Scholarly Works, Chemical Engineering

Permanent URI for this collection

https://hdl.handle.net/10919/24288

Research articles, presentations, and other scholarship

Browse

Now showing 1 - 20 of 156

Chain-length-controllable upcycling of polyolefins to sulfate detergents
Munyaneza, Nuwayo Eric; Ji, Ruiyang; DiMarco, Adrian; Miscall, Joel; Stanley, Lisa; Rorrer, Nicholas; Qiao, Rui; Liu, Guoliang (Springer Nature, 2024-11-18)
Escalating global plastic pollution and the depletion of fossil-based resources underscore the urgent need for innovative end-of-life plastic management strategies in the context of a circular economy. Thermolysis is capable of upcycling end-of-life plastics to intermediate molecules suitable for downstream conversion to eventually high-value chemicals, but tuning the molar mass distribution of the products is challenging. Here we report a temperature-gradient thermolysis strategy for the conversion of polyethylene and polypropylene into hydrocarbons with tunable molar mass distributions. The whole thermolysis process is catalyst- and hydrogen-free. The thermolysis of polyethylene and polyethylene/polypropylene mixtures with tailored temperature gradients generated oil with an average chain length of ~C14. The oil featured a high concentration of synthetically useful α-olefins. Computational fluid dynamics simulations revealed that regulating the reactor wall temperature was the key to tuning the hydrocarbon distributions. Subsequent oxidation of the obtained α-olefins by sulfuric acid and neutralization by potassium hydroxide afforded sulfate detergents with excellent foaming behaviour and emulsifying capacity and low critical micelle concentration. Overall, this work provides a viable approach to producing value-added chemicals from end-of-life plastics, improving the circularity of the anthropogenic carbon cycle.
Non-Nucleoside Lycorine-Based Analogs as Potential DENV/ZIKV NS5 Dual Inhibitors: Structure-Based Virtual Screening and Chemoinformatic Analysis
Rodríguez-Ararat, Adrián Camilo; Hayek-Orduz, Yasser; Vásquez, Andrés-Felipe; Sierra-Hurtado, Felipe; Villegas-Torres, María-Francisca; Caicedo-Burbano, Paola A.; Achenie, Luke E. K.; Barrios, Andrés Fernando González (MDPI, 2024-09-26)
Dengue (DENV) and Zika (ZIKV) virus continue to pose significant challenges globally due to their widespread prevalence and severe health implications. Given the absence of effective vaccines and specific therapeutics, targeting the highly conserved NS5 RNA-dependent RNA polymerase (RdRp) domain has emerged as a promising strategy. However, limited efforts have been made to develop inhibitors for this crucial target. In this study, we employed an integrated in silico approach utilizing combinatorial chemistry, docking, molecular dynamics simulations, MM/GBSA, and ADMET studies to target the allosteric N-pocket of DENV3-RdRp and ZIKV-RdRp. Using this methodology, we designed lycorine analogs with natural S-enantiomers (LYCS) and R-enantiomers (LYCR) as potential inhibitors of non-structural protein 5 (NS5) in DENV3 and ZIKV. Notably, 12 lycorine analogs displayed a robust binding free energy (<−9.00 kcal/mol), surpassing that of RdRp-ribavirin (<−7.00 kcal/mol) along with promising ADMET score predictions (<4.00), of which (LYCR728-210, LYCS728-210, LYCR728-212, LYCS505-214) displayed binding properties to both DENV3 and ZIKV targets. Our research highlights the potential of non-nucleoside lycorine-based analogs with different enantiomers that may present different or even completely opposite metabolic, toxicological, and pharmacological profiles as promising candidates for inhibiting NS5-RdRp in ZIKV and DENV3, paving the way for further exploration for the development of effective antiviral agents.
