Compact reverse time migration: A real-time approach for full waveform ultrasound imaging for breast.

We present compact reverse time migration (CRTM), a real-time ultrasound imaging method that can exploit the full waveform information of ultrasonic wave records for imaging breast tissue. Conventional reverse time migration (RTM) computes the gradient of the reflective ultrasound data with respect to the perturbation of the velocity model of the soft tissues and the gradient can indicate the interface between different types of body tissue. In contrast to conventional reflection ultrasound (B-mode), which is based on the high-frequency approximation to the wave equation, the RTM algorithm is based on the complete wave equation, and can thus exploit the full waveform (wide-spectrum) information of the data and provide an image with higher resolution. Unfortunately, the computational burden of RTM is noticeably higher than the ray-based B-mode. This precludes real-time applications, one of the most important features of ultrasound imaging. The proposed CRTM algorithm can significantly reduce the computational costs of RTM, such that it can be applied for real-time imaging. We demonstrate the performance of CRTM through a synthetic experiment of ultrasound breast imaging. CRTM can be potentially adapted to related signal-processing fields, such as seismic imaging, acoustic camera systems, and radar imaging.

Mechanical modeling of the maturation process for tissue-engineered implants: Application to biohybrid heart valves.

The development of tissue-engineered cardiovascular implants can improve the lives of large segments of our society who suffer from cardiovascular diseases. Regenerative tissues are fabricated using a process called tissue maturation. Furthermore, it is highly challenging to produce cardiovascular regenerative implants with sufficient mechanical strength to withstand the loading conditions within the human body. Therefore, biohybrid implants for which the regenerative tissue is reinforced by standard reinforcement material (e.g. textile or 3d printed scaffold) can be an interesting solution. In silico models can significantly contribute to characterizing, designing, and optimizing biohybrid implants. The first step towards this goal is to develop a computational model for the maturation process of tissue-engineered implants. This paper focuses on the mechanical modeling of textile-reinforced tissue-engineered cardiovascular implants. First, an energy-based approach is proposed to compute the collagen evolution during the maturation process. Then, the concept of structural tensors is applied to model the anisotropic behavior of the extracellular matrix and the textile scaffold. Next, the newly developed material model is embedded into a special solid-shell finite element formulation with reduced integration. Finally, our framework is used to compute two structural problems: a pressurized shell construct and a tubular-shaped heart valve. The results show the ability of the model to predict collagen growth in response to the boundary conditions applied during the maturation process. Consequently, the model can predict the implant's mechanical response, such as the deformation and stresses of the implant.

Synchronization-Inspired Interpretable Neural Networks.

Synchronization is a ubiquitous phenomenon in nature that enables the orderly presentation of information. In the human brain, for instance, functional modules such as the visual, motor, and language cortices form through neuronal synchronization. Inspired by biological brains and previous neuroscience studies, we propose an interpretable neural network incorporating a synchronization mechanism. The basic idea is to constrain each neuron, such as a convolution filter, to capture a single semantic pattern while synchronizing similar neurons to facilitate the formation of interpretable functional modules. Specifically, we regularize the activation map of a neuron to surround its focus position of the activated pattern in a sample. Moreover, neurons locally interact with each other, and similar ones are synchronized together during the training phase adaptively. Such local aggregation preserves the globally distributed representation nature of the neural network model, enabling a reasonably interpretable representation. To analyze the neuron interpretability comprehensively, we introduce a series of novel evaluation metrics from multiple aspects. Qualitative and quantitative experiments demonstrate that the proposed method outperforms many state-of-the-art algorithms in terms of interpretability. The resulting synchronized functional modules show module consistency across data and semantic specificity within modules.

In-vivo assessment of vascular injury for the prediction of in-stent restenosis.

