Search

This paper has developed a Finite Element Method (FEM) to calculate the dynamic response of a continuous sandwich beam resting on elastic support subjected to moving vehicles. The equation of motion is derived using the classical beam theory and FEM. The vehicle model is a two Degree of Freedom (2DOF) system that moves with a constant velocity. The governing equation of motion is integrated by applying the Wilson-θ time integration method to obtain the dynamic response in each time step. Numerical examples investigate the displacement of the sandwich beam with various values of the structure and vehicle velocity. The effects of the stiffness of elastic support and the vehicle velocity on displacement are studied.

Large language models (LLMs) have significantly advanced the field of natural language processing, with GPT models at the forefront. While their remarkable performance spans a range of tasks, adapting LLMs for real-world business scenarios still poses challenges warranting further investigation. This paper presents an empirical analysis aimed at bridging the gap in adapting LLMs to practical use cases. To do that, we select the question answering (QA) task of insurance as a case study due to its challenge of reasoning. Based on the task we design a new model relied on LLMs which are empowered by additional knowledge extracted from insurance policy rulebooks and DBpedia. The additional knowledge helps LLMs to understand new concepts of insurance for domain adaptation. Preliminary results on two QA datasets show that knowledge enhancement significantly improves the reasoning ability of GPT-3.5 (55.80% and 57.83% in terms of accuracy). The analysis also indicates that existing public knowledge bases, e.g., DBPedia is beneficial for knowledge enhancement. Our findings reveal that the inherent complexity of business scenarios often necessitates the incorporation of domain-specific knowledge and external resources for effective problem-solving., Ongoing work to adapt LLMs for business scenarios

Nguyen Duy Hung,1,2,&ast; Nguyen-Thi Tam,2 Dang Khanh Huyen,2 Nguyen-Van Thi,3 Nguyen Minh Duc4,&ast; 1Department of Radiology, Viet Duc Hospital, Hanoi, Vietnam; 2Department of Radiology, Hanoi Medical University, Hanoi, Vietnam; 3Department of Radiology, Vietnam National Cancer Hospital, Hanoi, Vietnam; 4Department of Radiology, Pham Ngoc Thach University of Medicine, Ho Chi Minh City, Vietnam&ast;These authors contributed equally to this workCorrespondence: Nguyen Minh Duc, Department of Radiology, Pham Ngoc Thach University of Medicine, 2 Duong Quang Trung Ward 12 District 10, Ho Chi Minh City, 700000, Vietnam, Email bsnguyenminhduc@pnt.edu.vnObjective: The study aimed to evaluate the role of magnetic resonance imaging (MRI) in differentiating between primary benign and malignant soft tissue tumors (STTs).Patients and Methods: The study was carried out on 110 patients with histopathological diagnoses of STTs. All patients underwent routine MRI before surgery/biopsy at Viet Duc University Hospital or Vietnam National Cancer Hospital, Hanoi, Vietnam, from January 2020 to October 2022. Data on preoperative MRI as well as the clinical features and pathological results of the patients were collected retrospectively. Univariate and multivariate linear regression were used to analyze the relationship between imaging, clinical parameters, and the ability to differentiate malignant from benign STTs.Results: Among 110 patients (59 men and 51 women), 66 had benign tumors and 44 had malignant tumors. The qualitative values that were significant in distinguishing between benign and malignant STTs were hypointensity on T1-weighted images (T1W; p< 0.001), hypointensity on T2-weighted images (T2W; p=0.003), cysts (p=0.003)), necrosis (p< 0.001), fibrosis (p=0.023), hemorrhage (p< 0.001), lobulated margin (p< 0.001), ill-defined border (p< 0.001), peritumoral edema (p< 0.001), vascular involvement (p< 0.001), and heterogeneous enhancement (p< 0.001). Regarding quantitative values, age (p=0.009), size (p< 0.001), T1W signal quantification value (p=0.002), and T2W signal quantification value (p=0.007) showed statistically significant differences between benign and malignant tumors. Multivariate linear regression analysis showed that the combination of peritumoral edema and heterogeneous enhancement was the most valuable in the differential diagnosis of malignant tumors from benign tumors.Conclusion: MRI is valuable in discriminating between malignant and benign STTs. The presence of cysts, necrosis, hemorrhage, lobulated margin, ill-defined border, peritumoral edema, heterogeneous enhancement, vascular involvement, and T2W hypointensity is suggestive of malignant lesions, especially signs of peritumoral edema and heterogeneous enhancement. Advanced age and large tumor size are also suggestive of soft tissue sarcomas.Keywords: soft tissue tumors, soft tissue sarcoma, magnetic resonance imaging, tissue characterization, size, prediction