Epigenomic tomography for probing spatially defined chromatin state in the brain
Liu, Zhengzhi; Deng, Chengyu; Zhou, Zirui; Ya, Xiao; Jiang, Shan; Zhu, Bohan; Naler, Lynette B.; Jia, Xiaoting; Yao, Danfeng (Daphne); Lu, Chang (Cell Press, 2024-03-25)
Spatially resolved epigenomic profiling is critical for understanding biology in the mammalian brain. Singlecell spatial epigenomic assays were developed recently for this purpose, but they remain costly and labor intensive for examining brain tissues across substantial dimensions and surveying a collection of brain samples. Here, we demonstrate an approach, epigenomic tomography, that maps spatial epigenomes of mouse brain at the scale of centimeters. We individually profiled neuronal and glial fractions of mouse neocortex slices with 0.5 mm thickness. Tri-methylation of histone 3 at lysine 27 (H3K27me3) or acetylation of histone 3 at lysine 27 (H3K27ac) features across these slices were grouped into clusters based on their spatial variation patterns to form epigenomic brain maps. As a proof of principle, our approach reveals striking dynamics in the frontal cortex due to kainic-acid-induced seizure, linked with transmembrane ion transporters, exocytosis of synaptic vesicles, and secretion of neurotransmitters. Epigenomic tomography provides a powerful and cost-effective tool for characterizing brain disorders based on the spatial epigenome.
Noncovalently particle-anchored cytokines with prolonged tumor retention safely elicit potent antitumor immunity
Niu, Liqian; Jang, Eungyo; Chin, Ai Lin; Huo, Ziyu; Wang, Wenbo; Cai, Wenjun; Rong, Tong (American Association for the Advancement of Science, 2024-04-19)
Preclinical studies have shown that immunostimulatory cytokines elicit antitumor immune responses but their clinical use is limited by severe immune-related adverse events upon systemic administration. Here, we report a facile and versatile strategy for noncovalently anchoring potent Fc-fused cytokine molecules to the surface of size-discrete particles decorated with Fc-binding peptide for local administration. Following intratumoral injection, particle-anchored Fc cytokines exhibit size-dependent intratumoral retention. The 1-micrometer particle prolongs intratumoral retention of Fc cytokine for over a week and has minimal systemic exposure, thereby eliciting antitumor immunity while eliminating systemic toxicity caused by circulating cytokines. In addition, the combination of these particle-anchored cytokines with immune checkpoint blockade antibodies safely promotes tumor regression in various syngeneic tumor models and genetically engineered murine tumor models and elicits systemic antitumor immunity against tumor rechallenge. Our formulation strategy renders a safe and tumor-agnostic approach that uncouples cytokines’ immunostimulatory properties from their systemic toxicities for potential clinical application.
Robust and Transparent Silver Oxide Coating Fabricated at Room Temperature Kills Clostridioides difficile Spores, MRSA, and Pseudomonas aeruginosa
Hosseini, Mohsen; Huang, Jinge; Williams, Myra D.; Gonzalez, Gerardo Alexander; Jiang, Xiuping; Falkinham, Joseph O.; Ducker, William A. (MDPI, 2023-12-31)
Antimicrobial coatings can inhibit the transmission of infectious diseases when they provide a quick kill that is achieved long after the coating application. Here, we describe the fabrication and testing of a glass coating containing Ag₂O microparticles that was prepared from sodium silicate at room temperature. The half-lives of both methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa on this coating are only 2–4 min. The half-life of Clostridioides difficile spores is about 9–12 min, which is extremely short for a spore. Additional tests on MRSA demonstrate that the coating retains its antimicrobial activity after abrasion and that an increased loading of Ag₂O leads to a shorter half-life. This coating combines the properties of optical transparency, robustness, fast kill, and room temperature preparation that are highly desirable for an antimicrobial coating.
