Method For Tissue Characterization Research Articles

Intraoperative characterization of cardiac tissue: the potential of light scattering spectroscopy.

Damage to the cardiac conduction system remains one of the most significant risks associated with surgical interventions to correct congenital heart disease. This work demonstrates how light-scattering spectroscopy (LSS) can be used to non-destructively characterize cardiac tissue regions. To present an approach for associating tissue composition information with location-specific LSS data and further evaluate an LSS and machine learning system as a method for non-destructive tissue characterization. A custom LSS probe was used to gather spectral data from locations across 14 excised human pediatric nodal tissue samples (8 sinus nodes, 6 atrioventricular nodes). The LSS spectra were used to train linear and neural-network-based regressor models to predict tissue composition characteristics derived from the 3D models. Nodal tissue region nuclear densities were reported. A linear model trained to regress nuclear density from spectra achieved a prediction r-squared of 0.64 and a concordance correlation coefficient of 0.78. These methods build on previous studies suggesting that LSS measurements combined with machine learning signal processing can provide clinically relevant cardiac tissue composition.

Validity of dynamic contrast-enhanced magnetic resonance imaging of the breast versus diffusion-weighted imaging and magnetic resonance spectroscopy in predicting the malignant nature of non-mass enhancement lesions

BackgroundBreast cancer is the commonest cancer affecting women worldwide. So, it is important to accurately detect and classify different breast lesions. Noninvasive methods for tissue characterization have increased interest, particularly for early diagnosis. Non-mass enhancement (NME) breast lesions are described in magnetic resonance imaging (MRI) as the presence of enhancement without space-occupying lesions. Several studies have described that certain characteristics can be used as new indicators of malignancy in breast NME lesions. We aimed to study the role of multiparametric-MRI (Mp-MRI) as diffusion-weighted imaging (DWI) and magnetic resonance spectroscopy (MRS) in assessment of NME lesions and to suggest which one offers the greatest diagnostic accuracy.MethodsThis retrospective study was conducted from March 2017 to December 2023 on 220 NME breast lesions. All lesions were analyzed to study the features of benign and malignant NME lesions using different MRI techniques including dynamic contrast-enhanced MRI (DCE-MRI), DWI, and MRS. Breast MRI was performed at 1.5 Tesla, findings were correlated with histopathological results of all cases.ResultsPatients’ mean age was 46.56 years with 220 NME breast lesions (54 were benign and 166 were malignant). Invasive ductal carcinoma with ductal carcinoma in situ was the most malignant type representing 93 cases. We found that segmental distribution, heterogeneous enhancement, type III curve, restricted diffusion, lower apparent diffusion coefficient, and positive choline peak were more with malignancy (P = 0.008, 0.02, 0.004, 0.001, and < 0.001). We detected that Mp-MRI has higher diagnostic accuracy than DCE-MRI and combined other functional sequences (DWI, MRS), it was 91.2% with sensitivity 89.9%, specificity 87.8%, positive predictive value 89.2%, and negative predictive value 82.2%.ConclusionsFunctional MRI techniques, such as DWI and MRS, can provide helpful information in assessment of NME lesions. They have high diagnostic accuracy, sensitivity, and specificity in characterizing NME breast lesions as benign or malignant. However, DCE-MRI is mandatory for lesion characterization and delineation of its nature and cannot be replaced by them alone in cases of lesion visualization. So, multiparametric-MRI can improve the diagnostic accuracy of NME breast lesions when combined with dynamic contrast-enhanced MRI and can help in reducing negative biopsy rates.

Parametric imaging based on horizontally normalized weight-adjustable Shannon entropy for tissue characterization