BACKGROUND: Despite optimizations of coronary stenting technology, a residual risk of in-stent restenosis (ISR) remains. Vessel wall injury has important impact on the development of ISR. While injury can be assessed in histology, there is no injury score available to be used in clinical practice. METHODS: Seven rats underwent abdominal aorta stent implantation. At 4 weeks after implantation, animals were euthanized, and strut indentation, defined as the impression of the strut into the vessel wall, as well as neointimal growth were assessed. Established histological injury scores were assessed to confirm associations between indentation and vessel wall injury. In addition, stent strut indentation was assessed by optical coherence tomography (OCT) in an exemplary clinical case. RESULTS: Stent strut indentation was associated with vessel wall injury in histology. Furthermore, indentation was positively correlated with neointimal thickness, both in the per-strut analysis (r = 0.5579) and in the per-section analysis (r = 0.8620; both p ≤ 0.001). In a clinical case, indentation quantification in OCT was feasible, enabling assessment of injury in vivo. CONCLUSION: Assessing stent strut indentation enables periprocedural assessment of stent-induced damage in vivo and therefore allows for optimization of stent implantation. The assessment of stent strut indentation might become a valuable tool in clinical practice.

A Novel Deep Learning Model as a Donor-Recipient Matching Tool to Predict Survival after Liver Transplantation.

BACKGROUND: The "digital era" in the field of medicine is the new "here and now". Artificial intelligence has entered many fields of medicine and is recently emerging in the field of organ transplantation. Solid organs remain a scarce resource. Being able to predict the outcome after liver transplantation promises to solve one of the long-standing problems within organ transplantation. What is the perfect donor recipient match? Within this work we developed and validated a novel deep-learning-based donor-recipient allocation system for liver transplantation. METHOD: In this study we used data collected from all liver transplant patients between 2004 and 2019 at the university transplantation centre in Munich. We aimed to design a transparent and interpretable deep learning framework to predict the outcome after liver transplantation. An individually designed neural network was developed to meet the unique requirements of transplantation data. The metrics used to determine the model quality and its level of performance are accuracy, cross-entropy loss, and F1 score as well as AUC score. RESULTS: A total of 529 transplantations with a total of 1058 matching donor and recipient observations were added into the database. The combined prediction of all outcome parameters was 95.8% accurate (cross-entropy loss of 0.042). The prediction of death within the hospital was 94.3% accurate (cross-entropy loss of 0.057). The overall F1 score was 0.899 on average, whereas the overall AUC score was 0.940. CONCLUSION: With the achieved results, the network serves as a reliable tool to predict survival. It adds new insight into the potential of deep learning to assist medical decisions. Especially in the field of transplantation, an AUC Score of 94% is very valuable. This neuronal network is unique as it utilizes transparent and easily interpretable data to predict the outcome after liver transplantation. Further validation must be performed prior to utilization in a clinical context.

Long-term results of bacterial septic arthritis of the wrist.

Septic arthritis of the wrist is a serious condition, yet little is known about its long-term outcome. A retrospective analysis of 22 patients treated for bacterial septic arthritis of the wrist was conducted with subsequent follow-up of 18 patients with a median period of 44 months to assess functional results via DASH-Score and clinical examination. Arthrotomy was used to treat all patients; in 19 patients, multiple operations were needed to cure the infection. Follow-up revealed a mean DASH-score of 34 (SD 22) and a significant correlation with needed surgical radicality and number of needed operations. The range of motion of the wrist and grip strength of the affected side was 49% (SD 20%) and 70% (SD 28%) of the contralateral side, respectively. In conclusion, septic arthritis of the wrist leads to long-term functional restrictions with a strong correlation with the stage of the disease. Hence, early diagnosis and treatment are paramount.

Targeting density-enhanced phosphatase-1 (DEP-1) with antisense oligonucleotides improves the metabolic phenotype in high-fat diet-fed mice.

BACKGROUND: Insulin signaling is tightly controlled by tyrosine dephosphorylation of the insulin receptor through protein-tyrosine-phosphatases (PTPs). DEP-1 is a PTP dephosphorylating tyrosine residues in a variety of receptor tyrosine kinases. Here, we analyzed whether DEP-1 activity is differentially regulated in liver, skeletal muscle and adipose tissue under high-fat diet (HFD), examined the role of DEP-1 in insulin resistance in vivo, and its function in insulin signaling. RESULTS: Mice were fed an HFD for 10 weeks to induce obesity-associated insulin resistance. Thereafter, HFD mice were subjected to systemic administration of specific antisense oligonucleotides (ASOs), highly accumulating in hepatic tissue, against DEP-1 or control ASOs. Targeting DEP-1 led to improvement of insulin sensitivity, reduced basal glucose level, and significant reduction of body weight. This was accompanied by lower insulin and leptin serum levels. Suppression of DEP-1 in vivo also induced hyperphosphorylation in the insulin signaling cascade of the liver. Moreover, DEP-1 physically associated with the insulin receptor in situ, and recombinant DEP-1 dephosphorylated the insulin receptor in vitro. CONCLUSIONS: These results indicate that DEP-1 acts as an endogenous antagonist of the insulin receptor, and downregulation of DEP-1 results in an improvement of insulin sensitivity. DEP-1 may therefore represent a novel target for attenuation of metabolic diseases.