Long-term in situ ruminal degradation of biodegradable polymers in Holstein dairy cattle
Galyon, Hailey; Vibostok, Samuel; Duncan, Jane; Ferreira, Gonzalo; Whittington, Abby; Cockrum, Rebecca R. (American Dairy Science Association, 2022-12-22)
Using biodegradable materials such as polyhydroxyalkanoates (PHA) and poly(butylene succinate-co-adipate) (PBSA) to develop single-use agricultural plastics like bale netting may reduce the negative effects of plastic accumulation in the rumens of cattle. The objective of this research was to assess the long-term degradation of PHA, PBSA, and a PBSA:PHA blend (Blend) compared with a low-density polyethylene (LDPE) control. Polyhydroxyalkanoate, PBSA, Blend, and LDPE films were incubated in the rumens of 3 cannulated, nonlactating Holsteins for up to 150 d. In situ disappearance (ISD) and residue length were assessed after every incubation time. Data were analyzed with PROC MIXED in SAS and adjusted by Tukey's method to determine least squares differences between polymer treatments, incubation time, and their interaction. By 30 d, PHA achieved 100% degradation, with initiation occurring at 14 d indicated by ISD and a reduction in residue length. Poly(butylene succinate-co-adipate) and Blend did not achieve any significant ISD, but fragmentation of PBSA occurred at 60 d and fragmentation of Blend at just 1 d, likely due to abiotic hydrolysis. Low-density polyethylene achieved no ISD, and residue length did not change over incubation time. We propose that a PBSA:PHA blend is a valid alternative to polyethylene single-use agricultural plastic products based on its fragmentation within 1 d of incubation.
Bayesian-optimization-assisted discovery of stereoselective aluminum complexes for ring-opening polymerization of racemic lactide
Wang, Xiaoqian; Huang, Yang; Xie, Xiaoyu; Liu, Yan; Huo, Ziyu; Lin, Maverick; Xin, Hongliang; Tong, Rong (Nature Research, 2023-06-20)
Stereoselective ring-opening polymerization catalysts are used to produce degradable stereoregular poly(lactic acids) with thermal and mechanical properties that are superior to those of atactic polymers. However, the process of discovering highly stereoselective catalysts is still largely empirical.We aim to develop an integrated computational and experimental framework for efficient, predictive catalyst selection and optimization. As a proof of principle, we have developed a Bayesian optimization workflow on a subset of literature results for stereoselective lactide ring-opening polymerization, and using the algorithm, we identify multiple new Al complexes that catalyze either isoselective or heteroselective polymerization. In addition, feature attribution analysis uncovers mechanistically meaningful ligand descriptors, such as percent buried volume (%Vbur) and the highest occupied molecular orbital energy (Eₕₒₘₒ), that can access quantitative and predictivemodels for catalyst development.
Profiling renal dysfunction using Raman chemometric urinalysis, with special reference to COVID19, lupus nephritis, and diabetic nephropathy
Robertson, John L.; Issa, Amr Sayed; Gomez, Mariana; Sullivan, Kathleen; Senger, Ryan S. (Knowledge Enterprise Journals, 2023-09-30)
Background: Many systemic and urinary tract diseases alter renal structure and function, including changing the composition of urine. While routine urinalysis (physical properties, sediment evaluation, urine chemistry analytes) is useful in screening, it has limitations on separating disease processes, structural changes, and functional abnormalities. Likewise, while many individual ‘biomarkers’ have been used to screen for disease, they have not met with widespread clinical adoption. The recent COVID19 Pandemic and the recognition of post-acute sequelae SARS-CoV-2 infection (PASC) have highlighted the need for rapid, scalable, economical, and accurate screening tools for managing disease. Aims: Validate a Raman spectroscopy-based screening technology for urine analysis that could be used for recognition and quantification of systemic and renal effects of acute and PASC COVID19 disease. Methods: One hundred ten (110) urine specimens were obtained from consented adults diagnosed with COVID19 disease by RT-PCR and/or proximate (household) contact With RT-PCR-confirmed COVID19 disease. Samples were analyzed using Raman chemometric urinalysis, a technology that detects hundreds of discrete chemicals in urine and applies computational comparison-machine learning to detect COVID19-associated molecular patterns (‘fingerprints’). Results: When compared with the urine multimolecular ‘fingerprints’ of healthy individuals and patients with known systemic diseases (diabetes mellitus, lupus) that alter renal structure and function, patients with acute and PASC COVID19 had unique ‘fingerprints’ indicative of alterations in renal function (i.e. – infection altered urine composition). Differences in disease severity (mild to severe) were reflected by different ‘fingerprints’ in urine. Roughly 20% of hospitalized patients developed a degree of renal dysfunction (decrements in eGFR) that were correlated with distinct changes in urine fingerprints. Conclusion: Raman chemometric urinalysis may be a useful tool in management of patients with COVID19 disease, particularly in detecting patients with evolving renal dysfunction for whom there should be attention to medication use and renal health restoration/preservation.