Gliosarcoma (GS) is an uncommon central nervous system tumor with several characteristics of a malignant neoplasm and poor prognosis. The majority of GS reports describe a predilection for the cerebral hemispheres, and cases of intraventricular GS are extremely rare, with only a few reported. In addition, intraventricular GS has not been associated with any unique radiographic or clinical features, which can result in misdiagnosis as other intraventricular tumor types. In this report, we present the case of a 32-year-old woman with GS in the trigone of the lateral ventricle and provide a retrospective review of similar, previously reported cases.

Today AI systems are rarely made without Machine Learning (ML) and this inspires us to explore what aptly called composite argumentation systems with ML components. Concretely, against two theoretical backdrops of PABA (Probabilistic Assumption-based Argumentation) and DST (Dempster-Shafer Theory), we present a framework for such systems called c-PABA. It is argued that c-PABA lends itself to a development tool as well and to demonstrate we show that DST-based ML classifier combination and multi-source data fusion can be implemented as simple c-PABA frameworks.

Chordoid gliomas are uncommon neoplasms located within the anterior portion of the third ventricle. In this article, we aimed to describe the clinical presentation, magnetic resonance imaging characteristics, histological findings, and surgical treatment applied to a case of chordoid glioma. Chordoid gliomas are typically observed as solid masses within the anterior segment of the third ventricle, characterized by homogenous and vivid enhancement. Despite being classified as a low-grade neoplasm, the outcome of choroid glioma is often uncertain.

Introduction: Neurocysticercosis (NCC), a major contributor to the burden of seizure disorders and epilepsy in the world, is one of the most common parasitic infections of the central nervous system, which is usually located in the brain. Medical therapy for NCC should initially cover appropriate symptom control and then the use of antiparasitic agents should be considered. Antiparasitic treatment is of benefit in most cases of viable and degenerating NCC. Nevertheless, cysticercosis of the spinal cord is very uncommon. Case series: In this article, we recorded 5 cases of extramedullary-intradural lumbar spinal cysticercosis, in which one case displayed cystic lesions in both brain and spine, one case showed an independent cystic lesion in the spine, and three remaining cases showed diffuse lesions in the spinal canal. Conclusion: Thus, in any case of single or numerous cystic lesions or dispersed lesions entering the spinal canal, spinal cysticercosis should be considered.

Argumentative agents in AI are inspired by how humans reason by exchange of arguments. Given the same set of arguments possibly attacking one another (Dung's AA framework) these agents are bound to accept the same subset of those arguments (aka extension) unless they reason by different argumentation semantics. However humans may not be so predictable, and in this paper we assume that this is because any real agent's reasoning is inevitably influenced by her own preferences over the arguments. Though such preferences are usually unobservable, their effects on the agent's reasoning cannot be washed out. Hence by reconstructing her reasoning process, we might uncover her hidden preferences, which then allow us to predict what else the agent must accept. Concretely we formalize and develop algorithms for such problems as uncovering the hidden argument preference relation of an agent from her expressed opinion, by which we mean a subset of arguments or attacks she accepted from a given AA framework; and uncovering the collective preferences of a group from a dataset of individual opinions. A major challenge we addressed in this endeavor is to deal with “answer sets” of argument preference relations which are generally exponential or even infinite. So we start by developing a compact representation for such answer sets called preference states. Preference revelation tasks are then structured as derivations of preference states from data, and reasoning prediction tasks are reduced to manipulations of derived preference states without enumerating the underlying (possibly infinite) answer sets. We also apply the presented results to two non-trivial problems: learning preferences over rules in structured argumentation with priorities – an open problem so far; and analyzing public polls in apparently deeper ways than existing social argumentation frameworks allow.

