Multi-modal trained artificial intelligence solution to triage chest X-ray for COVID-19 using pristine ground-truth, versus radiologists.

The front-line imaging modalities computed tomography (CT) and X-ray play important roles for triaging COVID patients. Thoracic CT has been accepted to have higher sensitivity than a chest X-ray for COVID diagnosis. Considering the limited access to resources (both hardware and trained personnel) and issues related to decontamination, CT may not be ideal for triaging suspected subjects. Artificial intelligence (AI) assisted X-ray based application for triaging and monitoring require experienced radiologists to identify COVID patients in a timely manner with the additional ability to delineate and quantify the disease region is seen as a promising solution for widespread clinical use. Our proposed solution differs from existing solutions presented by industry and academic communities. We demonstrate a functional AI model to triage by classifying and segmenting a single chest X-ray image, while the AI model is trained using both X-ray and CT data. We report on how such a multi-modal training process improves the solution compared to single modality (X-ray only) training. The multi-modal solution increases the AUC (area under the receiver operating characteristic curve) from 0.89 to 0.93 for a binary classification between COVID-19 and non-COVID-19 cases. It also positively impacts the Dice coefficient (0.59 to 0.62) for localizing the COVID-19 pathology. To compare the performance of experienced readers to the AI model, a reader study is also conducted. The AI model showed good consistency with respect to radiologists. The DICE score between two radiologists on the COVID group was 0.53 while the AI had a DICE value of 0.52 and 0.55 when compared to the segmentation done by the two radiologists separately. From a classification perspective, the AUCs of two readers was 0.87 and 0.81 while the AUC of the AI is 0.93 based on the reader study dataset. We also conducted a generalization study by comparing our method to the-state-art methods on independent datasets. The results show better performance from the proposed method. Leveraging multi-modal information for the development benefits the single-modal inferencing.

Oxidative impairment and histopathological alterations in kidney and brain of mice following subacute lambda-cyhalothrin exposure.

Lambda cyhalothrin (LCT), a broad-spectrum type II (α-cyano) synthetic pyrethroid pesticide, is widely employed in various agricultural and animal husbandry practices for the control of pests. Acute and chronic exposure to LCT can elicit several adverse effects including oxidative stress. With the objective to investigate nephrotoxicity and neurotoxicity of LCT in mice, we evaluated oxidative stress parameters and histological changes in the kidney and brain of LCT exposed mice. Swiss albino mice were divided randomly into four groups ( n = 6 per group) as: (A) corn oil/vehicle control; (B) 0.5 mg/kg body weight (b.w.) LCT; (C) 1 mg/kg b.w. LCT; (D) 2 mg/kg b.w. LCT. Mice were treated orally for 28 days. LCT exposure significantly increased serum urea nitrogen, creatinine and urea levels. LCT exposure also increased lipid peroxidation, superoxide anion generation, nitrite level and decreased the level of reduced glutathione. The activities of superoxide dismutase, catalase and glutathione- S-transferase were depleted significantly in both kidney and brain. Histological examination revealed marked histopathological changes in the kidney and brain of mice that were more pronounced at high dose of LCT. Thus, results of the present study indicate that 28 days oral exposure of LCT causes oxidative damage to the kidney and brain of mice which in turn could be responsible for nephrotoxicity and neurotoxicity. Nevertheless, further detailed studies are required to prove these effects especially after long-term exposure.

Oral nanoparticulate curcumin combating arsenic-induced oxidative damage in kidney and brain of rats.

Arsenic exposure through drinking water causes oxidative stress and tissue damage in the kidney and brain. Curcumin (CUR) is a good antioxidant with limited clinical application because of its hydrophobic nature and limited bioavailability, which can be overcome by the encapsulation of CUR with nanoparticles (NPs). The present study investigates the therapeutic efficacy of free CUR and NP-encapsulated CUR (CUR-NP) against sodium arsenite-induced renal and neuronal oxidative damage in rat. The CUR-NP prepared by emulsion technique and particle size ranged between 120 and 140 nm, with the mean particle size being 130.8 nm. Rats were divided into five groups (groups 1-5) with six animals in each group. Group 1 served as control. Group 2 rats were exposed to sodium arsenite (25 ppm) daily through drinking water for 42 days. Groups 3, 4, and 5 were treated with arsenic as in Group 2; however, these animals were also administered with empty NPs, CUR (100 mg/kg body weight), and CUR-NP (100 mg/kg), respectively, by oral gavage during the last 14 days of arsenic exposure. Arsenic exposure significantly increased serum urea nitrogen and creatinine levels. Arsenic increased lipid peroxidation (LPO), reduced glutathione content and the activities of superoxide dismutase, catalase, glutathione peroxidase, and glutathione reductase were depleted significantly in both kidney and brain. Treatment with free CUR and CUR-NP decreased the LPO and increased the enzymatic and nonenzymatic antioxidant system in kidney and brain. Histopathological examination showed that kidney and brain injury mediated by arsenic was ameliorated by treatment. However, the amelioration percentage indicates that CUR-NP had marked therapeutic effect on arsenic-induced oxidative damage in kidney and brain tissues.

