Single­Dose Pharmacokinetics and Safety of the Oral Galectin­3 Inhibitor, Selvigaltin (GB1211), in Participants with Hepatic Impairment.

BACKGROUND AND OBJECTIVES: Selvigaltin (GB1211), an orally available small molecule galectin-3 inhibitor developed as a treatment for liver fibrosis and cirrhosis, was evaluated to assess the effect of hepatic impairment on its pharmacokinetics and safety to address regulatory requirements. METHODS: GULLIVER-2 was a Phase Ib/IIa three-part study. Parts 1 and 3 had single-dose, open-label designs assessing pharmacokinetics (plasma [total and unbound] and urine), safety, and tolerability of 100 mg oral selvigaltin in participants with moderate (Child-Pugh B, Part 1) or severe (Child-Pugh C, Part 3) hepatic impairment, compared with healthy-matched participants (n = 6 each). RESULTS: All participants received selvigaltin and completed the study. No adverse events were reported. The median time to reach maximum total plasma concentration following drug administration was of 3.49 and 4.00 h post-dose for Child-Pugh B and C participants, respectively; comparable with controls. Total plasma exposure was higher for participants with hepatic impairment compared with controls. Whilst maximum plasma concentration (Cmax) was unaffected in Child-Pugh B participants, area under the plasma concentration-time curve from time zero to infinity (AUC∞) increased by ~ 1.7-fold compared with controls, and half-life was prolonged (geometric mean 28.15 vs 16.38 h). In Child-Pugh C participants, Cmax increased by ~ 1.3-fold, AUC∞ increased by ~ 1.5-fold, and half-life was prolonged (21.05 vs 16.14 h). No trend was observed in plasma unbound fractions or urinary excretion of unchanged selvigaltin in either group. CONCLUSION: Hepatic impairment increased selvigaltin exposure without safety concerns. These data can inform dose recommendations for future clinical programmes. TRIAL REGISTRATION: Clinicaltrials.gov NCT05009680.

Expression of Myeloperoxidase in Patient-Derived Endothelial Colony-Forming Cells-Associations with Coronary Artery Disease and Mitochondrial Function.

BACKGROUND AND AIMS: Myeloperoxidase (MPO) plays a critical role in the innate immune response and has been suggested to be a surrogate marker of oxidative stress and inflammation, with elevated levels implicated in cardiovascular diseases, such as atherosclerosis and heart failure, as well as in conditions like rheumatoid arthritis and cancer. While MPO is well-known in leukocytes, its expression and function in human endothelial cells remain unclear. This study investigates MPO expression in patient-derived endothelial colony-forming cells (ECFCs) and its potential association with CAD and mitochondrial function. METHODS: ECFCs were cultured from the peripheral blood of 93 BioHEART-CT patients. MPO expression and associated functions were examined using qRT-PCR, immunochemistry, flow cytometry, and MPO activity assays. CAD presence was defined using CT coronary angiography (CACS > 0). RESULTS: We report MPO presence in patient-derived ECFCs for the first time. MPO protein expression occurred in 70.7% of samples (n = 41) which had nuclear co-localisation, an atypical observation given its conventional localisation in the granules of neutrophils and monocytes. This suggests potential alternative roles for MPO in nuclear processes. MPO mRNA expression was detected in 66.23% of samples (n = 77). CAD patients had a lower proportion of MPO-positive ECFCs compared to non-CAD controls (57.45% vs. 80%, p = 0.04), a difference that persisted in the statin-naïve sub-cohort (53.85% vs. 84.62%, p = 0.02). Non-CAD patients with MPO expression showed upregulated mitochondrial-antioxidant genes (AIFM2, TXNRD1, CAT, PRDX3, PRDX6). In contrast, CAD patients with MPO gene expression had heightened mROS production and mitochondrial mass and decreased mitochondrial function compared to that of CAD patients without MPO gene expression. CONCLUSIONS: MPO is present in the nucleus of ECFCs. In non-CAD ECFCs, MPO expression is linked to upregulated mitochondrial-antioxidant genes, whereas in CAD ECFCs, it is associated with greater mitochondrial dysfunction.