Nguyen Duy Hung,1,2,* Vuong Kim Ngan,3 Nguyen Minh Duc4,5,* 1Department of Radiology, Hanoi Medical University, Hanoi, Vietnam; 2Department of Radiology, Viet Duc Hospital, Hanoi, Vietnam; 3Department of Radiology, Vinmec Times City International Hospital, Hanoi, Vietnam; 4Department of Radiology, Pham Ngoc Thach University of Medicine, Ho Chi Minh City, Vietnam; 5Department of Radiology, Children’s Hospital 2, Ho Chi Minh City, Vietnam*These authors contributed equally to this workCorrespondence: Nguyen Minh DucDepartment of Radiology, Pham Ngoc Thach University of Medicine, Ho Chi Minh City, VietnamEmail bsnguyenminhduc@pnt.edu.vnBackground: Craniopharyngiomas are common lesions that occur in the suprasellar region; however, strictly intrinsic third ventricular craniopharyngiomas are rare.Case Series: We aimed to describe the magnetic resonance imaging features observed in five cases of strictly intrinsic third ventricular papillary craniopharyngiomas, including two cases of mixed cystic and solid tumors and three cases of pure solid masses.Conclusion: Among the adult population, intrinsic third ventricular papillary craniopharyngiomas should be considered when either solid or mixed cystic and solid masses are observed, in which the solid component shows heterogeneous intensity, heterogeneous and strong enhancement, and is strictly located in the third ventricle.Keywords: strictly intrinsic third ventricle, craniopharyngioma, magnetic resonance imaging

Although it has been extensively studied for decades, the α-Al2O3:Cr3+ phosphor has rarely been investigated for horticultural lighting. In this work, for the first time, a prototype of a plant growth light-emitting diode (LED) has been fabricated by coating a deep-red-emitting α-Al2O3:Cr3+ phosphor onto a near-ultraviolet (NUV) chip. The α-Al2O3:Cr3+ phosphor, synthesized by a co-precipitation method and annealed at 1500 °C for 2 h, emits an outstanding narrow peak at 695 nm. The α-Al2O3:0.6%Cr3+ phosphor shows a high activation energy of 0.29 eV, a long lifetime of 3.4 ms, and a superior color purity of 100%. The chromatic coordinates and the QE value of the red LED obtained by coating an α-Al2O3:0.6%Cr3+ phosphor on a NUV chip are (x = 0.5650, y = 0.2429) and 87.1%, respectively.

The development of full-visible-spectrum phosphors is essential for next-generation light-emitting devices with better light quality. Herein, we report on a novel broad-band-emitting phosphor based on single-composition Al-doped ZnO phosphors. Under the UV excitation of 325 nm, the ZnO : Al phosphor exhibits a full spectrum emission in the visible wavelength range from 400 to 800 nm with a CIE chromaticity coordinate of (0.42, 0.48), a quantum efficiency of 43%, a color rendering index (CRI) of 74, a correlated color temperature (CCT) value of 3873 K and an activation energy of 0.22 eV. A prototype of a UV-pumped warm WLED with a high CRI of 87 and a CCT of 4067 K has been achieved by using only this broad-band-emitting Al3+-doped ZnO phosphor. The obtained results indicate that the single-composition Al3+-singly doped warm white emitting phosphor is a promising candidate for UV-pump warm white light-emitting diodes.