Sexual dimorphism in obesity-mediated left ventricular hypertrophy.

In the present study we investigated the influence of sex difference on the development of left ventricular hypertrophy (LVH) during obesity. Male and female C57BL/6J mice were fed for 15 and 25 wk with a high-fat diet (HFD) or low-fat control diet (LFD). Analysis of body composition, monitoring of body weight (BW), and echocardiographic analysis were performed, as well as analysis of expression of different adipocytokines in epicardial adipose tissue. The increment in left ventricular mass (LVM) after HFD (25 wk) was significantly stronger in male mice compared with female mice [LVM: male, 116.9 ± 2.9 (LFD) vs. 142.2 ± 9.3 mg (HFD); female, 84.3 ± 3.3 (LFD) vs. 93.9 ± 1.7 mg (HFD), Psex < 0.01]. In parallel, males developed a higher BW and fat mass after 25 wk HFD than female mice [BW: male, 33 ± 0.9 (LFD) vs. 53 ± 0.8 g (HFD); fat mass: male, 8.8 ± 0.9 (LFD) vs. 22.8 ± 0.7 g (HFD); BW: female, 22.5 ± 0.4 (LFD) vs. 33.7 ± 1.3 g (HFD); fat mass: female, 4.0 ± 0.2 (LFD) vs. 13.2 ± 1.2 g (HFD)] (P < 0.01 for BW+ fat mass female vs. male). The mRNA expression of adipocytokines in epicardial fat after 25 wk of diet showed higher levels of adiponectin (2.8-fold), leptin (4.2-fold), and vaspin (11.9-fold) in male mice compared with female mice (P < 0.05). To identify new adipose-derived molecular mediators of LVH, we further elucidated the cardiac impact of vaspin. Murine primary cardiac fibroblast proliferation was significantly induced by vaspin (1.8-fold, vaspin 1 µg/l, P < 0.05 vs. control) compared with 1.9-fold induction by angiotensin II (10 µM). The present study demonstrates a sex-dependent regulation of diet-induced LVH associated with sexual dimorphic expression of adipocytokines in epicardial adipose tissue.

Moderate correlations of in vitro versus in vivo pharmacokinetics questioning the need of early microsomal stability testing.

BACKGROUND/AIMS: Putative in vitro-in vivo correlations of pharmacokinetic (PK) parameters are regarded as a prerequisite to filter hits derived from high-throughput screening (HTS) approaches for subsequent murine in vivo PK studies. METHODS: In this study, we assessed stabilities in rat and human microsomes of 121 compounds from an early, academic drug discovery programme targeting the (pro)renin receptor and correlated the respective data with single-dose, in vivo PK parameters of 22 hits administered intravenously in rats. RESULTS: After transformation of in vitro half-lives to predicted in vivo hepatic clearances, r(2) regarding in vitro-in vivo clearance correlations were 0.31 and 0.27 for the rat and human species, respectively. CONCLUSIONS: Our data concerning structurally diverse real-world compounds indicate that microsomal stability testing is not a tool to triage early compounds for in vivo PK testing.

Sexual dimorphic regulation of body weight dynamics and adipose tissue lipolysis.