Vascular Cytokines and Atherosclerosis: Differential Serum Levels of TRAIL, IL-18, and OPG in Obstructive Coronary Artery Disease.

The risk-factor-based prediction of atherosclerotic coronary artery disease (CAD) remains suboptimal, particularly in the absence of any of the standard modifiable cardiovascular risk factors (SMuRFs), making the discovery of biomarkers that correlate with atherosclerosis burden critically important. We hypothesized that cytokines and receptors associated with inflammation in CAD-tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), interleukin-18 (IL-18), and osteoprotegerin (OPG)-would be independently associated with CAD. To determine this, we measured the serum biomarker levels of 993 participants from the BioHEART study who had CT coronary angiograms that were scored for severity of stenosis and plaque composition. We found that the quartiles of TRAIL, OPG, and IL-18 were significantly associated with disease scores, and that the IL-18/TRAIL and OPG/TRAIL ratios demonstrated significant differences between no CAD vs. STEMI whereas only the OPG/TRAIL ratio showed differences between no CAD and obstructive CAD (stenosis > 50%). However, these associations did not persist after adjustment for age, sex, SMuRFs, and a family history of CAD. In conclusion, TRAIL, IL-18, and OPG and the derived ratios of IL-18/TRAIL and OPG/TRAIL demonstrate significant associations with raw disease scores and risk factors, but these markers are not discriminatory biomarkers for the prediction of CAD when incorporated into multi-variable risk models.

OceanNet: a principled neural operator-based digital twin for regional oceans.

While data-driven approaches demonstrate great potential in atmospheric modeling and weather forecasting, ocean modeling poses distinct challenges due to complex bathymetry, land, vertical structure, and flow non-linearity. This study introduces OceanNet, a principled neural operator-based digital twin for regional sea-suface height emulation. OceanNet uses a Fourier neural operator and predictor-evaluate-corrector integration scheme to mitigate autoregressive error growth and enhance stability over extended time scales. A spectral regularizer counteracts spectral bias at smaller scales. OceanNet is applied to the northwest Atlantic Ocean western boundary current (the Gulf Stream), focusing on the task of seasonal prediction for Loop Current eddies and the Gulf Stream meander. Trained using historical sea surface height (SSH) data, OceanNet demonstrates competitive forecast skill compared to a state-of-the-art dynamical ocean model forecast, reducing computation by 500,000 times. These accomplishments demonstrate initial steps for physics-inspired deep neural operators as cost-effective alternatives to high-resolution numerical ocean models.

Mesenchymal stem cell cryopreservation with cavitation-mediated trehalose treatment.

Dimethylsulfoxide (DMSO) has conventionally been used for cell cryopreservation both in research and in clinical applications, but has long-term cytotoxic effects. Trehalose, a natural disaccharide, has been proposed as a non-toxic cryoprotectant. However, the lack of specific cell membrane transporter receptors inhibits transmembrane transport and severely limits its cryoprotective capability. This research presents a method to successfully deliver trehalose into mesenchymal stem cells (MSCs) using ultrasound in the presence of microbubbles. The optimised trehalose concentration was shown to be able to not only preserve membrane integrity and cell viability but also the multipotency of MSCs, which are essential for stem cell therapy. Confocal imaging revealed that rhodamine-labelled trehalose was transported into cells rather than simply attached to the membrane. Additionally, the membranes were successfully preserved in lyophilised cells. This study demonstrates that ultrasonication with microbubbles facilitated trehalose delivery, offering promising cryoprotective capability without the cytotoxicity associated with DMSO-based methods.

Quantitative evaluation of anti-biofilm cavitation activity seeded from microbubbles or protein cavitation nuclei by passive acoustic mapping.