Introduction: Tractor semi-trailer vehicles are playing an increasingly important role in the global freight chain. However, due to the heavy total load and height of the center of gravity, this type of vehicle is often at a higher risk of instability than other vehicles. This paper focuses on improving the vehicle roll stability by using an active anti-roll bar system. Methods: The Linear Quadratic Regulator (LQR) approach is used for this purpose with the control signal being the torque generated by the active anti-roll bar system. In order to synthesize the controller, the roll angle of the vehicle body and the normalized load transfer at all axles of the tractor semi-trailer vehicle are considered as the optimal goals. Results: The simulation results in time and frequency domains clearly show the effectiveness of the proposed method for the active anti-roll bar system, because the reduction of the desired criterias is about 40% less when compared to a vehicle using the passive anti-roll bar system. Conclusions: The effectiveness of the active anti-roll bar system on improving the vehicle roll stability, has been verified in this theoretical study with the LQR optimal controller. This is an important basis for conducting more in-depth studies and future experiments.

Background: This study aimed to evaluate the diagnostic function of 3-Tesla (T) magnetic resonance imaging (MRI) during the assessment of brachial plexus injury (BPI), in comparison with intraoperative findings. Methods: A retrospective study was performed on 60 patients (47 men and 13 women), who had clinical manifestations of BPI, underwent 3T MRI of the brachial plexus, and were surgically treated at the Viet Duc and Vinmec Times City hospitals, in Hanoi, Vietnam, from March 2016 to December 2019. Preganglionic and postganglionic lesion features were identified on MRI. The diagnostic function of MRI features for the determination of BPI was evaluated and correlated with intraoperative findings. Results: The root avulsion and pseudomeningocele preganglionic injuries were observed in 57% and 43% of MRIs, respectively, and were commonly observed at the C7 and C8 roots. Nerve disruption and never edema were observed in 47.56% and 33.53% of MRIs, respectively, and were commonly observed at the C5 and C6 roots. The sensitivity, specificity, accuracy, positive prognostic value, and negative prognostic value of 3T MRI were 64.12%, 92.90%, 80.33%, 87.50%, and 76.96%, respectively, for the diagnosis of total avulsion, and 68.52%, 83.33%, 80.67%, 47.44%, and 92.34%, respectively, for the diagnosis of nerve disruption. Conclusion: MRI offers valuable details regarding the location, morphology, and severity of both preganglionic and postganglionic injuries during the preoperative diagnosis of BPI. However, this modality played a moderate diagnostic role. Therefore, 3T MRI should be used as a supplemental evaluation, coupled with clinical tests and electromyography, to determine the most appropriate treatment strategies for BPI patients.

Background/Aim: Hirayama disease, a rare neurological disorder, is characterized by progressive, asymmetric muscle atrophy of the forearm and hand in young male patients. Case Report: We noted two cases of Hirayama disease that appeared in the right upper limb and identified the cervical neutral and flexion position magnetic resonance imaging (MRI) characteristics. Conclusion: In the flexion posture, the cervical MRI allows early diagnosis of Hirayama disease and should be considered in patients with the typical clinical symptoms.