Ultrasonic B-mode imaging provides real-time and non-invasive imaging for soft tissue diagnosis in clinical use, but its limited contrast leads to the challenge of detection accuracy. Quantitative ultrasound techniques have been proposed as a promising method for soft tissue characterization and reflecting the microstructure of lesions. This study proposed a novel entropy called horizontally normalized weight-adjustable Shannon entropy (hNWASE). An adjustable weight factor was added to this entropy, so that the entropy value can be changed and the imaging performance can be adjusted to lesions according to different positions and acoustic characteristics. The relationship between the adjustable parameter n and the imaging performance was explored. Moreover, the performance of the proposed hNWASE imaging was compared with weighted Shannon entropy (WSE) imaging, and horizontally normalized Shannon entropy (hNSE) imaging by both simulations and clinical data. hNSE imaging obtained a Matthews correlation coefficient (MCC) of 0.68 ± 0.11 in the thyroid nodule diagnostic tests, which underestimated the periphery of the nodule. WSE imaging got the largest area difference of 3.70 ± 1.4 mm2 between the ground truth and predicted area, which indicated that the delineation of the nodule boundary by the WSE was too large. hNWASE imaging got superior lesion area prediction with the MCC of 0.81 ± 0.06, F1 score of 0.81 ± 0.07, and generalized contrast-to-noise ratio of 0.98 ± 0.03. These findings suggested that hNWASE imaging could improve image quality and be a promising technique for tissue characterization.

Ultrasound k-nearest neighbor entropy imaging: Theory, algorithm, and applications

Ultrasound information entropy is a flexible approach for analyzing ultrasound backscattering. Shannon entropy imaging based on probability distribution histograms (PDHs) has been implemented as a promising method for tissue characterization and diagnosis. However, the bin number affects the stability of entropy estimation. In this study, we introduced the k-nearest neighbor (KNN) algorithm to estimate entropy values and proposed ultrasound KNN entropy imaging. The proposed KNN estimator leveraged the Euclidean distance between data samples, rather than the histogram bins by conventional PDH estimators. We also proposed cumulative relative entropy (CRE) imaging to analyze time-series radiofrequency signals and applied it to monitor thermal lesions induced by microwave ablation (MWA). Computer simulation phantom experiments were conducted to validate and compare the performance of the proposed KNN entropy imaging, the conventional PDH entropy imaging, and Nakagami-m parametric imaging in detecting the variations of scatterer densities and visualizing inclusions. Clinical data of breast lesions were analyzed, and porcine liver MWA experiments ex vivo were conducted to validate the performance of KNN entropy imaging in classifying benign and malignant breast tumors and monitoring thermal lesions, respectively. Compared with PDH, the entropy estimation based on KNN was less affected by the tuning parameters. KNN entropy imaging was more sensitive to changes in scatterer densities and performed better visualizable capability than typical Shannon entropy (TSE) and Nakagami-m parametric imaging. Among different imaging methods, KNN-based Shannon entropy (KSE) imaging achieved the higher accuracy in classification of benign and malignant breast tumors and KNN-based CRE imaging had larger lesion-to-normal contrast when monitoring the ablated areas during MWA at different powers and treatment durations. Ultrasound KNN entropy imaging is a potential quantitative ultrasound approach for tissue characterization.

MULTISCALE ASSESSMENT OF ACHILLES TENDON MECHANICS AND MECHANOBIOLOGY

Achilles tendon mechanical properties depend on a complex hierarchical design, with collagen being the smallest load-bearing unit. At the nanoscale, collagen molecules are organized into fibrils, which at the microscale are assembled into fibers, followed by larger structures such as sub-tendons or fascicles. Degree of in vivo loading affects the collagen content, and organization and consequently the tissue's mechanical response. We aim to unravel how composition, structural organization, and mechanical response are affected by degree of in vivo loading at each length scale. The presentation will outline the results to date about to the use of high-resolution synchrotron-based tissue characterisation methods on several length scales in combination with in situ mechanical tests. We use a rat model, where the tendons are subjected to varying loading in vivo. To characterize the tissue microstructure, phase-contrast enhanced synchrotron micro-tomography is performed. The 3D fiber organization in fully loaded tendons is highly aligned, whereas the fibers in unloaded tendons are significantly more heterogeneously arranged and crimped. To characterize the collagen fibril response, Small Angle X-ray Scattering is performed. Two types of fibril organizations are found; a single population oriented towards the main load direction and two fibril subpopulations with clearly distinct orientations. Scattering during loading showed that the fibrils in unloaded tendons did not strain as much in fully loaded. In situ loading concurrently with high resolution synchrotron experiments show the complex tendon response to in situ load and its relation to in vivo loading and tendon hierarchical structure. Unloading seems to alter the organization of the fibrils and fibers, e.g. increased crimping and more pronounced sub-tendon twists.Acknowledgements: Funding from Knut and Alice Wallenberg Foundation and European Research Council (101002516). Paul Scherrer Institut, Switzerland for beamtime at cSAXS and TOMCAT.

