Ammonolysis of Glyoxal at the Air-Water Nanodroplet Interface.

The reactions of glyoxal (CHO)2 ) with amines in cloud processes contribute to the formation of brown carbon and oligomer particles in the atmosphere. However, their molecular mechanisms remain unknown. Herein, we investigate the ammonolysis mechanisms of glyoxal with amines at the air-water nanodroplet interface. We identified three and two distinct pathways for the ammonolysis of glyoxal with dimethylamine and methylamine by using metadynamics simulations at the air-water nanodroplet interface, respectively. Notably, the stepwise pathways mediated by the water dimer for the reactions of glyoxal with dimethylamine and methylamine display the lowest free energy barriers of 3.6 and 4.9 kcal ⋅ mol-1 , respectively. These results showed that the air-water nanodroplet ammonolysis reactions of glyoxal with dimethylamine and methylamine were more feasible and occurred at faster rates than the corresponding gas phase ammonolysis, the OH+(CHO)2 reaction, and the aqueous phase reaction of glyoxal, leading to the dominant removal of glyoxal. Our results provide new and important insight into the reactions between carbonyl compounds and amines, which are crucial in forming nitrogen-containing aerosol particles.

Simple Death Risk Models to Predict In-hospital Outcomes in Acute Aortic Dissection in Emergency Department.

Objective: We sought to find a bedside prognosis prediction model based on clinical and image parameters to determine the in-hospital outcomes of acute aortic dissection (AAD) in the emergency department. Methods: Patients who presented with AAD from January 2010 to December 2019 were retrospectively recruited in our derivation cohort. Then we prospectively collected patients with AAD from January 2020 to December 2021 as the validation cohort. We collected the demographics, medical history, treatment options, and in-hospital outcomes. All enrolled patients underwent computed tomography angiography. The image data were systematically reviewed for anatomic criteria in a retrospective fashion by three professional radiologists. A series of radiological parameters, including the extent of dissection, the site of the intimal tear, entry tear diameter, aortic diameter at each level, maximum false lumen diameter, and presence of pericardial effusion were collected. Results: Of the 449 patients in the derivation cohort, 345 (76.8%) were male, the mean age was 61 years, and 298 (66.4%) had a history of hypertension. Surgical repair was performed in 327 (72.8%) cases in the derivation cohort, and the overall crude in-hospital mortality of AAD was 10.9%. Multivariate logistic regression analysis showed that predictors of in-hospital mortality in AAD included age, Marfan syndrome, type A aortic dissection, surgical repair, and maximum false lumen diameter. A final prognostic model incorporating these five predictors showed good calibration and discrimination in the derivation and validation cohorts. As for type A aortic dissection, 3-level type A aortic dissection clinical prognosis score (3ADPS) including 5 clinical and image variables scored from -2 to 5 was established: (1) moderate risk of death if 3ADPS is <0; (2) high risk of death if 3ADPS is 1-2; (3) very high risk of death if 3ADPS is more than 3. The area under the receiver operator characteristic curves in the validation cohorts was 0.833 (95% CI, 0.700-0.967). Conclusion: Age, Marfan syndrome, type A aortic dissection, surgical repair, and maximum false lumen diameter can significantly affect the in-hospital outcomes of AAD. And 3ADPS contributes to the prediction of in-hospital prognosis of type A aortic dissection rapidly and effectively. As multivariable risk prediction tools, the risk models were readily available for emergency doctors to predict in-hospital mortality of patients with AAD in extreme clinical risk.

Cascaded deep transfer learning on thoracic CT in COVID-19 patients treated with steroids.

Purpose: Given the recent COVID-19 pandemic and its stress on global medical resources, presented here is the development of a machine intelligent method for thoracic computed tomography (CT) to inform management of patients on steroid treatment. Approach: Transfer learning has demonstrated strong performance when applied to medical imaging, particularly when only limited data are available. A cascaded transfer learning approach extracted quantitative features from thoracic CT sections using a fine-tuned VGG19 network. The extracted slice features were axially pooled to provide a CT-scan-level representation of thoracic characteristics and a support vector machine was trained to distinguish between patients who required steroid administration and those who did not, with performance evaluated through receiver operating characteristic (ROC) curve analysis. Least-squares fitting was used to assess temporal trends using the transfer learning approach, providing a preliminary method for monitoring disease progression. Results: In the task of identifying patients who should receive steroid treatments, this approach yielded an area under the ROC curve of 0.85 ± 0.10 and demonstrated significant separation between patients who received steroids and those who did not. Furthermore, temporal trend analysis of the prediction score matched expected progression during hospitalization for both groups, with separation at early timepoints prior to convergence near the end of the duration of hospitalization. Conclusions: The proposed cascade deep learning method has strong clinical potential for informing clinical decision-making and monitoring patient treatment.

RPW: a hybrid reverse prediction method for level of detail.

In this paper, we propose a way of using multiple domain analysis methodology to speed up the image rendering process. We first apply wavelet transform to the original image, and then compress the wavelets in the frequency domain using histogram equalization techniques. When rendering the image, we uncompress the wavelets and reverse predict the upper level images. This process continues until it reaches a certain criteria. We use two terms-total image size (TIS) and total loading time (TLT) to measure the performance of level of detail (LOD) in a network environment. We compare traditional image-based LOD methods with the new method we are proposing. Experiments show that the proposed method can reduce both TIS and TLT. The image rendering speed on a remote client is approximately 2.5 times faster than the common image compression methods. Applications such as remote diagnostic systems and online museums can use this technique to achieve better real-time animation effects.

Processing X-ray images to eliminate irrelevant structures that mask important features.

In a plane radiographic image, there generally is an important area of interest (AOI). Too often, the AOI is partially masked by images of other overlapping and underlying structures that may be in front of or behind the AOI. An important adjunct to radiological diagnosis would be the capability of eliminating images of such masking structures to isolate the AOI for more detailed examination. We described a computerized method that utilizes a stereo pair of plane X-ray images to enable radiologists to interact with these images for first identifying for the computer the AOI and then directing the computer to eliminate all structures in front of and behind the AOI. The result is a plane X-ray image or a stereo X-ray image pair that includes only the AOI, but not any overlapping or underlying structures. The method uses a stereo pair of X-rays and the 3D perception of radiologists. 3D perception involves eye convergence and lens focus as well as cues, such as parallax and relative sizes. Convergence of the eyes is by far the strongest factor in 3D visualization. The horizontal separation or disparity between points in the left and right eye images on a screen or X-ray film produces convergence which determines an object's perceived depth in visual 3D space. All points in a given perceived depth plane have the same disparity on the screen. In theory, a given depth plane can be eliminated from the 3D image by shifting one image and then the other image of a stereo pair horizontally by the distance of the disparity of the depth plane, and subtracting. A new stereo image pair is thereby produced in which points only of the depth plane do not appear. However, in practical situations, certain artifacts arise that must be considered. The method has the potential for important applications in many areas of medical imaging processing.