Development of Artificial Intelligence Image Classification Models for Determination of Umbilical Cord Vascular Anomalies.

OBJECTIVE: The goal of this work was to develop robust techniques for the processing and identification of SUA using artificial intelligence (AI) image classification models. METHODS: Ultrasound images obtained retrospectively were analyzed for blinding, text removal, AI training, and image prediction. After developing and testing text removal methods, a small n-size study (40 images) using fastai/PyTorch to classify umbilical cord images. This data set was expanded to 286 lateral-CFI images that were used to compare: different neural network performance, diagnostic value, and model predictions. RESULTS: AI-Optical Character Recognition method was superior in its ability to remove text from images. The small n-size mixed single umbilical artery determination data set was tested with a pretrained ResNet34 neural network and obtained and error rate average of 0.083 (n = 3). The expanded data set was then tested with several AI models. The majority of the tested networks were able to obtain an average error rate of <0.15 with minimal modifications. The ResNet34-default performed the best with: an image-classification error rate of 0.0175, sensitivity of 1.00, specificity of 0.97, and ability to correctly infer classification. CONCLUSION: This work provides a robust framework for ultrasound image AI classifications. AI could successfully classify umbilical cord types of ultrasound image study with excellent diagnostic value. Together this study provides a reproducible framework to develop AI-specific ultrasound classification of umbilical cord or other diagnoses to be used in conjunction with physicians for optimal patient care.

Role of interleukin-6 in a glucan-induced model of granulomatous vasculitis.

The role of interleukin-6 (IL-6) in granulomatous vasculitis is not well understood. To investigate its involvement in this type of vasculitis a model of glucan-induced pulmonary vasculitis employed interleukin-6 deficient (IL-6-/-) mice. Briefly, IL-6-/- mice and C57B/J6 wild type (IL-6+/+) mice were injected intravenously with a suspension of glucan isolated from the cell wall of bakers yeast which results in a granulomatous vasculitis primarily in the pulmonary vasculature. Histological examination demonstrated no significant difference in the number of infiltrating leukocytes between the IL-6+/+ and IL-6-/- glucan-injured mice. Similar numbers of granulomas were noted in both the IL-6+/+ and IL-6-/- injured animals, while no granulomas were seen in saline injected control mice. Cells recovered from the bronchoalveolar lavage (BAL) fluid were differentially stained and counted. While there was a significant increase in infiltrating leukocytes recovered from the BAL following glucan-induced injury, there was no significant difference between the IL-6+/+ and IL-6-/- mice. In addition, no difference was demonstrated in total protein content in the BAL fluid between IL-6+/+ and IL-6-/- mice. However, myeloperoxidase (MPO) activity in the lungs of the IL-6-/- mice was less than in their IL-6+/+ counterparts suggesting that these animals have a partial defect in their ability to recruit neutrophils in this model. Studies done to look for levels of other cytokines/chemokines in these animals to compensate for the loss of IL-6 revealed that only IL-10 in the sera (p<0.016) and BAL fluid (p<0.05) of IL-6-/- mice was significantly higher then their IL-6+/+-injured counterparts. These studies suggest that IL-6, while possibly involved in early neutrophil accumulation in this model does not appear critical to the development of the TH-2 mediated granulomatous vasculitis.

Screening of serum samples from Wegener's granulomatosis patients using antibody microarrays.

Wegener's Granulomatosis (WG) is an idiopathic granulomatosis autoimmune vasculitis that primarily affects small vessels and is associated with glomerulonephritis and pulmonary granulomatous vasculitis. Anti-neutrophil cytoplasmic auto-antibodies (cANCA) against proteinase-3 are used to identify WG, but ANCA titers are not present in some patients with the localized disease. The objective of this study was to develop an antibody array to help identify protein expression patterns in serum from patients with WG as compared to normals. The arrays were tested for limits of detection, background, and cross reactivity using standard proteins. The arrays were hybridized with either normal patient serum (n = 30) or with serum samples from a population of WG patients (n = 26) that were age and sex matched. Data analysis and curve fitting of the standard dilution series calculated r(2) values and determined a sensitivity of <50 pg/mL for the majority of proteins. A total of 24 proteins were assessed. Several statistically significant increases (p<0.05) were seen in the expression of: angiotensin converting enzyme-I, IFN-γ, IL-8, s-ICAM-1 and s-VCAM in WG patients as compared to controls. Utilizing the antibody microarray technology has led to the identification of potential biomarkers of vascular injury in the serum of WG patients.

Matrix metalloproteinase-9 is an important factor in hepatic regeneration after partial hepatectomy in mice.

Partial hepatectomy triggers hepatocyte proliferation, hepatic matrix remodeling, and hepatocyte apoptosis, all of which are important processes in the regenerating liver. Previous studies have shown an increase in the levels of matrix metalloproteinases gelatinase A (MMP-2) and gelatinase B (MMP-9) after partial hepatectomy. The goal of this study was to investigate the role of MMP-9 in liver regeneration after partial hepatectomy. A 70% hepatectomy or sham laparotomy was performed in wild-type or MMP-9-deficient (MMP-9-/-) mice. Hepatic regeneration was determined by liver weight/total body weight ratios and BrdU staining, which was used to a calculate mitotic index at several times postoperatively. Cytokine and growth factor expression was evaluated by Luminex bead-based ELISA and Western blots. Finally, the effect of MMP-9 on apoptosis was measured using TUNEL and caspase expression. The MMP-9-/- animals had a delayed hepatic regenerative response when compared with wild-type controls. The MMP-9-deficient animals expressed significantly less VEGF, HGF, and TNF-alpha between days 2 and 3 post-hepatectomy. Apoptosis, as measured by TUNEL staining and caspase expression, was decreased in the MMP-9-/-. In conclusion, MMP-9 plays an important role in liver regeneration after partial hepatectomy by affecting matrix remodeling, as well as cytokine, growth factor, and caspase expression.