BACKGROUND: Successful reduction of body weight (BW) is often followed by recidivism to obesity. BW-changes including BW-loss and -regain is associated with marked alterations in energy expenditure (EE) and adipose tissue (AT) metabolism. Since these processes are sex-specifically controlled, we investigated sexual dimorphisms in metabolic processes during BW-dynamics (gain-loss-regain). RESEARCH DESIGN: Obesity was induced in C57BL/6J male (m) and female (f) mice by 15 weeks high-fat diet (HFD) feeding. Subsequently BW was reduced (-20%) by caloric restriction (CR) followed by adaptive feeding, and a regain-phase. Measurement of EE, body composition, blood/organ sampling were performed after each feeding period. Lipolysis was analyzed ex-vivo in gonadal AT. RESULTS: Male mice exhibited accelerated BW-gain compared to females (relative BW-gain m:140.5±3.2%; f:103.7±6.5%; p<0.001). In consonance, lean mass-specific EE was significantly higher in females compared to males during BW-gain. Under CR female mice reached their target-BW significantly faster than male mice (m:12.2 days; f:7.6 days; p<0.001) accompanied by a sustained sex-difference in EE. In addition, female mice predominantly downsized gonadal AT whereas the relation between gonadal and total body fat was not altered in males. Accordingly, only females exhibited an increased rate of forskolin-stimulated lipolysis in AT associated with significantly higher glycerol concentrations, lower RER-values, and increased AT expression of adipose triglyceride lipase (ATGL) and hormone sensitive lipase (HSL). Analysis of AT lipolysis in estrogen receptor alpha (ERα)-deficient mice revealed a reduced lipolytic rate in the absence of ERα exclusively in females. Finally, re-feeding caused BW-regain faster in males than in females. CONCLUSION: The present study shows sex-specific dynamics during BW-gain-loss-regain. Female mice responded to CR with an increase in lipolytic activity, and augmented lipid-oxidation leading to more efficient weight loss. These processes likely involve ERα-dependent signaling in AT and sexual dimorphic regulation of genes involved in lipid metabolism.

Prediction of Alzheimer's disease using individual structural connectivity networks.

Alzheimer's disease (AD) progressively degrades the brain's gray and white matter. Changes in white matter reflect changes in the brain's structural connectivity pattern. Here, we established individual structural connectivity networks (ISCNs) to distinguish predementia and dementia AD from healthy aging in individual scans. Diffusion tractography was used to construct ISCNs with a fully automated procedure for 21 healthy control subjects (HC), 23 patients with mild cognitive impairment and conversion to AD dementia within 3 years (AD-MCI), and 17 patients with mild AD dementia. Three typical pattern classifiers were used for AD prediction. Patients with AD and AD-MCI were separated from HC with accuracies greater than 95% and 90%, respectively, irrespective of prediction approach and specific fiber properties. Most informative connections involved medial prefrontal, posterior parietal, and insular cortex. Patients with mild AD were separated from those with AD-MCI with an accuracy of approximately 85%. Our finding provides evidence that ISCNs are sensitive to the impact of earliest stages of AD. ISCNs may be useful as a white matter-based imaging biomarker to distinguish healthy aging from AD.

Applying microdroplets as sensors for label-free detection of chemical reactions.

Despite its tremendous high-throughput screening capabilities, widespread applications of droplet-based microfluidics are still limited by the poor availability of appropriate analytical assays. Here we report on a novel sensor method that exploits the osmosis-driven change in droplet size as a quantitative and label-free marker for reactions inside the droplets. We present an analysis of the underlying mechanism and apply the method for monitoring metabolic activity at a single-cell level.

New similarity search based glioma grading.

INTRODUCTION: MR-based differentiation between low- and high-grade gliomas is predominately based on contrast-enhanced T1-weighted images (CE-T1w). However, functional MR sequences as perfusion- and diffusion-weighted sequences can provide additional information on tumor grade. Here, we tested the potential of a recently developed similarity search based method that integrates information of CE-T1w and perfusion maps for non-invasive MR-based glioma grading. METHODS: We prospectively included 37 untreated glioma patients (23 grade I/II, 14 grade III gliomas), in whom 3T MRI with FLAIR, pre- and post-contrast T1-weighted, and perfusion sequences was performed. Cerebral blood volume, cerebral blood flow, and mean transit time maps as well as CE-T1w images were used as input for the similarity search. Data sets were preprocessed and converted to four-dimensional Gaussian Mixture Models that considered correlations between the different MR sequences. For each patient, a so-called tumor feature vector (= probability-based classifier) was defined and used for grading. Biopsy was used as gold standard, and similarity based grading was compared to grading solely based on CE-T1w. RESULTS: Accuracy, sensitivity, and specificity of pure CE-T1w based glioma grading were 64.9%, 78.6%, and 56.5%, respectively. Similarity search based tumor grading allowed differentiation between low-grade (I or II) and high-grade (III) gliomas with an accuracy, sensitivity, and specificity of 83.8%, 78.6%, and 87.0%. CONCLUSION: Our findings indicate that integration of perfusion parameters and CE-T1w information in a semi-automatic similarity search based analysis improves the potential of MR-based glioma grading compared to CE-T1w data alone.