Decreasing the Energy of Evaporation Using Interfacial Water: Is This Useful for Solar Evaporation Efficiency?
Ducker, William A. (American Chemical Society, 2023-04)
Evaporation of water using solar power is an economical and environmentally friendly method for purification of aqueous solutions. It has been suggested that intermediate states can be used to lower the enthalpy of evaporation of water and therefore to increase the efficiency of evaporation that uses absorption of sunlight. However, the relevant quantity is the enthalpy of evaporation from bulk water to bulk vapor, which is fixed for a given temperature and pressure. The formation of an intermediate state does not alter the enthalpy of the overall process.
Supramolecular Peptide Nanostructures Regulate Catalytic Efficiency and Selectivity
Li, Zhao; Joshi, Soumil Y.; Wang, Yin; Deshmukh, Sanket A.; Matson, John B. (Wiley-V C H, 2023-05)
We report three constitutionally isomeric tetrapeptides, each comprising one glutamic acid (E) residue, one histidine (H) residue, and two lysine (K-S) residues functionalized with side-chain hydrophobic S-aroylthiooxime (SATO) groups. Depending on the order of amino acids, these amphiphilic peptides self-assembled in aqueous solution into different nanostructures:nanoribbons, a mixture of nanotoroids and nanoribbons, or nanocoils. Each nanostructure catalyzed hydrolysis of a model substrate, with the nanocoils exhibiting the greatest rate enhancement and the highest enzymatic efficiency. Coarse-grained molecular dynamics simulations, analyzed with unsupervised machine learning, revealed clusters of H residues in hydrophobic pockets along the outer edge of the nanocoils, providing insight for the observed catalytic rate enhancement. Finally, all three supramolecular nanostructures catalyzed hydrolysis of the l-substrate only when a pair of enantiomeric Boc-l/d-Phe-ONp substrates were tested. This study highlights how subtle molecular-level changes can influence supramolecular nanostructures, and ultimately affect catalytic efficiency.
Elucidating the Effect of Ion Exchange Protocol on the Copper Exchange Efficacy, Copper Siting, and SCR activity in Cu-SSZ-13
Koishybay, Aibolat; Umhey, Charles; Kuo, Chun-Te; Groden, Kyle; McEwen, Jean-Sabin; Karim, Ayman M.; Shantz, Daniel F. (Wiley-VCH, 2023-05)
The influence of the copper ion exchange protocol on SCR activity of SSZ-13 is quantified. Using the same parent SSZ-13 zeolite, four exchange protocols are used to assess how exchange protocol impacts metal uptake and SCR activity. Large differences in the SCR activity, nearly 30 percentage points at 160 degrees C at constant copper content, are observed for different exchange protocols implying that different exchange protocols lead to different copper species. Hydrogen temperature programmed reduction on selected samples and infrared spectroscopy of CO binding corroborates this conclusion as the reactivity at 160 degrees C correlates with the intensity of the IR band at 2162 cm(-1). DFT-based calculations show that such an IR assignment is consistent with CO adsorbed on a Cu(I) cation within an eight-membered ring. This work shows that SCR activity can be influenced by the ion exchange process even when different protocols lead to the same metal loading. Perhaps most interesting, a protocol used to generate Cu-MOR for methane to methanol studies led to the most active catalyst both on a unit mass or unit mole copper basis. This points to a yet not recognized means to tailor catalyst activity as the open literature is silent on this issue.