Abstract 15813: Machine Learning Improves Prediction of Heart Failure and Cardiovascular Death Through a Signal Analytical Approach of Echocardiography

Introduction: Recently, speckle tracking echocardiography (STE) and tissue Doppler imaging (TDI) have gained increasing traction as non-invasive tissue characterization methods within cardiology. But until now many patterns from the strain and TDI curves remain uninvestigated. Signal analytical methods like wavelet analysis have shown promising potential in effectively identifying previously unknown pathological patterns present in ECG signals. Therefore, we hypothesized that a signal analytic approach combined with supervised machine learning (ML) on strain and TDI curves could identify unknown pathophysiological deformation drivers of heart failure (HF) and cardiovascular death (CV death) in the general population. Methods: The analysis included 3781 subjects from the general population. In total 720 novel statistical parameters and wavelet signal parameters from 18 strain curves and 6 TDI curves were generated. The parameters were used to train an ensemble decision tree with 20-fold cross-validation and compared to a baseline ML model trained on 22 conventional echocardiographic parameters. Results: Follow up-time was four years. In total 108 subjects (2.9%) met the outcome. By including statistical and wavelet-derived parameters along with 22 conventional echocardiographic parameters in a combined model, the ML model was significantly improved compared to the baseline model (AUC for conventional model: 0.76 vs combined model: 0.825, p = 0.0086). Both statistical parameters, such as the mean and skewness of the TDI curve, as well as wavelet parameters, such as mean on the first decomposition level of the strain curve, were found to be important in the model. Conclusion Adding novel echocardiographic parameters based on signal analytical statistics and wavelet decomposition methods to preexisting conventional echocardiographic parameters significantly improved the prediction of incident HF or CV death in ML models.

Malignant versus normal breast tissue: Optical differentiation exploiting hyperspectral imaging system

Breast malignancy is a critical problem that severely affects women’s health globally with a high-frequency rate, necessitating fast, effective, and early diagnostic methods. The present study aims to measure the breast tissue’s optical properties by capturing the spectral signatures from malignant and normal breast tissue for therapeutic and diagnostic applications. The optical imaging system incorporates a hyperspectral (HS) camera to capture the spectral signatures for both the malignant and normal breast tissues within 400 ~ 1000 nm. The system was subdivided into two exploratory (reflection/transmission) to measure the tissue’s diffuse reflectance (Rd) and light transmission (Tr), respectively. The study involved 30 breast tissue (normal/tumor) samples from 30 females in the age range of 46 ~ 72 years, who were optically inspected in the visible and near-infrared (VIS-NIR) spectra. Then, the inverse adding doubling (IAD) method for breast tissue characterization and descriptive analysis (T-test) was exploited to verify the significant difference between the various types of breast tissues and select the optimum wavelength. Finally, comparing the study outcome with the histopathological examination to evaluate the system’s effectiveness by calculation (sensitivity, specificity, and accuracy). The average outcome values demonstrated that the optimal spectral bands distinguishing between the normal and the tumor tissues regarding the reflectance approach were 600 ~ 680 nm and 750 ~ 960 nm at the VIS and NIR spectrum, respectively. Then, for the transmission technique, the optimal spectral bands were 560 ~ 590 nm and 760 ~ 810 nm at the VIS and NIR spectra, respectively. Later, the T-test and the IAD verified that the highest Rd values for discrimination were 600 ~ 640 nm and 800 ~ 840 nm at the VIS and NIR spectra, respectively. On the other side, the highest Tr values were 600 ~ 640 nm and 760 ~ 800 nm at the VIS and NIR spectra, respectively. The investigation’s average reading accuracy, sensitivity, and specificity were 85%, 81.88%, and 88.8%, respectively. The experimental trials revealed that the system could identify the optimal wavelength for therapeutic and diagnostic applications through the light interaction behavior of the breast tissue’s optical properties.

Building a comprehensive cardiovascular magnetic resonance exam on a commercial 0.55 T system: A pictorial essay on potential applications.