Protective action of curcumin and nano-curcumin against arsenic-induced genotoxicity in rats in vivo.

We explored whether nanoformulation of curcumin can cause better protective effect than free curcumin against arsenic-induced genotoxicity. Curcumin-loaded Poly(lactic-co-glycolic acid) nanoparticles (CUR-NP) were prepared by emulsion technique. The CUR-NP were water soluble and showed biphasic release pattern. Rats were divided into 5 groups of 6 each. Group I served as the control. Group II rats were exposed to sodium arsenite (25 ppm) daily through drinking water for 42 days. Groups III, IV and V were maintained as in Group II, however, they were also administered empty nanoparticle, curcumin (100 mg/kg bw) and CUR-NP (100 mg/kg bw), respectively, by oral gavage during the last 14 days of arsenic exposure. On the 43rd day, genotoxic effects were evaluated in bone marrow cells. Arsenic increased chromosomal aberrations, micronuclei formation and DNA damage. Both free curcumin and CUR-NP attenuated these arsenic-mediated genotoxic effects. However, the result suggests that nanoformulation have better protective effect than free curcumin at the same dose level.

Acetaminophen increases the risk of arsenic-mediated development of hepatic damage in rats by enhancing redox-signaling mechanism.

We evaluated whether the commonly used analgesic-antipyretic drug acetaminophen can modify the arsenic-induced hepatic oxidative stress and also whether withdrawal of acetaminophen administration during the course of long-term arsenic exposure can increase susceptibility of liver to arsenic toxicity. Acetaminophen was co-administered orally to rats for 3 days following 28 days of arsenic pre-exposure (Phase-I) and thereafter, acetaminophen was withdrawn, but arsenic exposure was continued for another 28 days (Phase-II). Arsenic increased lipid peroxidation and reactive oxygen species (ROS) generation, depleted glutathione (GSH), and decreased superoxide dismutase (SOD), catalase, glutathione peroxidase (GPx), and glutathione reductase (GR) activities. Acetaminophen caused exacerbation of arsenic-mediated lipid peroxidation and ROS generation and further enhancement of serum alanine aminotransferase and aspartate aminotransferase activities. In Phase-I, acetaminophen caused further GSH depletion and reduction in SOD, catalase, GPx and GR activities, but in Phase-II, only GPx and GR activities were more affected. Arsenic did not alter basal and inducible nitric oxide synthase (iNOS)-mediated NO production, but decreased constitutive NOS (cNOS)-mediated NO release. Arsenic reduced expression of endothelial NOS (eNOS) and iNOS genes. Acetaminophen up-regulated eNOS and iNOS expression and NO production in Phase-I, but reversed these effects in Phase-II. Results reveal that acetaminophen increased the risk of arsenic-mediated hepatic oxidative damage. Withdrawal of acetaminophen administration also increased susceptibility of liver to hepatotoxicity. Both ROS and NO appeared to mediate lipid peroxidation in Phase-I, whereas only ROS appeared responsible for peroxidative damage in Phase-II.

Light mixed-supervised segmentation for 3D medical image data.

BACKGROUND: Accurate 3D semantic segmentation models are essential for many clinical applications. To train a model for 3D segmentation, voxel-level annotation is necessary, which is expensive to obtain due to laborious work and privacy protection. To accurately annotate 3D medical data, such as MRI, a common practice is to annotate the volumetric data in a slice-by-slice contouring way along principal axes. PURPOSE: In order to reduce the annotation effort in slices, weakly supervised learning with a bounding box (Bbox) was proposed to leverage the discriminating information via a tightness prior assumption. Nevertheless, this method requests accurate and tight Bboxes, which will significantly drop the performance when tightness is not held, that is when a relaxed Bbox is applied. Therefore, there is a need to train a stable model based on relaxed Bbox annotation. METHODS: This paper presents a mixed-supervised training strategy to reduce the annotation effort for 3D segmentation tasks. In the proposed approach, a fully annotated contour is only required for a single slice of the volume. In contrast, the rest of the slices with targets are annotated with relaxed Bboxes. This mixed-supervised method adopts fully supervised learning, relaxed Bbox prior, and contrastive learning during the training, which ensures the network exploits the discriminative information of the training volumes properly. The proposed method was evaluated on two public 3D medical imaging datasets (MRI prostate dataset and Vestibular Schwannoma [VS] dataset). RESULTS: The proposed method obtained a high segmentation Dice score of 85.3% on an MRI prostate dataset and 83.3% on a VS dataset with relaxed Bbox annotation, which are close to a fully supervised model. Moreover, with the same relaxed Bbox annotations, the proposed method outperforms the state-of-the-art methods. More importantly, the model performance is stable when the accuracy of Bbox annotation varies. CONCLUSIONS: The presented study proposes a method based on a mixed-supervised learning method in 3D medical imaging. The benefit will be stable segmentation of the target in 3D images with low accurate annotation requirement, which leads to easier model training on large-scale datasets.