Facile Implementation of Antimicrobial Coatings through Adhesive Films (Wraps) Demonstrated with Cuprous Oxide Coatings
Behzadinasab, Saeed; Williams, Myra D.; Falkinham, Joseph O.; Ducker, William A. (MDPI, 2023-05-17)
Antimicrobial coatings have a finite lifetime because of wear, depletion of the active ingredient, or surface contamination that produces a barrier between the pathogen and the active ingredient. The limited lifetime means that facile replacement is important. Here, we describe a generic method for rapidly applying and reapplying antimicrobial coatings to common-touch surfaces. The method is to deposit an antimicrobial coating on a generic adhesive film (wrap), and then to attach that modified wrap to the common-touch surface. In this scenario, the adhesion of the wrap and antimicrobial efficacy are separated and can be optimized independently. We demonstrate the fabrication of two antimicrobial wraps, both using cuprous oxide (Cu₂O) as the active ingredient. The first uses polyurethane (PU) as the polymeric binder and the second uses polydopamine (PDA). Our antimicrobial PU/Cu₂O and PDA/Cu₂O wraps, respectively, kill >99.98% and >99.82% of the human pathogen, P. aeruginosa, in only 10 min, and each of them kill >99.99% of the bacterium in 20 min. These antimicrobial wraps can be removed and replaced on the same object in <1 min with no tools. Wraps are already frequently used by consumers to coat drawers or cars for aesthetic or protective purposes.
International Workshop on Digital Twins for Smart Health
Deng, Jun; Ding, Ying; Achenie, Luke; Liu, Jinwei; Pan, Shimei; Purushotham, Sanjay; Wu, Huanmei; Wang, Qi (ACM, 2023-05)
Customized blends of polypropylene for extrusion based additive manufacturing
Das, Arit; Shanmugham, Nishanth; Bortner, Michael J. (Wiley, 2022-11)
Filament-based material extrusion (MatEx) additive manufacturing has garnered great interest due to its simplicity, customizability, and cost-effectiveness. However, MatEx of semicrystalline polymers is still largely relegated to prototyping applications. Major issues involving volumetric shrinkage and warpage of the printed parts must be addressed in order to employ them for printing functional parts. Moreover, the crystallization behavior and rheology of the polymer are dependent on the MatEx processing conditions. In the current work, the printability of blends of isotactic polypropylene with a soft, low crystallinity propylene based homopolymer is evaluated. Addition of the homopolymer resulted in an increase in the crystallization window of the blends by similar to 6 degrees C that had a profound impact on the interlayer adhesion and residual stress state. The shear-dependent melt flow behavior inside the printing nozzle as well as the interlayer chain diffusion and interlayer welding on the print bed were investigated. Rheological characterizations also indicate sufficient dispersion and miscibility of the homopolymer in the neat polypropylene matrix. The incorporation of the homopolymer as an additive significantly improved the dimensional accuracy of the printed parts through better dissipation of the entrapped residual stresses during MatEx. Moreover, the degree of mechanical anisotropy of the parts was significantly lower than that obtained using many 3D printable grade polymers. The findings from this study can be leveraged in toolpath planning, process parameter optimization, and new feedstock development, highlighting current limitations as well as providing valuable insights into necessary processing modifications in order to enable MatEx of next generation semicrystalline polymers.
Multi-axis alignment of Rod-like cellulose nanocrystals in drying droplets
Pritchard, Cailean Q.; Navarro, Fernando; Roman, Maren; Bortner, Michael J. (Elsevier, 2021-12)
Hypothesis: Radial capillary flow in evaporating droplets carry suspended nanoparticles to its periphery where they are deposited and form a coffee-ring. Rod-like nanoparticles seeking to minimize their capillary energy will align with their long-axis parallel to the contact line. Particles exhibiting electrostatic repulsion, such as cellulose nanocrystals (CNCs), establish a competition between capillary flow-induced impingement against a growing coffee-ring and entropic minimization leading to enhanced particle mobility. Therefore, balancing these effects by manipulating the local particle concentration in drying droplets should result in deposition with a controlled orientation of CNCs. Experiments: The dynamic local order in aqueous suspensions of CNCs in evaporating sessile droplets was investigated through time-resolved polarized light microscopy. The spatial distribution of alignment in deposited CNCs was explored as a function of nanoparticle concentration, droplet volume, initial degree of anisotropy, and substrate hydrophobicity. Computational analysis of the rotational Péclet number during evaporation was also investigated to evaluate any effects of shear-induced alignment. Findings: Multiple modes of orientation were identified suggesting local control over CNC orientation and subsequent properties can be attained via droplet-based patterning methods. Specifically, high local particle concentrations led to tangential alignment and lower local particle concentrations resulted in new evidence for radial alignment near the center of dried droplets.