OBJECTIVE: Bacterial biofilms represent a major challenge for effective antibiotic therapy as they confer physical and functional changes that protect bacteria from their surrounding environment. In this work, focused ultrasound in combination with cavitation nuclei was used to disrupt biofilms of Staphylococcus aureus and Pseudomonas aeruginosa, both of which are on the World Health Organization's priority list for new antimicrobial research. Approach: Single species biofilms were exposed to ultrasound (0.5 MHz centre frequency, 0.5-1.5 MPa peak rarefactional pressure, 200 cycle pulses, 5 Hz repetition frequency, 30 s duration), in the presence of two different types of cavitation nuclei. Quantitative passive acoustic mapping (PAM) was used to monitor cavitation emissions during treatment using a calibrated linear array. Main Results: It was observed that the cumulative energy of acoustic emissions during treatment was positively correlated with biofilm disruption, with differences between bacterial species attributed to differences in biofilm morphology. PCaN provided increased biofilm reduction compared to microbubbles due in large part to their persistence over the duration of ultrasound exposure. There was also good correlation between the spatial distribution of cavitation as characterized by PAM and the extent of biofilm disruption observed with microscopy. Significance: Collectively, the results from this work indicate the potential broad applicability of cavitation for eliminating biofilms of priority pathogens and the opportunity presented by Passive Acoustic Mapping for real-time monitoring of antimicrobial processes.

Inorganic polyphosphate and the stringent response coordinately control cell division and cell morphology in Escherichia coli.

Bacteria encounter numerous stressors in their constantly changing environments and have evolved many methods to deal with stressors quickly and effectively. One well known and broadly conserved stress response in bacteria is the stringent response, mediated by the alarmone (p)ppGpp. (p)ppGpp is produced in response to amino acid starvation and other nutrient limitations and stresses and regulates both the activity of proteins and expression of genes. Escherichia coli also makes inorganic polyphosphate (polyP), an ancient molecule evolutionary conserved across most bacteria and other cells, in response to a variety of stress conditions, including amino acid starvation. PolyP can act as an energy and phosphate storage pool, metal chelator, regulatory signal, and chaperone, among other functions. Here we report that E. coli lacking both (p)ppGpp and polyP have a complex phenotype indicating previously unknown overlapping roles for (p)ppGpp and polyP in regulating cell division, cell morphology, and metabolism. Disruption of either (p)ppGpp or polyP synthesis led to formation of filamentous cells, but simultaneous disruption of both pathways resulted in cells with heterogenous cell morphologies, including highly branched cells, severely mislocalized Z-rings, and cells containing substantial void spaces. These mutants also failed to grow when nutrients were limited, even when amino acids were added. These results provide new insights into the relationship between polyP synthesis and the stringent response in bacteria and point towards their having a joint role in controlling metabolism, cell division, and cell growth.

Preclinical Characterization of ARX517, a Site-specific Stable PSMA-Targeted Antibody Drug Conjugate for Treatment of Metastatic Castration-Resistant Prostate Cancer.