Objectives: This study compared the diagnostic value of 3D T1-weighted (T1W) gradient-echo (GRE) and 2D T1W in-phase and out-of-phase GRE sequences for appendicitis diagnosis in pregnant women. Material and Methods: This retrospective study included 25 pregnant patients with suspected appendicitis who underwent 1.5 Tesla abdominal magnetic resonance imaging and had definitive diagnoses. Four doctors approached four separate imaging groups: A (only T2-weighted image [T2WI] sequences), B (T2WI and 3D T1W GRE sequences), C (T2WI and T1W in-phase and out-of-phase GRE sequences), and D (T2WI, 3D T1W GRE, and T1W in-phase and out-of-phase GRE sequences). The kappa (κ) index was used to compare the appendicitis diagnostic results between groups. The diagnostic value of these sequences in the diagnosis of pregnancy appendicitis was also evaluated. Results: Groups A and C had average consistency with definitive diagnosis (κ = 0.6), lower than Groups B and D (κ = 0.865), indicating a high definite diagnosis consistency. Groups B and D had similarly high sensitivity (80%), specificity (100%), positive predictive value (100%), negative predictive value (95.2%), and accuracy (ACC) (96%), higher than Groups A and C (60%, 95%, 75%, 90.5%, and 88%, respectively). Conclusion: 3D T1W-GRE sequences improve appendicitis diagnosis in pregnancy compared to T2W sequences alone. Adding in and out phase GRE sequences do not increase diagnostic ACC.

Background Medulloblastoma is well known as the most common malignant brain tumor identified in children, frequently found at an intra-axial location in the posterior cranial fossa. Extra-axial medulloblastoma is uncommon and often misdiagnosed. We believe that a thorough understanding of atypical medulloblastoma cases is important in daily practice. Case presentation We present the unique case of a 39-year-old woman of Asian descent who suffered from headaches and right-sided hearing impairment. A right extra-axial medulloblastoma with an extremely low apparent diffusion coefficient of 0.404 × 10−3 mm2/second was detected on magnetic resonance imaging. The initial diagnosis suggested schwannoma or hemangioblastoma. However, the postoperative histopathologic findings indicated medulloblastoma (World Health Organization grade IV). Pre- and postoperative magnetic resonance imaging revealed no drop metastasis, but adjuvant radiation therapy was still required as a standard treatment therapy Conclusions Extra-axial medulloblastoma is an uncommon tumor that is often mistaken for other cerebellopontine angle neoplasms. We describe a rare example of extra-axial medulloblastoma, characterized by a low apparent diffusion coefficient. When evaluating an atypical cerebellopontine angle neoplasm, the apparent diffusion coefficient should be considered a relevant indicator.

In this study, we synthesized K3AlF6:Mn4+ phosphor by co-precipitation method. The crystal structure and photoluminescence properties were investigated. By surface modification of K3AlF6:Mn4+ using K3AlF6, the moisture resistance performances of the phosphor can be significantly improved. It was found that the luminescence performance of K3AlF6:Mn4+@K3AlF6, which was dispersed in water for 2 h, is unchanged but the uncoated sample reduces dramatically. White light emitting diodes (WLEDs) based on the phosphor combined with commercial YAG:Ce3+ coated on a blue LED showed significantly improved of performance with color correlated temperature (CCT) from 5307 K down to 3528 K and color rendering index (CRI) from 64 up to 87. The results exhibit the potential for the application of K3AlF6:Mn4+@K3AlF6 as a red phosphor in warm WLEDs.

The aim of this paper is to estimate the bulk modulus of composite containing spherical inclusion surrounded by an interphase whose elastic properties vary with radius. Based on the sphere assemblage model and the differential substitution scheme, the differential equation for the bulk modulus is established. The effective bulk modulus of the material are predicted by numerical integration. The effect of interphase zone is also investigated.