Encapsulation of PI3K Inhibitor LY294002 within Polymer Nanoparticles Using Ion Pairing Flash Nanoprecipitation
Fergusson, Austin D.; Zhang, Rui; Riffle, Judy S.; Davis, Richey M. (MDPI, 2023-04-06)
Flash nanoprecipitation (FNP) is a turbulent mixing process capable of reproducibly producing polymer nanoparticles loaded with active pharmaceutical ingredients (APIs). The nanoparticles produced with this method consist of a hydrophobic core surrounded by a hydrophilic corona. FNP produces nanoparticles with very high loading levels of nonionic hydrophobic APIs. However, hydrophobic compounds with ionizable groups are not as efficiently incorporated. To overcome this, ion pairing agents (IPs) can be incorporated into the FNP formulation to produce highly hydrophobic drug salts that efficiently precipitate during mixing. We demonstrate the encapsulation of the PI3K inhibitor, LY294002, within poly(ethylene glycol)-b-poly(D,L lactic acid) nanoparticles. We investigated how incorporating two hydrophobic IPs (palmitic acid (PA) and hexadecylphosphonic acid (HDPA)) during the FNP process affected the LY294002 loading and size of the resulting nanoparticles. The effect of organic solvent choice on the synthesis process was also examined. While the presence of either hydrophobic IP effectively increased the encapsulation of LY294002 during FNP, HDPA resulted in well-defined colloidally stable particles, while the PA resulted in ill-defined aggregates. The incorporation of hydrophobic IPs with FNP opens the door for the intravenous administration of APIs that were previously deemed unusable due to their hydrophobic nature.
Clearance of Biodegradable Polymer and Polyethylene Films from the Rumens of Holstein Bull Calves
Galyon, Hailey; Vibostok, Samuel; Duncan, Jane; Ferreira, Gonzalo; Whittington, Abby; Havens, Kirk; McDevitt, Jason; Cockrum, Rebecca R. (MDPI, 2023-03-03)
Due to the occurrence of plastic impaction in ruminants and its deleterious effects on health and production, it is necessary to determine the suitability of biodegradable polymers to replace polyethylene-based agricultural plastics, such as hay netting. The objectives of this study were to evaluate the clearance of a polyhydroxyalkanoate (PHA) and poly(butylene succinate-co-adipate) (PBSA) melt-blend polymer from the rumen when fed to cattle and subsequent animal health. Twelve Holstein bull calves were dosed with an encapsulated 13.6 g of PBSA:PHA (Blend), 13.6 g of low-density polyethylene (LDPE), or four empty gelatin capsules (Control) for 30 d. The feed intake, body weight, and body temperature were evaluated, and hemograms were run on d 0 and d 30. On d 31, calves were euthanized to evaluate gross rumen measurements and pathology, papillae length, and polymer residues in rumen contents. No calves presented any signs related to plastic impaction. The feed intake; body weight; rectal temperature; hematological parameters; gross rumen measurements and pathology; and rumen pH and temperature were not affected by treatments. Calves dosed with LDPE had 27 g of undegraded polymer retained in the rumen while Blend calves had only 2 g of fragmented polymers that were 10% of their original size. Agricultural plastics developed from PBSA:PHA may be a suitable alternative to LDPE-based products in the case of animal ingestion and may reduce the incidence of plastic impaction.
Mechanism and Efficacy of Cu2O-Treated Fabric
Benmamoun, Zachary; Wyhopen, Trent; Li, You; Ducker, William A. (MDPI, 2022-11-16)
Pathogenic bacteria can remain viable on fabrics for several days and therefore are a source of infection. Antimicrobial fabrics are a potential method of reducing such infections, and advances in antimicrobial fabrics can be enhanced by knowledge of how the fabric kills bacteria. Metal oxides have been considered and used as antimicrobial ingredients in self-sanitizing surfaces, including in clinical settings. In this work, we examine how the addition of cuprous oxide (Cu₂O) particles to polypropylene fibers kills bacteria. First, we show that the addition of the Cu₂O particles reduces the viability of common hospital pathogens, Staphylococcus aureus, Pseudomonas aeruginosa, and Streptococcus pneumoniae, by 99.9% after 30 min of contact with the treated polypropylene. Then, we demonstrate that the main killing effect is due to the drying of the bacteria onto the cuprous oxide particles. There is also a weaker effect due to free Cu⁺ ions that dissolve into the liquid. Other dissolved species were unimportant. Chelation of these Cu⁺ ions in soluble form or precipitation removes their antimicrobial activity.