Metastatic castration-resistant prostate cancer (mCRPC) is an advanced disease in which patients ultimately fail standard of care androgen-deprivation therapies and exhibit poor survival rates. The prostate-specific membrane antigen (PSMA) has been validated as a mCRPC tumor antigen with over-expression in tumors and low expression in healthy tissues. Using our proprietary technology for incorporating synthetic amino acids (SAAs) into proteins at selected sites, we have developed ARX517, an antibody drug conjugate (ADC) which is composed of a humanized anti-PSMA antibody site-specifically conjugated to a tubulin inhibitor at a drug-to-antibody ratio of 2. After binding PSMA, ARX517 is internalized and catabolized, leading to cytotoxic payload delivery and apoptosis. To minimize premature payload release and maximize delivery to tumor cells, ARX517 employs a non-cleavable PEG linker and stable oxime conjugation enabled via SAA protein incorporation to ensure its overall stability. In vitro studies demonstrate that ARX517 selectively induces cytotoxicity of PSMA-expressing tumor cell lines. ARX517 exhibited a long terminal half-life and high serum exposure in mice, and dose-dependent anti-tumor activity in both enzalutamide-sensitive and -resistant CDX and PDX prostate cancer models. Repeat dose toxicokinetic studies in non-human primates demonstrated ARX517 was tolerated at exposures well above therapeutic exposures in mouse pharmacology studies, indicating a wide therapeutic index. In summary, ARX517 inhibited tumor growth in diverse mCRPC models, demonstrated a tolerable safety profile in monkeys, and had a wide therapeutic index based on preclinical exposure data. Based on the encouraging preclinical data, ARX517 is currently being evaluated in a Phase 1 clinical trial ([NCT04662580]).

Targeted opening of the blood-brain barrier using VCAM-1 functionalised microbubbles and "whole brain" ultrasound.

Metastatic tumours in the brain now represent one of the leading causes of death from cancer. Current treatments are largely ineffective owing to the combination of late diagnosis and poor delivery of therapies across the blood-brain barrier (BBB). Conjugating magnetic resonance imaging (MRI) contrast agents with a monoclonal antibody for VCAM-1 (anti-VCAM1) has been shown to enable detection of micrometastases, two to three orders of magnitude smaller in volume than those currently detectable clinically. The aim of this study was to exploit this targeting approach to enable localised and temporary BBB opening at the site of early-stage metastases using functionalised microbubbles and ultrasound. Methods: Microbubbles functionalised with anti-VCAM1 were synthesised and shown to bind to VCAM-1-expressing cells in vitro. Experiments were then conducted in vivo in a unilateral breast cancer brain metastasis mouse model using Gadolinium-DTPA (Gd-DTPA) enhanced MRI to detect BBB opening. Following injection of Gd-DTPA and targeted microbubbles, the whole brain volume was simultaneously exposed to ultrasound (0.5 MHz, 10% duty cycle, 0.7 MPa peak negative pressure, 2 min treatment time). T1-weighted MRI was then performed to identify BBB opening, followed by histological confirmation via immunoglobulin G (IgG) immunohistochemistry. Results: In mice treated with targeted microbubbles and ultrasound, statistically significantly greater extravasation of Gd-DTPA and IgG was observed in the left tumour-bearing hemisphere compared to the right hemisphere 5 min after treatment. No acute adverse effects were observed. There was no investigation of longer term bioeffects owing to the nature of the study. Conclusion: The results demonstrate the feasibility of using targeted microbubbles in combination with low intensity ultrasound to localise opening of the BBB to metastatic sites in the brain. This approach has potential application in the treatment of metastatic tumours whose location cannot be established a priori with conventional imaging methods.

The role of metals in hypothiocyanite resistance in Escherichia coli.