Nanocurcumin ameliorates Staphylococcus aureus-induced mastitis in mouse by suppressing NF­κB signaling and inflammation.

Mastitis is the inflammation of the mammary glands caused by bacteria. It causes severe economic loss to dairy industry. Curcumin, a polyphenol obtained from turmeric, has considerable anti-inflammatory effect. Since it is rapidly eliminated from the body, its oral bioavailability is low. However, nanoformulation of curcumin significantly enhances its therapeutic efficiency by improving its oral bioavailability. We evaluated whether nanocurcumin could be more effective than normal curcumin against bovine Staphylococcus aureus mastitis in mouse model. Curcumin-loaded PLGA nanoparticles (CUR-NP) were prepared by solid-in-oil-in-water emulsion method. The mouse model of mastitis was induced by inoculation of a field strain of S. aureus (bovine mastitis isolate) on the 9th day of parturition through the duct of the mammary gland. CUR-NP and curcumin were given orally for 7 days (day 2 to day 8 of parturition) prior to S. aureus inoculation. We determined the levels of inflammatory cytokines and the mRNA expression of NF­κB. S. aureus infection increased the levels of tumor necrosis factor­α, interleukin­1ß and myeloperoxidase in mammary tissues and C-reactive protein in serum. Both CUR-NP and curcumin significantly attenuated the levels of these cytokines. However, comparatively, the ameliorative efficiency of CUR-NP was better than normal curcumin. S. aureus infection-induced NF­κB mRNA expression was significantly reduced to the healthy control level by CUR-NP. Our study demonstrates that the nanoformulation of curcumin can reduce pro-inflammatory mediators in S. aureus-infected mammary tissues by improving NF­κB signaling. Besides, compared to normal curcumin, this nanoformulation appears to be a better alternative against murine mastitis.

Resolution and noise trade-off analysis for volumetric CT.

Until recently, most studies addressing the trade-off between spatial resolution and quantum noise were performed in the context of single-slice CT. In this study, we extend the theoretical framework of previous works to volumetric CT and further extend it by taking into account the actual shapes of the preferred reconstruction kernels. In the experimental study, we also attempt to explore a three-dimensional approach for spatial resolution measurement, as opposed to the conventional two-dimensional approaches that were widely adopted in previously published studies. By scanning a finite-sized sphere phantom, the MTF was measured from the edge profile along the spherical surface. Cases of different resolutions (and noise levels) were generated by adjusting the reconstruction kernel. To reduce bias, the total photon fluxes were matched: 120 kVp, 200 mA, and 1 s per gantry rotation. All data sets were reconstructed using a modified FDK algorithm under the same condition: Scan field-of-view (SFOV) = 10 cm, and slice thickness = 0.625 mm. The theoretical analysis indicated that the variance of noise is proportional to > 4th power of the spatial resolution. Our experimental results supported this conclusion by showing the relationship is 4.6th (helical) or 5th (axial) power.

Optimization of slice sensitivity profile for radiographic tomosynthesis.

Similar to other tomographic imaging modalities, the slice sensitivity profile (SSP) is an important image quality metric for radiographic tomosynthesis. In this study, the relationship between the acquisition angular range (Theta) and the SSP for the linear trajectory system was carefully investigated from both theoretical and experimental perspectives. A mathematical SSP model was derived for arbitrary points in the reconstructed volume. We used a newly developed flat-panel tomosynthesis prototype system to experimentally validate the mathematical model from 20 degrees (+/-10 degrees) to 60 degrees (+/-30 degrees) angular ranges. The SSP was measured by imaging an edge phantom placed at an angle with respect to the detector plane using the modulation transfer function degradation (MTF-d) method. In addition to the experiments, computer simulations were performed to investigate the relationship in a wider angular range (2.5 degrees to 60 degrees). Furthermore, image data from an anthropomorphic phantom were collected to corroborate the system analysis. All the images in this study were constructed using a 3D view-weighted cone-beam filtered backprojection algorithm (3D VW CB-FBP). The theoretical analysis reveals that the SSP of linear trajectory tomosynthesis is inversely proportional to tan(Theta/2). This theory was supported by both simulation (chi2=1.415, DF=7, p =0.985) and phantom experiment (r=0.999, p < 0.001) and was further confirmed by an analysis of the reconstructed images of an anthropomorphic phantom. The results imply that the benefit of narrower SSP by increasing angular range quickly diminishes once beyond 40 degrees. The advantages of the MTF-d method were also demonstrated.