Subpopulation-specific machine learning prognosis for underrepresented patients with double prioritized bias correction
Afrose, Sharmin; Song, Wenjia; Nemeroff, Charles B.; Yao, Danfeng (Nature Research, 2022-09)
Background: Many clinical datasets are intrinsically imbalanced, dominated by overwhelming majority groups. Off-the-shelf machine learning models that optimize the prognosis of majority patient types (e.g., healthy class) may cause substantial errors on the minority prediction class (e.g., disease class) and demographic subgroups (e.g., Black or young patients). In the typical one-machine-learning-model-fits-all paradigm, racial and age disparities are likely to exist, but unreported. In addition, some widely used whole-population metrics give misleading results. Methods: We design a double prioritized (DP) bias correction technique to mitigate representational biases in machine learning-based prognosis. Our method trains customized machine learning models for specific ethnicity or age groups, a substantial departure from the one-model-predicts-all convention. We compare with other sampling and reweighting techniques in mortality and cancer survivability prediction tasks. Results: We first provide empirical evidence showing various prediction deficiencies in a typical machine learning setting without bias correction. For example, missed death cases are 3.14 times higher than missed survival cases for mortality prediction. Then, we show DP consistently boosts the minority class recall for underrepresented groups, by up to 38.0%. DP also reduces relative disparities across race and age groups, e.g., up to 88.0% better than the 8 existing sampling solutions in terms of the relative disparity of minority class recall. Crossrace and cross-age-group evaluation also suggests the need for subpopulation-specific machine learning models. Conclusions: Biases exist in the widely accepted one-machine-learning-model-fits-all-population approach. We invent a bias correction method that produces specialized machine learning prognostication models for underrepresented racial and age groups. This technique may reduce potentially life-threatening prediction mistakes for minority populations.
Focused Ultrasound Biofilm Ablation: Investigation of Histotripsy for the Treatment of Catheter-Associated Urinary Tract Infections (CAUTIs)
Childers, Christopher; Edsall, Connor; Gannon, Jessica; Whittington, Abby R.; Muelenaer, Andre A.; Rao, Jayasimha; Vlaisavljevich, Eli (IEEE, 2021-09-01)
Urinary catheters often become contaminated with biofilms, resulting in catheter-associated urinary tract infections (CAUTIs) that adversely affect patient outcomes. Histotripsy is a non-invasive focused ultrasound therapy previously developed for the non-invasive ablation of cancerous tumors and soft tissues. Histotripsy has also previously shown the ability to treat biofilms on glass slides and surgical meshes. Here, we investigate the potential of histotripsy for the treatment of CAUTIs for the first time in vitro. Clinically relevant catheter materials (Tygon, Silicone, and latex catheter mimics) and commonly used clinical catheters were tested to determine the feasibility of producing luminal histotripsy bubble clouds. A Pseudomonas aeruginosa (strain PA14) biofilm model was developed and tested to produce luminal biofilms in an in vitro Tygon catheter mimic. This model was treated with histotripsy to determine the ability to remove a luminal biofilm. Finally, the bactericidal effects of histotripsy were tested by treating PA14 suspended inside the Tygon catheter mimic. Results showed that histotripsy produced precise luminal cavitation within all tested catheter mimics and clinical catheters. Histotripsy treatment of a PA14 biofilm with histotripsy reduced luminal biofilm OD590 signal down to background levels. Further, the treatment of suspended PA14 in LB showed a 3.45±0.11 log10 reduction in CFU/mL after 6 histotripsy scans across the catheter mimics. Overall, the results of this study demonstrate the potential of histotripsy to provide a new modality for removing bacterial biofilms from catheter-based medical devices and suggest that additional work is warranted to investigate histotripsy for the treatment of CAUTIs and other biomaterial-associated infections.

Browse

Recent Submissions