions, and increase structure size in all directions (x/y/z axes). These discrepancies question the accuracy of the MNI template, as well as its utility for studies that examine and form conclusions from group-level data. Given that surgical precision in obsessive-compulsive disorder is essential to patient outcomes, we retrospectively investigated lesion size and location in patients (n = 21) who underwent capsulotomy for intractable OCD, comparing deviations in the native scans to those in standard space. MNI measurements were significantly larger than native measurements across several structures in both coronal and axial slices, and we found that MNI transformation increases the size of many subcortical structures in a significant and proportional way for both females and males. These findings urge caution when using MNI as a reference space, as well as a stronger consideration of population-specific brain templates when examining connectivity-based networks.

Partial volume effect in SPECT & PET imaging and impact on radionuclide dosimetry estimates.

The spatial resolution of emission tomographic imaging systems can lead to a significant underestimation in the apparent radioactivity concentration in objects of size comparable to the resolution volume of the system. The aim of this study was to investigate the impact of the partial volume effect (PVE) on clinical imaging in PET and SPECT with current state-of-the-art instrumentation and the implications that this has for radionuclide dosimetry estimates. Using the IEC Image Quality Phantom we have measured the underestimation in observed uptake in objects of various sizes for both PET and SPECT imaging conditions. Both single pixel measures (i.e., SUVmax) and region of interest mean values were examined over a range of object sizes. We have further examined the impact of the PVE on dosimetry estimates in OLINDA in 177Lu SPECT imaging based on a subject with multiple somatostatin receptor positive paragangliomas in the head and neck. In PET, single pixel estimates of uptake are affected for objects less than approximately 18 mm in minor axis with existing systems. In SPECT imaging with medium energy collimators (e.g., for 177Lu imaging), however, the underestimates are far greater, where single pixel estimates in objects less than 2-3×the resolution volume are significantly impacted. In SPECT, region of interest mean values are underestimated in objects less than 10 cm in diameter. In the clinical case example, the dosimetry measured with SPECT ranged from more than 60% underestimate in the largest lesion (28×22 mm in maximal cross-section; 10.2 cc volume) to >99% underestimate in the smallest lesion (4×5 mm; 0.06 cc). The partial volume effect remains a significant factor when estimating radionuclide uptake in vivo, especially in small volumes. Accurate estimates of absorbed dose from radionuclide therapy will be particularly challenging until robust solutions to correct for the PVE are found.

Differential effect of dementia etiology on cortical stiffness as assessed by MR elastography

BackgroundAging and dementia involve the disruption of brain molecular pathways leading to the alterations in tissue composition and gross morphology of the brain. Phenotypic and biomarker overlap between various etiologies of dementia supports a need for new modes of information to more accurately distinguish these disorders. Brain mechanical properties, which can be measured noninvasively by MR elastography, represent one understudied feature that are sensitive to neurodegenerative processes. In this study, we used two stiffness estimation schemes to test the hypothesis that different etiologies of dementia are associated with unique patterns of mechanical alterations across the cerebral cortex. MethodsMR elastography data were acquired for six clinical groups including amyloid-negative cognitively unimpaired (CU), amyloid-positive cognitively unimpaired (A + CU), amyloid-positive participants with mild cognitive impairment (A + MCI), amyloid-positive participants with Alzheimer’s clinical syndrome (A + ACS), dementia with Lewy bodies (DLB), and frontotemporal dementia (FTD). Stiffness maps were computed using two neural network inversions with the objective to at least partially separate the parenchyma-specific and morphological effects of neurodegeneration on mechanical property estimates. A tissue-confined inversion algorithm was designed to obtain the best estimate of stiffness in the brain parenchyma itself, while a regionally-aware inversion algorithm was used to measure the tissue stiffness along with the surroundings. Mean stiffness of 15 bilateral gray matter cortical regions were considered for statistical analysis. First, we tested the hypothesis that cortical stiffness changes in the aging brain. Next, we tested the overall study hypothesis by first comparing stiffness in each clinical group to the CU group, and then comparing the clinical groups against one another. Finally, we assessed the spatial and statistical overlap between atrophy and stiffness changes for both inversions. ResultsCortical brain regions become softer with age for both inversions with larger effects observed using regionally-aware stiffness. Stiffness decreases in the range 0.010–0.027 kPa per year were observed. Pairwise comparisons of each clinical group with cognitively unimpaired participants demonstrated 5 statistically significant differences in stiffness for tissue-confined measurements and 19 statistically different stiffness changes for the regionally-aware stiffness measurements. Pairwise comparisons between clinical groups further demonstrated unique patterns of stiffness differences. Analysis of the atrophy-versus-stiffness relationship showed that regionally-aware stiffness measurements exhibit higher sensitivity to neurodegeneration with findings that are not fully explained by partial volume effects or atrophy. ConclusionsBoth tissue-confined and regionally-aware stiffness estimates exhibited unique and complementary stiffness differences in various etiologies of dementia. Our results suggest that mechanical alterations measured by MRE reflect both tissue-specific differences as well as environmental effects. Multi-inversion schemes in MRE may provide new insights into the relationships between neuropathology and brain biomechanics.