Contemporary advances in low-field magnetic resonance imaging systems can potentially widen access to cardiovascular magnetic resonance (CMR) imaging. We present our initial experience in building a comprehensive CMR protocol on a commercial 0.55 T system with a gradient performance of 26 mT/m amplitude and 45 T/m/s slew rate. To achieve sufficient image quality, we adapted standard imaging techniques when possible, and implemented compressed-sensing (CS) based techniques when needed in an effort to compensate for the inherently low signal-to-noise ratio at lower field strength. A prototype CMR exam was built on an 80 cm, ultra-wide bore commercial 0.55 T MR system. Implementation of all components aimed to overcome the inherently lower signal of low-field and the relatively longer echo and repetition times owing to the slower gradients. CS-based breath-held and real-time cine imaging was built utilizing high acceleration rates to meet nominal spatial and temporal resolution recommendations. Similarly, CS 2D phase-contrast cine was implemented for flow. Dark-blood turbo spin echo sequences with deep learning based denoising were implemented for morphology assessment. Magnetization-prepared single-shot myocardial mapping techniques incorporated additional source images. CS-based dynamic contrast-enhanced imaging was implemented for myocardial perfusion and 3D MR angiography. Non-contrast 3D MR angiography was built with electrocardiogram-triggered, navigator-gated magnetization-prepared methods. Late gadolinium enhanced (LGE) tissue characterization methods included breath-held segmented and free-breathing single-shot imaging with motion correction and averaging using an increased number of source images. Proof-of-concept was demonstrated through porcine infarct model, healthy volunteer, and patient scans. Reasonable image quality was demonstrated for cardiovascular structure, function, flow, and LGE assessment. Low-field afforded utilization of higher flip angles for cine and MR angiography. CS-based techniques were able to overcome gradient speed limitations and meet spatial and temporal resolution recommendations with imaging times comparable to higher performance scanners. Tissue mapping and perfusion imaging require further development. We implemented cardiac applications demonstrating the potential for comprehensive CMR on a novel commercial 0.55 T system. Further development and validation studies are needed before this technology can be applied clinically.

Tissue Characterization of Puborectalis Muscle From 3-D Ultrasound.

Pelvic floor (PF) muscles have the role of preventing pelvic organ descent. The puborectalis muscle (PRM), which is one of the female PF muscles, can be damaged during child delivery. This damage can potentially cause irreversible muscle trauma and even lead to an avulsion, which is disconnection of the muscle from its insertion point, the pubic bone. Ultrasound imaging allows diagnosis of such trauma based on comparison of geometric features of a damaged muscle with the geometric features of a healthy muscle. Although avulsion, which is considered severe damage, can be diagnosed, microdamage within the muscle itself leading to structural changes cannot be diagnosed by visual inspection through imaging only. Therefore, we developed a quantitative ultrasound tissue characterization method to obtain information on the state of the tissue of the PRM and the presence of microdamage in avulsed PRMs. The muscle was segmented as the region of interest (ROI) and further subdivided into six regions of interest (sub-ROIs). Mean echogenicity, entropy and shape parameter of the statistical distribution of gray values were analyzed on two of these sub-ROIs nearest to the bone. The regions nearest to the bones are also the most likely regions to exhibit damage in case of disconnection or avulsion. This analysis was performed for both the muscle at rest and the muscle in contraction. We found that, for PRMs with unilateral avulsion compared with undamaged PRMs, the mean echogenicity (p=0.02) and shape parameter (p < 0.01) were higher, whereas the entropy was lower (p < 0.01). This method might be applicable to quantification of PRM damage within the muscle.

One-class machine learning classification of skin tissue based on manually scanned optical coherence tomography imaging

We investigated a method for automatic skin tissue characterization based on optical coherence tomography (OCT) imaging. We developed a manually scanned single fiber OCT instrument to perform in vivo skin imaging and tumor boundary assessment. The goal is to achieve more accurate tissue excision in Mohs micrographic surgery (MMS) and reduce the time required for MMS. The focus of this study was to develop a novel machine learning classification method to automatically identify abnormal skin tissues through one-class classification. We trained a deep convolutional neural network (CNN) with a U-Net architecture for automatic skin segmentation, used the pre-trained U-Net as a feature extractor, and trained one-class support vector machine (SVM) classifiers to detect abnormal tissues. The novelty of this study is the use of a neural network as a feature extractor and the use of a one-class SVM for abnormal tissue detection. Our approach eliminated the need to engineer the features for classification and eliminated the need to train the classifier with data obtained from abnormal tissues. To validate the effectiveness of the one-class classification method, we assessed the performance of our algorithm using computer synthesized data, and experimental data. We also performed a pilot study on a patient with skin cancer.