Although it has been extensively studied for decades, the α-Al

Gd2O3:Er3+ nanophosphors were fabricated by the combustion method in presence of Na2 ethylene diamine tetra acetic acid (EDTA-Na2) as fuel at not high temperature (≤350 oC) within a very short time of 5 min. The added concentration of Er3+ ions in Gd2O3 matrix was changed from 0.5 mol% to 5.0 mol%. The X-ray diffraction pattern of samples indicates the monoclinic structure of Gd2O3:Er3+. The morphology and chemical composition analysis of the Gd2O3:Er3+ samples are characterized by a field emission scanning electron microscope (FESEM) and a Fourier-transform infrared spectrometer (FTIR). The photoluminescence (PL), photoluminescence excitation (PLE) and upconversion (UC) at room temperature of the prepared materials with different concentrations of Er3+ were investigated. The PL of Gd2O3:Er3+ nanomaterials are shown in visible at 545, 594, 623, 648, 688 nm under excitation at 275 nm. The emission bands from transitions of Er3+ from 2P3/2 to 4F9/2 are observed. UC luminescent spectra of the Gd2O3:Er3+/silica nanocomposites under 976 nm excitation show the bands at 548 and 670 nm. The influence of excitation power at 980 nm for transitions were measured and calculated. The results indicate that the upconversion process of Gd2O3:Er3+/silica is two photons absorption mechanism. The low temperature dependence of UC luminescent intensities of the main bands of Gd2O3:Er3+ was investigated towards development of a nanotemperature sensor in the range of 10–300 K.

Analogical reasoning is a complex process based on a comparison between two pairs of concepts or states of affairs (aka. the source and the target) for characterizing certain features from one to a...

This paper investigates the variability of dynamic responses of a beam resting on an elastic foundation, which is subjected to a vehicle with uncertain parameters, such as random mass, stiffness, damping of the vehicle and random fields of mass density, and the elastic modulus of the beam and stiffness of elastic foundation. The vehicle is modeled as a two-degree-of-freedom spring-damper-mass system. The equations of motion of the beam was constructed using a finite element method. The mass and elastic properties of the beam, and the stiffness of foundation are assumed to be Gaussian random fields and were simulated by the spectral represent method. Masses, stiffness of the spring, and the damping coefficient of the vehicle are assumed as Gaussian random variables. The numerical analyses were performed using the finite element method (FEM) in conjunction with the Monte Carlo simulation (MCS). The variability of dynamic responses of the beam were investigated with various cases of random parameters. For each sample, the equations of motions were solved with the Wilson-q integral method to find dynamic responses. The influence of random system parameters and their correlation on the response variability is discussed in detail.

Ureteral fibroepithelial polyps are benign, mesodermal neoplasms that can cause upper urinary tract obstruction. The preoperative diagnosis of this disease is challenging, but multidetector computed tomography (CT), particularly CT urography in the excretory phase, can be used to detect the features of ureteral polyps. We illustrated a case preoperatively diagnosed as upper ureteral tract obstruction due to ureteral tumor based on clinical presentation and CT imaging. However, the histopathological report revealed ureteral fibroepithelial polyps. Ureteral fibroepithelial polyps should be considered as a rare cause of urinary obstruction when the clinical presentation and imaging findings are atypical for more common etiologies.

Consumer preference studies in economics rest heavily on the behavioural interpretation of preference especially in the form of Revealed Preference Theory (RPT). Viewing purchasing decisions as a species of human reasoning, in this paper we are interested in generalising behaviourism to preference-based argumentation where existing frameworks are universally governed by the opposing mentalistic interpretation of preference. Concretely we re-construct and unify two main approaches to RPT then develop a so-called Revealed Preference Argumentation (RPA) framework which identifies preference as observed reasoning behaviour of an agent. We show that RPA subsumes RPT, by showing that key RPT-based consumer analyses can be translated to and solved as RPA computational tasks. It is argued that RPA may pave the way for future applications of argumentation to behavioural economics.

Automatic summarization of legal texts is an important and still a challenging task since legal documents are often long and complicated with unusual structures and styles. Recent advances of deep models trained end-to-end with differentiable losses can well-summarize natural text, yet when applied to legal domain, they show limited results. In this paper, we propose to use reinforcement learning to train current deep summarization models to improve their performance on the legal domain. To this end, we adopt proximal policy optimization methods and introduce novel reward functions that encourage the generation of candidate summaries satisfying both lexical and semantic criteria. We apply our method to training different summarization backbones and observe a consistent and significant performance gain across 3 public legal datasets.