Immunomodulatory effects of nanocurcumin in arsenic-exposed rats.

We evaluated whether the nanoformulation of curcumin could be more effective than free curcumin against arsenic-induced immune dysfunction in rats. Curcumin was encapsulated in polylactic-co-glycolic acid (PLGA). Nanocurcumin (CUR-NP) exhibited a spherical shape with the mean particle size of 130.8 nm. Rats were randomly divided into five groups of six each. Group I was kept as the control. In Group II, rats were exposed to sodium arsenite (25 ppm) daily through drinking water for 42 days. Groups III, IV and V were treated with arsenic as in Group II, however, they were administered with nanoparticle, curcumin (100 mg/kg bw) and CUR-NP (100 mg/kg bw), respectively, by oral gavage during the last 14 days of arsenic exposure. At term, serum and spleen were collected. Immune dysfunction was evaluated by assessing cellular and humoral immunities. Arsenic significantly decreased the splenic lymphocyte proliferation in response to the antigen -- Keyhole Limpet Hemocyanin (KLH) and mitogen -- concanavalin-A. Arsenic reduced both the delayed type hypersensitivity response and secondary antibody (IgG) response to KLH. It also reduced the lipopolysaccharide-stimulated nitric oxide production in splenic lymphocytes. Free curcumin and CUR-NP treatment significantly attenuated these arsenic-mediated effects. However, the magnitude of the effects indicates that CUR-NP has better ameliorative potential than free curcumin at the equivalent dose level.

Low-dose CT of the abdomen: evaluation of image improvement with use of noise reduction filters pilot study.

A prospective assessment of improvement in image quality at low-radiation-dose computed tomography (CT) of the abdomen by using noise reduction filters was performed. CT images acquired at standard and 50% reduced tube current were processed with six noise reduction filters and evaluated by three radiologists for image noise, sharpness, contrast, and overall image quality in terms of abdominal organ depiction. Quantitative image noise and contrast-to-noise ratio were measured. Baseline low-dose CT images were significantly worse than standard-dose CT images (P <.05). A statistically significant reduction of noise in low-dose images processed with three filters was noted. In conclusion, use of noise reduction filters decreased image noise at low-dose CT.

Detection and characterization of lesions on low-radiation-dose abdominal CT images postprocessed with noise reduction filters.

PURPOSE: To assess the effect of noise reduction filters on detection and characterization of lesions on low-radiation-dose abdominal computed tomographic (CT) images. MATERIALS AND METHODS: Low-dose CT images of abdominal lesions in 19 consecutive patients (11 women, eight men; age range, 32-78 years) were obtained at reduced tube currents (120-144 mAs). These baseline low-dose CT images were postprocessed with six noise reduction filters; the resulting postprocessed images were then randomly assorted with baseline images. Three radiologists performed independent evaluation of randomized images for presence, number, margins, attenuation, conspicuity, calcification, and enhancement of lesions, as well as image noise. Side-by-side comparison of baseline images with postprocessed images was performed by using a five-point scale for assessing lesion conspicuity and margins, image noise, beam hardening, and diagnostic acceptability. Quantitative noise and contrast-to-noise ratio were obtained for all liver lesions. Statistical analysis was performed by using the Wilcoxon signed rank test, Student t test, and kappa test of agreement. RESULTS: Significant reduction of noise was observed in images postprocessed with filter F compared with the noise in baseline nonfiltered images (P =.004). Although the number of lesions seen on baseline images and that seen on postprocessed images were identical, lesions were less conspicuous on postprocessed images than on baseline images. A decrease in quantitative image noise and contrast-to-noise ratio for liver lesions was noted with all noise reduction filters. There was good interobserver agreement (kappa = 0.7). CONCLUSION: Although the use of currently available noise reduction filters improves image noise and ameliorates beam-hardening artifacts at low-dose CT, such filters are limited by a compromise in lesion conspicuity and appearance in comparison with lesion conspicuity and appearance on baseline low-dose CT images.

Can noise reduction filters improve low-radiation-dose chest CT images? Pilot study.

Effect of noise reduction filters on chest computed tomographic (CT) images acquired with 50% radiation dose reduction was evaluated. Two sets of images were acquired with multi-detector row CT at standard (220-280 mA) and 50% reduced (110-140 mA) tube current at the level of the carina. After postprocessing with six noise reduction filters, images were compared with baseline standard-dose images for noise, sharpness, and contrast in lungs, mediastinum, and chest wall. Quantitative image noise was measured in descending thoracic aorta. Modulation transfer functions were calculated from CT images of 50-microm wire. Noise reduction filters reduced image noise on low-radiation-dose chest CT images, with some compromise in image sharpness and contrast assessed qualitatively, and slightly altered modulation transfer function at higher spatial frequencies.