Virtual pathology: Reaching higher standards for noninvasive CTA tissue characterization capability by using histology as a truth standard

AimsDespite advances in therapy, reduction in myocardial infarction or death remains elusive. Whereas computed tomography angiography (CTA) is increasingly appreciated, the analyses are often subjective or qualitative. Methods for specific tissue characterization using histopathologic correlates have recently been reported. We extend this here to demonstrate accurate discrimination between, and quantitation of, lipid-rich necrotic core (LRNC), intraplaque hemorrhage (IPH), and fibrotic tissues. MethodsNCT02143102 collected 576 tissue samples with paired CTA. Cardiovascular pathologists annotated LRNC, IPH, and dense calcification (CALC) regions as a reference standard. Blinded to histology, CTA was analyzed using ElucidVivo (Elucid Bioimaging Inc., Boston, MA USA). Structure and tissue characteristics of atherosclerotic plaque from CTA, accounting for both the imaging acquisition process and the biology, accounting for differences in density distributions that result from the different cellular and molecular level milieu of the relevant tissue types. ResultsLRNC was tested across a true range of 0–10 mm2, with a difference of 0.15 mm2 and a slope of 0.92. IPH was tested across a true range of 0–18 mm2, with a difference from histology of 1.68 mm2 and a slope of 0.95. CALC was tested across a range of 0–14 mm2, with a difference of −0.06 mm2 and a slope of 0.99. Matrix tissue (MATX) was tested across a range of 4–52 mm2, with a difference of 0.02 mm2 and a slope of 0.91. ConclusionLRNC, IPH, CALC, and MATX may be objectively quantified using histopathologic correlates automatically from CTA for use singly or in combination to optimize patient care. The availability of objectively validated quantitative markers that may be followed longitudinally may extend the clinical utility of CTA. Additionally, these measures contribute efficacy variables for developing novel drugs and clinical decision support tools for tailored therapeutics.

High-contrast spectroscopic photoacoustic characterization of thermal tissue ablation in the visible spectrum.

High-contrast tissue characterization of thermal ablation has been desired to evaluate therapeutic outcomes accurately. This paper presents a photoacoustic (PA) characterization of thermal tissue ablation in the visible spectrum, in which higher light absorbance can produce spectral contrast starker than in the near-infrared range. Ex vivo experiments were performed to measure visible PA spectra (480-700 nm) from fresh porcine liver tissues that received a thermal dose in a range of cumulative equivalent minutes at 43°C (CEM43). The local hemoglobin lobe area between 510-600 nm and wholespectral area under the curve were evaluated to represent the transition of hemoglobin into methemoglobin (MetHb) in the target tissue. The thermal process below an estimated therapeutic CEM43 threshold (80-340 minutes) presented a progressive elevation of the PA spectrum and an eventual loss of local hemoglobin peaks in the visible spectrum, closer to the MetHb spectrum. Interestingly, an excessive CEM43 produced a substantial drop in the PA spectrum. In the spectral analysis, the visible spectrum yielded 13.9-34.1 times higher PA sensitivity and 1.42 times higher contrast change than at a near-infrared wavelength. This novel method of PA tissue characterization in the visible spectrum could be a potential modality to evaluate various thermal therapeutic modalities at high-contrast resolution.