Background: Central nervous system (CNS) lymphoma presents as the dense infiltration of tumor cells in the perivascular space and blood-brain barrier disruption, on histopathological examination. The Ki-67 expression has been significantly correlated with tumor proliferation and is considered to be a prognostic factor. Objectives: This study aimed at analyzing the correlations among the apparent diffusion coefficient (ADC), the relative cerebral blood volume (rCBV), and the Ki-67 proliferation index in CNS lymphoma. Methods: From August 2019 to March 2020, 26 patients (14 men and 12 women) who underwent biopsy or surgery and were histologically confirmed as CNS lymphoma were included in this retrospective study. Diffusion and perfusion acquisitions were performed in 26 and 10 examinations, respectively. The Ki-67 proliferation index was available for all cases. Results: The mean tADC, rADC, and rCBV values were 0.61 ± 0.12 × 10-3 mm2/s, 0.73 ± 0.14, and 1.1 ± 0.32, respectively. Negative correlations were identified between both tADC and rADC and the Ki-67 proliferation index (r = -0.656, P < 0.01 and r = -0.540, P < 0.01, respectively). No significant correlations were found between rCBV values and the Ki-67 proliferation index, between rCBV and rADC, or between rCBV and tADC. Conclusions: tADC and rADC values can be used as noninvasive indicators to predict cell proliferation in CNS lymphoma.

Abdominal solid organ trauma accounts for 86.9% of all blunt abdominal trauma cases, among which 5%-25% of cases are associated with arterial injuries. Aim: In this study, we aimed to determine the characteristics and diagnostic functions of computed tomography (CT) scans during the diagnosis of arterial injuries, including active extravasation, pseudoaneurysm, and arteriovenous fistula associated with solid organ trauma compared with digital subtraction angiography (DSA).From July 2019 to March 2020, a retrospective study was performed on 44 patients who presented with clinical manifestations of blunt abdominal injury and CT diagnosis of arterial injuries due to solid organ trauma, including active extravasation, pseudoaneurysm, and arteriovenous fistula, and later underwent DSA at Vietduc Hospital, in Hanoi, Vietnam. The features of arterial lesions on CT scan was described. Value of CT scan for the diagnosis of arterial injuries was analyzed compared to the DSA findings.There were 53 arterial lesions observed on CT scan including 15 active extravasations, 34 pseudoaneurysms, and 4 arteriovenous fistulas while 51 arterial lesions were observed on DSA including 15 active extravasations, 30 pseudoaneurysms, and 6 arteriovenous fistulas. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of active extravasation, pseudoaneurysm, and arteriovenous fistula on CT scan were 93.3%, 97.7%, 93.3%, 97.7% and 96.6% ; 90%, 75%, 79.4%, 87.5% and 82.8%; and 66.7%, 100%, 100%, 96.3% and 96.6%, respectively.Our study showed that CT scans had high sensitivity and specificity for the diagnosis of active extravasation; however, CT scans demonstrated low specificity for detecting pseudoaneurysm and low sensitivity for the diagnosis of arteriovenous fistula.

There are still few works on application of deep learning for Myanmar language. This paper presents an approach to use a convolutional neural network (CNN) model to classify sentence sentiment in Myanmar texts. A CNN model is constructed on the top of a word embedding model (i.e., Word2Vec), which converts words into vectors. The model classifies the input sentences and labels each sentence with positive, negative, neutral, unrelated and unreadable sentiments. The model is learnt from 1,152 sentences taken from the customers' reviews of products, which is provided by a telecommunication company. Then, the model is tested on 495 unseen sentences with the result of 86.26% accuracy and 82.58% average f-measure in prediction. The model is compared with the traditional machine learning (ML) classifiers, especially support vector machine (SVM), naive Bayes (NB), and logistic regression (LR). The model outperforms these classifiers since SVM results in 64.44% accuracy, NB obtains 60.20% accuracy and LR gets 55.15% accuracy.