Comparison of antibody array substrates and the use of glycerol to normalize spot morphology.

Antibody microarrays are a high-throughput proteomic technology used to examine the expression of multiple proteins in complex solutions. Antibody microarrays can be manufactured on a variety of commercially available activated glass or coated slides. The goal of this study was to compare Hydrogeltrade mark, nitrocellulose, aldehyde-silane and epoxy-silane slides to determine the amount of antibody bound. The optimal substrate was defined as one that bound the greatest amount of antibody with minimal background. Our studies found that epoxy-silane enhanced surface (ES) slides gave the greatest degree of binding along with a minimal background. However, larger antibody microarrays showed variability in spot size, high intra-spot coefficient of variation and drying artifacts. Increasing the amount of glycerol in the spotting buffer caused a dose-dependent improvement in overall spot morphology. Glycerol was tested on 128 different antibodies and showed decreased: mean spot diameter, intra-spot coefficient of variation and drying artifacts. These studies revealed that the optimal slide substrate was epoxy-silane ES microarray slides. Furthermore, glycerol could normalize spot size, decrease intra-spot coefficient of variability, decrease drying artifacts and increase antibody-spotting density.

Harmful and protective roles of neutrophils in sepsis.

The current studies demonstrate protective and harmful effects of neutrophils (PMN) during experimental sepsis after cecal ligation and puncture (CLP). It is known that CLP induces signaling defects in blood PMN. When PMN were depleted 12 h after CLP, there were dramatic reductions in levels of bacteremia, evidence for reduced liver and renal dysfunction, sharp reductions in serum levels of cytokines (IL-1beta, IL-6, IL-10, TNF-alpha, and IL-2), and improved survival. In contrast, PMN depletion before CLP resulted in substantial increases in bacteremia and no evidence for attenuation of liver and renal failure dysfunction. These data suggest that at the onset of sepsis, PMN are important in regulating the levels of bacteremia, whereas after the onset of sepsis, as they lose innate immune functions, their presence is associated with higher levels of bacteremia and intensified organ dysfunction.

Development of an internally controlled antibody microarray.

Antibody microarrays are a high throughput technology used to concurrently screen for protein expression. Most antibody arrays currently used are based on the ELISA sandwich approach that uses two antibodies to screen for the expression of a limited number of proteins. Also because antigen-antibody interactions are concentration-dependent, antibody microarrays need to normalize the amount of antibody that is used. In response to the limitations with the currently existing technology we have developed a single antibody-based microarray where the quantity of antibody spotted is used to standardize the antigen concentration. In addition, this new array utilizes an internally controlled system where one color represents the amount of antibody spotted, and the other color represents the amount of the antigen that is used to quantify the level of protein expression. When compared with median fluorescence intensity alone, normalization for antibody spot intensity decreased variability and lowered the limits of detection. This new antibody array was tested using standard cytokine proteins and also cell lysates obtained from mouse macrophages stimulated in vitro and evaluated for the expression of the cytokine proteins interleukin (IL)-1beta, IL-5, IL-6, and macrophage inflammatory proteins 1alpha and 1beta. The levels of protein expression seen with the antibody microarray was compared with that obtained with Western blot analysis, and the magnitude of protein expression observed was similar with both technologies with the antibody array actually showing a greater degree of sensitivity. In summary, we have developed a new type of antibody microarray to screen for protein expression that utilizes a single antibody and controls for the amount of antibody spotted. This type of array appears at least as sensitive as Western blot analysis, and the technology can be scaled up for high throughput screening for hundreds of proteins in complex biofluids such as blood.

Role of interleukin-6 in immune complex induced models of vascular injury.

Previous studies have suggested that Interleukin-6 (IL-6) acts as a marker of vasculitis. To determine the role of IL-6 in vasculitis we utilized two models of immune complex induced vascular injury (dermal Arthus and acute pulmonary alveolitis) in IL-6 deficient (IL-6(-/-)) and IL-6 sufficient (IL-6(+/+)) mice. Plasma and bronchoalveolar lavage (BAL) levels of IL-6 were elevated in the injured IL-6(+/+) mice with acute alveolitis and in the plasma of IL-6(+/+) mice with dermal Arthus vasculitis. While, IL-6 levels in IL-6(-/-) mice were near or below the levels of detection. Histological examination of the intensity of vascular injury response demonstrated no significant differences between IL-6(-/-) and IL6(+/+) mice. More specifically, lung permeability (total protein in the BAL) in the lung injury model in IL-6(-/-) mice was the same as injured IL-6(+/+) mice. As a corollary, assessment of vascular permeability in both models was the same in the IL-6(-/-) as the IL-6(+/+) mice. Quantification of leukocyte influx into the injured tissues in both models also revealed no differences between the IL-6(-/-) and IL-6(+/+) mice. These data demonstrate that while IL-6 is upregulated in acute vascular injury it does not appear to be critical in the development of the vascular inflammatory response.