Characterization of mechanical properties of soft tissues using sub-microscale tensile testing and 3D-Printed sample holder

Obtaining the mechanical properties of soft tissues is critical in many medical fields, such as regenerative medicine and surgical simulation training. Although various tissue-characterization methods have been developed, such as AFM, indentation, and elastography, there remain some limitations on their accuracy and validity for measuring small and fragile soft tissues. This paper presents a tensile testing technique to measure the mechanical properties of soft tissues directly and accurately. Tensile testing was chosen as the primary method because of its simple procedure and ability to derive mechanical properties without requiring many assumptions or complicated models. However, tensile testing on soft tissues presents challenges related to gripping the tissue sample without affecting its inherent properties, applying minuscule forces to the sample, and measuring the cross-section area and strain of the sample. To solve these issues, this study presents a sub-micro scale tensile testing system that uses a flexure mechanism and a novel 3D-printed sample holder for gripping the tissue samples. The system also measures tested samples' cross-section area and strain using two high-resolution cameras. The system was validated by testing standard materials and used to characterize the elastic modulus, yield stress, and yield strain of lung tissue slices from six different mice. The results from the validation tests showed a less than 2.5% error for elastic modulus values measured using the tensile tester. At the same time, results from the mice lung tissue measurements revealed qualitative findings that closely matched those seen in the literature and displayed low coefficient of variation values, demonstrating the high repeatability of the system.

One-step iterative reconstruction approach based on eigentissue decomposition for spectral photon-counting computed tomography.

Purpose: We propose a one-step tissue characterization method for spectral photon-counting computed tomography (SPCCT) using eigentissue decomposition (ETD), tailored for highly accurate human tissue characterization in radiotherapy. Methods: The approach combines a Poisson likelihood, a spatial prior, and a quantitative prior constraining eigentissue fractions based on expected values for tabulated tissues. There are two regularization parameters: for the quantitative prior, and for the spatial prior. The approach is validated in a realistic simulation environment for SPCCT. The impact of and is evaluated on a virtual phantom. The framework is tested on a virtual patient and compared with two sinogram-based two-step methods [using respectively filtered backprojection (FBP) and an iterative method for the second step] and a post-reconstruction approach with the same quantitative prior. All methods use ETD. Results: Optimal performance with respect to bias or RMSE is achieved with different combinations of and on the cylindrical phantom. Evaluated in tissues of the virtual patient, the one-step framework outperforms two-step and post-reconstruction approaches to quantify proton-stopping power (SPR). The mean absolute bias on the SPR is 0.6% (two-step FBP), 0.6% (two-step iterative), 0.6% (post-reconstruction), and 0.2% (one-step optimized for low bias). Following the same order, the RMSE on the SPR is 13.3%, 2.5%, 3.2%, and 1.5%. Conclusions: Accurate and precise characterization with ETD can be achieved with noisy SPCCT data without the need to rely on post-reconstruction methods. The one-step framework is more accurate and precise than two-step methods for human tissue characterization.

An Improved Parameter Estimator of the Homodyned K Distribution Based on the Maximum Likelihood Method for Ultrasound Tissue Characterization.

The homodyned K distribution (HK) can generally describe the ultrasound backscatter envelope statistics distribution with parameters that have specific physical meaning. However, creating robust and reliable HK parameter estimates remains a crucial concern. The maximum likelihood estimator (MLE) usually yields a small variance and bias in parameter estimation. Thus, two recent studies have attempted to use MLE for parameter estimation of HK distribution. However, some of the statements in these studies are not fully justified and they may hinder the application of parameter estimation of HK distribution based on MLE. In this study, we propose a new parameter estimator for the HK distribution based on the MLE (i.e., MLE1), which overcomes the disadvantages of conventional MLE of HK distribution. The MLE1 was compared with other estimators, such as XU estimator (an estimation method based on the first moment of the intensity and tow log-moments) and ANN estimator (an estimation method based on artificial neural networks). We showed that the estimations of parameters α and k are the best overall (in terms of the relative bias, normalized standard deviation, and relative root mean squared errors) using the proposed MLE1 compared with the others based on the simulated data when the sample size was N = 1000. Moreover, we assessed the usefulness of the proposed MLE1 when the number of scatterers per resolution cell was high (i.e., α up to 80) and when the sample size was small (i.e., N = 100), and we found a satisfactory result. Tests on simulated ultrasound images based on Field II were performed and the results confirmed that the proposed MLE1 is feasible and reliable for the parameter estimation from the ultrasonic envelope signal. Therefore, the proposed MLE1 can accurately estimate the HK parameters with lower uncertainty, which presents a potential practical value for further ultrasonic applications.

Advances in Multimodality Cardiovascular Imaging in the Diagnosis of Heart Failure With Preserved Ejection Fraction.