Measuring forces between two single actin filaments during bundle formation.

Bundles of filamentous actin are dominant cytoskeletal structures, which play a crucial role in various cellular processes. As yet quantifying the fundamental interaction between two individual actin filaments forming the smallest possible bundle has not been realized. Applying holographic optical tweezers integrated with a microfluidic platform, we were able to measure the forces between two actin filaments during bundle formation. Quantitative analysis yields forces up to 0.2 pN depending on the concentration of bundling agents.

Sex differences in physiological cardiac hypertrophy are associated with exercise-mediated changes in energy substrate availability.

Exercise-induced cardiac hypertrophy has been recently identified to be regulated in a sex-specific manner. In parallel, women exhibit enhanced exercise-mediated lipolysis compared with men, which might be linked to cardiac responses. The aim of the present study was to assess if previously reported sex-dependent differences in the cardiac hypertrophic response during exercise are associated with differences in cardiac energy substrate availability/utilization. Female and male C57BL/6J mice were challenged with active treadmill running for 1.5 h/day (0.25 m/s) over 4 wk. Mice underwent cardiac and metabolic phenotyping including echocardiography, small-animal PET, peri-exercise indirect calorimetry, and analysis of adipose tissue (AT) lipolysis and cardiac gene expression. Female mice exhibited increased cardiac hypertrophic responses to exercise compared with male mice, measured by echocardiography [percent increase in left ventricular mass (LVM): female: 22.2 ± 0.8%, male: 9.0 ± 0.2%; P < 0.05]. This was associated with increased plasma free fatty acid (FFA) levels and augmented AT lipolysis in female mice after training, whereas FFA levels from male mice decreased. The respiratory quotient during exercise was significantly lower in female mice indicative for preferential utilization of fatty acids. In parallel, myocardial glucose uptake was reduced in female mice after exercise, analyzed by PET {injection dose (ID)/LVM [%ID/g]: 36.8 ± 3.5 female sedentary vs. 28.3 ± 4.3 female training; P < 0.05}, whereas cardiac glucose uptake was unaltered after exercise in male counterparts. Cardiac genes involved in fatty acid uptake/oxidation in females were increased compared with male mice. Collectively, our data demonstrate that sex differences in exercise-induced cardiac hypertrophy are associated with changes in cardiac substrate availability and utilization.

Automated detection of brain atrophy patterns based on MRI for the prediction of Alzheimer's disease.

Subjects with mild cognitive impairment (MCI) have an increased risk to develop Alzheimer's disease (AD). Voxel-based MRI studies have demonstrated that widely distributed cortical and subcortical brain areas show atrophic changes in MCI, preceding the onset of AD-type dementia. Here we developed a novel data mining framework in combination with three different classifiers including support vector machine (SVM), Bayes statistics, and voting feature intervals (VFI) to derive a quantitative index of pattern matching for the prediction of the conversion from MCI to AD. MRI was collected in 32 AD patients, 24 MCI subjects and 18 healthy controls (HC). Nine out of 24 MCI subjects converted to AD after an average follow-up interval of 2.5 years. Using feature selection algorithms, brain regions showing the highest accuracy for the discrimination between AD and HC were identified, reaching a classification accuracy of up to 92%. The extracted AD clusters were used as a search region to extract those brain areas that are predictive of conversion to AD within MCI subjects. The most predictive brain areas included the anterior cingulate gyrus and orbitofrontal cortex. The best prediction accuracy, which was cross-validated via train-and-test, was 75% for the prediction of the conversion from MCI to AD. The present results suggest that novel multivariate methods of pattern matching reach a clinically relevant accuracy for the a priori prediction of the progression from MCI to AD.

Liver-specific peroxisome proliferator-activated receptor alpha target gene regulation by the angiotensin type 1 receptor blocker telmisartan.