Heart failure with preserved ejection fraction (HFpEF) is a syndrome defined by the presence of heart failure symptoms and increased levels of circulating natriuretic peptide (NP) in patients with preserved left ventricular ejection fraction and various degrees of diastolic dysfunction (DD). HFpEF is a complex condition that encompasses a wide range of different etiologies. Cardiovascular imaging plays a pivotal role in diagnosing HFpEF, in identifying specific underlying etiologies, in prognostic stratification, and in therapeutic individualization. Echocardiography is the first line imaging modality with its wide availability; it has high spatial and temporal resolution and can reliably assess systolic and diastolic function. Cardiovascular magnetic resonance (CMR) is the gold standard for cardiac morphology and function assessment, and has superior contrast resolution to look in depth into tissue changes and help to identify specific HFpEF etiologies. Differently, the most important role of nuclear imaging [i.e., planar scintigraphy and/or single photon emission CT (SPECT)] consists in the screening and diagnosis of cardiac transthyretin amyloidosis (ATTR) in patients with HFpEF. Cardiac CT can accurately evaluate coronary artery disease both from an anatomical and functional point of view, but tissue characterization methods have also been developed. The aim of this review is to critically summarize the current uses and future perspectives of echocardiography, nuclear imaging, CT, and CMR in patients with HFpEF.

Abstract 9479: Novel Predictors of Left Ventricular Reverse Remodeling in Patients With Recent-onset Dilated Cardiomyopathy Using T1 and T2 Mapping on Cardiac Magnetic Resonance

Introduction: The prognosis of patients with idiopathic dilated cardiomyopathy (DCM) has improved in recent decades with guideline-directed medical therapy. There is an essential need for better biomarkers to predict left ventricular reverse remodeling (LVRR) in DCM. Native T1, T2 and left ventricular (LV) global strain by cardiac magnetic resonance (CMR) without contrast agents provides a novel quantitative method for myocardial tissue characterization. Hypothesis: CMR parameters predict LVRR in patients with DCM. Methods: A total 34 patients ( 58 ± 13 years, 25 male) with DCM underwent baseline CMR imaging (3T Philips), including cine, late gadolinium enhancement (LGE) CMR, and triple-slice T1 and T2 mapping (short modified look locker inversion recovery ShMOLLI sequence). The 6SD-LGE (% area), global circumferential strain (GCS), global radial strain (GRS), global longitudinal strain (GLS), global radial strain (GRS), LV mass index (LVMI), LV end-diastolic volume (EDV), end-systolic volume (ESV) and ejection fraction (EF) were calculated from multiple short-axis views. Echocardiography was conducted before and &gt; 1 year after after optimal medical therapy. Results: The patients were divided into two groups, LVRR with LVEF increase of &gt;10% and the final value of ≥40% for those without. There were no significant differences in LV mass index, LVEDV, LVESV, and LVEF between the two groups (80 ± 20 g/m 2 vs 78 ± 19 g/m 2 ; 174 ± 54 ml vs 165 ± 48 ml; 134 ± 51 ml vs 122 ± 46 ml; 24 ± 9% vs 28 ± 9%, respectively). The any LV global strain showed no significant difference between the two groups. The 6SD-LGE, native T1, and T2 values were significantly lower in LVRR+ group than in LVRR-group (15.3 ± 15.5 % vs 5.9 ± 6.8%;1376 ± 59 ms vs 1301 ± 42 ms;55 ± 3.5 ms vs 50 ± 2.5 ms,respectively). The LGE-6SD &lt; 13%. native T1 value &lt; 1330 ms and T2 values &lt; 53ms were the best threshold values for predicting LVRR (The LGE-6SD: sensitivity, 82%; specificity, 41%; AUC 0.72, T1:sensitivity, 82%; specificity, 71%; AUC 0.87, and T2: sensitivity, 82%; specificity, 63%; AUC 0.88). Conclusions: Native T1 and T2 mapping values may be useful for assessing of the LVRR in patients with DCM even without contrast.

Ten Fast Transfer Learning Models for Carotid Ultrasound Plaque Tissue Characterization in Augmentation Framework Embedded with Heatmaps for Stroke Risk Stratification.

Background and Purpose: Only 1–2% of the internal carotid artery asymptomatic plaques are unstable as a result of >80% stenosis. Thus, unnecessary efforts can be saved if these plaques can be characterized and classified into symptomatic and asymptomatic using non-invasive B-mode ultrasound. Earlier plaque tissue characterization (PTC) methods were machine learning (ML)-based, which used hand-crafted features that yielded lower accuracy and unreliability. The proposed study shows the role of transfer learning (TL)-based deep learning models for PTC. Methods: As pertained weights were used in the supercomputer framework, we hypothesize that transfer learning (TL) provides improved performance compared with deep learning. We applied 11 kinds of artificial intelligence (AI) models, 10 of them were augmented and optimized using TL approaches—a class of Atheromatic™ 2.0 TL (AtheroPoint™, Roseville, CA, USA) that consisted of (i–ii) Visual Geometric Group-16, 19 (VGG16, 19); (iii) Inception V3 (IV3); (iv–v) DenseNet121, 169; (vi) XceptionNet; (vii) ResNet50; (viii) MobileNet; (ix) AlexNet; (x) SqueezeNet; and one DL-based (xi) SuriNet-derived from UNet. We benchmark 11 AI models against our earlier deep convolutional neural network (DCNN) model. Results: The best performing TL was MobileNet, with accuracy and area-under-the-curve (AUC) pairs of 96.10 ± 3% and 0.961 (p < 0.0001), respectively. In DL, DCNN was comparable to SuriNet, with an accuracy of 95.66% and 92.7 ± 5.66%, and an AUC of 0.956 (p < 0.0001) and 0.927 (p < 0.0001), respectively. We validated the performance of the AI architectures with established biomarkers such as greyscale median (GSM), fractal dimension (FD), higher-order spectra (HOS), and visual heatmaps. We benchmarked against previously developed Atheromatic™ 1.0 ML and showed an improvement of 12.9%. Conclusions: TL is a powerful AI tool for PTC into symptomatic and asymptomatic plaques.

Intra-plaque Hemorrhage And Lipid-rich Necrotic Core May Be Objectively Quantified Using Histopathologic Correlates Automatically From CTA

Introduction: Whereas CTA is increasingly appreciated for atherosclerosis phenotyping, the analyses are often subjective or qualitative. Methods for specific tissue characterization using histopathologic correlates have recently been reported. We extend this here to demonstrate accurate quantitation of both lipid-rich necrotic core (LRNC) and intraplaque hemorrhage (IPH), with the aim to reliably discriminate between them.

Accelerated 3D whole-brain T1, T2, and proton density mapping: feasibility for clinical glioma MR imaging

PurposeAdvanced MRI-based biomarkers offer comprehensive and quantitative information for the evaluation and characterization of brain tumors. In this study, we report initial clinical experience in routine glioma imaging with a novel, fully 3D multiparametric quantitative transient-state imaging (QTI) method for tissue characterization based on T1 and T2 values.MethodsTo demonstrate the viability of the proposed 3D QTI technique, nine glioma patients (grade II–IV), with a variety of disease states and treatment histories, were included in this study. First, we investigated the feasibility of 3D QTI (6:25 min scan time) for its use in clinical routine imaging, focusing on image reconstruction, parameter estimation, and contrast-weighted image synthesis. Second, for an initial assessment of 3D QTI-based quantitative MR biomarkers, we performed a ROI-based analysis to characterize T1 and T2 components in tumor and peritumoral tissue.ResultsThe 3D acquisition combined with a compressed sensing reconstruction and neural network-based parameter inference produced parametric maps with high isotropic resolution (1.125 × 1.125 × 1.125 mm3 voxel size) and whole-brain coverage (22.5 × 22.5 × 22.5 cm3 FOV), enabling the synthesis of clinically relevant T1-weighted, T2-weighted, and FLAIR contrasts without any extra scan time. Our study revealed increased T1 and T2 values in tumor and peritumoral regions compared to contralateral white matter, good agreement with healthy volunteer data, and high inter-subject consistency.Conclusion3D QTI demonstrated comprehensive tissue assessment of tumor substructures captured in T1 and T2 parameters. Aiming for fast acquisition of quantitative MR biomarkers, 3D QTI has potential to improve disease characterization in brain tumor patients under tight clinical time-constraints.