OBJECTIVE: The angiotensin type 1 receptor blocker (ARB) and peroxisome proliferator-activated receptor (PPAR) gamma modulator telmisartan has been recently demonstrated to reduce plasma triglycerides in nondiabetic and diabetic hypertensive patients. The present study investigates the molecular mechanisms of telmisartans hypolipidemic actions, in particular its effect on the PPARalpha pathway. RESEARCH DESIGN AND METHODS; Regulation of PPARalpha target genes by telmisartan was studied by real-time PCR and Western immunoblotting in vitro and in vivo in liver/skeletal muscle of mice with diet-induced obesity. Activation of the PPARalpha ligand binding domain (LBD) was investigated using transactivation assays. RESULTS: Telmisartan significantly induced the PPARalpha target genes carnitine palmitoyl transferase 1A (CPT1A) in human HepG2 cells and acyl-CoA synthetase long-chain family member 1 (ACSL1) in murine AML12 cells in the micromolar range. Telmisartan-induced CPT1A stimulation was markedly reduced after small interfering RNA-mediated knockdown of PPARalpha. Telmisartan consistently activated the PPARalpha-LBD as a partial PPARalpha agonist. Despite high in vitro concentrations required for PPARalpha activation, telmisartan (3 mg x kg(-1) x day(-1)) potently increased ACSL1 and CPT1A expression in liver from diet-induced obese mice associated with a marked decrease of hepatic and serum triglycerides. Muscular CPT1B expression was not affected. Tissue specificity of telmisartan-induced PPARalpha target gene induction may be the result of previously reported high hepatic concentrations of telmisartan. CONCLUSIONS: The present study identifies the ARB/PPARgamma modulator telmisartan as a partial PPARalpha agonist. As a result of its particular pharmacokinetic profile, PPARalpha activation by telmisartan seems to be restricted to the liver. Hepatic PPARalpha activation may provide an explanation for telmisartan's antidyslipidemic actions observed in recent clinical trials.

Enhancing instance-based classification with local density: a new algorithm for classifying unbalanced biomedical data.

MOTIVATION: Classification is an important data mining task in biomedicine. In particular, classification on biomedical data often claims the separation of pathological and healthy samples with highest discriminatory performance for diagnostic issues. Even more important than the overall accuracy is the balance of a classifier, particularly if datasets of unbalanced class size are examined. RESULTS: We present a novel instance-based classification technique which takes both information of different local density of data objects and local cluster structures into account. Our method, which adopts the basic ideas of density-based outlier detection, determines the local point density in the neighborhood of an object to be classified and of all clusters in the corresponding region. A data object is assigned to that class where it fits best into the local cluster structure. The experimental evaluation on biomedical data demonstrates that our approach outperforms most popular classification methods. AVAILABILITY: The algorithm LCF is available for testing under http://biomed.umit.at/upload/lcfx.zip.

Functional cluster analysis of CT perfusion maps: a new tool for diagnosis of acute stroke?

CT perfusion imaging constitutes an important contribution to the early diagnosis of acute stroke. Cerebral blood flow (CBF), cerebral blood volume (CBV) and time-to-peak (TTP) maps are used to estimate the severity of cerebral damage after acute ischemia. We introduce functional cluster analysis as a new tool to evaluate CT perfusion in order to identify normal brain, ischemic tissue and large vessels. CBF, CBV and TTP maps represent the basis for cluster analysis applying a partitioning (k-means) and density-based (density-based spatial clustering of applications with noise, DBSCAN) paradigm. In patients with transient ischemic attack and stroke, cluster analysis identified brain areas with distinct hemodynamic properties (gray and white matter) and segmented territorial ischemia. CBF, CBV and TTP values of each detected cluster were displayed. Our preliminary results indicate that functional cluster analysis of CT perfusion maps may become a helpful tool for the interpretation of perfusion maps and provide a rapid means for the segmentation of ischemic tissue.

Modelling of classification rules on metabolic patterns including machine learning and expert knowledge.

Machine learning has a great potential to mine potential markers from high-dimensional metabolic data without any a priori knowledge. Exemplarily, we investigated metabolic patterns of three severe metabolic disorders, PAHD, MCADD, and 3-MCCD, on which we constructed classification models for disease screening and diagnosis using a decision tree paradigm and logistic regression analysis (LRA). For the LRA model-building process we assessed the relevance of established diagnostic flags, which have been developed from the biochemical knowledge of newborn metabolism, and compared the models' error rates with those of the decision tree classifier. Both approaches yielded comparable classification accuracy in terms of sensitivity (>95.2%), while the LRA models built on flags showed significantly enhanced specificity. The number of false positive cases did not exceed 0.001%.