The innate immune system employs a variety of antimicrobial oxidants to control and kill host-associated bacteria. Hypothiocyanite/hypothiocyanous acid (-OSCN/HOSCN) is one such antimicrobial oxidant that is synthesized by lactoperoxidase, myeloperoxidase, and eosinophil peroxidase at sites throughout the human body. HOSCN has potent antibacterial activity while being largely non-toxic toward human cells. The molecular mechanisms by which bacteria sense and defend themselves against HOSCN have only recently begun to be elaborated, notably by the discovery of bacterial HOSCN reductase (RclA), an HOSCN-degrading enzyme widely conserved among bacteria that live on epithelial surfaces. In this paper, I show that Ni2+ sensitizes Escherichia coli to HOSCN by inhibiting glutathione reductase and that inorganic polyphosphate protects E. coli against this effect, probably by chelating Ni2+ ions. I also found that RclA is very sensitive to inhibition by Cu2+ and Zn2+, metals that are accumulated to high levels by innate immune cells, and that, surprisingly, thioredoxin and thioredoxin reductase are not involved in HOSCN stress resistance in E. coli. These results advance our understanding of the contribution of different oxidative stress responses and redox buffering pathways to HOSCN resistance in E. coli and illustrate important interactions between metal ions and the enzymes bacteria use to defend themselves against oxidative stress. IMPORTANCE: Hypothiocyanite (HOSCN) is an antimicrobial oxidant produced by the innate immune system. The molecular mechanisms by which host-associated bacteria defend themselves against HOSCN have only recently begun to be understood. The results in this paper are significant because they show that the low molecular weight thiol glutathione and enzyme glutathione reductase are critical components of the Escherichia coli HOSCN response, working by a mechanism distinct from that of the HOSCN-specific defenses provided by the RclA, RclB, and RclC proteins and that metal ions (including nickel, copper, and zinc) may impact the ability of bacteria to resist HOSCN by inhibiting specific defensive enzymes (e.g., glutathione reductase or RclA).

Lipidomic Risk Score to Enhance Cardiovascular Risk Stratification for Primary Prevention.

BACKGROUND: Accurate risk stratification is vital for primary prevention of cardiovascular disease (CVD). However, traditional tools such as the Framingham Risk Score (FRS) may underperform within the diverse intermediate-risk group, which includes individuals requiring distinct management strategies. OBJECTIVES: This study aimed to develop a lipidomic-enhanced risk score (LRS), specifically targeting risk prediction and reclassification within the intermediate group, benchmarked against the FRS. METHODS: The LRS was developed via a machine learning workflow using ridge regression on the Australian Diabetes, Obesity, and Lifestyle Study (AusDiab; n = 10,339). It was externally validated with the Busselton Health Study (n = 4,492), and its predictive utility for coronary artery calcium scoring (CACS)-based outcomes was independently validated in the BioHEART cohort (n = 994). RESULTS: LRS significantly improved discrimination metrics for the intermediate-risk group in both AusDiab and Busselton Health Study cohorts (all P < 0.001), increasing the area under the curve for CVD events by 0.114 (95% CI: 0.1123-0.1157) and 0.077 (95% CI: 0.0755-0.0785), with a net reclassification improvement of 0.36 (95% CI: 0.21-0.51) and 0.33 (95% CI: 0.15-0.49), respectively. For CACS-based outcomes in BioHEART, LRS achieved a significant area under the curve improvement of 0.02 over the FRS (0.76 vs 0.74; P < 1.0 × 10-5). A simplified, clinically applicable version of LRS was also created that had comparable performance to the original LRS. CONCLUSIONS: LRS, augmenting the FRS, presents potential to improve intermediate-risk stratification and to predict atherosclerotic markers using a simple blood test, suitable for clinical application. This could facilitate the triage of individuals for noninvasive imaging such as CACS, fostering precision medicine in CVD prevention and management.

Proteobacteria impair anti-tumor immunity in the omentum by consuming arginine.

Gut microbiota influence anti-tumor immunity, often by producing immune-modulating metabolites. However, microbes consume a variety of metabolites that may also impact host immune responses. We show that tumors grow unchecked in the omenta of microbe-replete mice due to immunosuppressive Tregs. By contrast, omental tumors in germ-free, neomycin-treated mice or mice colonized with altered Schaedler's flora (ASF) are spontaneously eliminated by CD8+ T cells. These mice lack Proteobacteria capable of arginine catabolism, causing increases in serum arginine that activate the mammalian target of the rapamycin (mTOR) pathway in Tregs to reduce their suppressive capacity. Transfer of the Proteobacteria, Escherichia coli (E. coli), but not a mutant unable to catabolize arginine, to ASF mice reduces arginine levels, restores Treg suppression, and prevents tumor clearance. Supplementary arginine similarly decreases Treg suppressive capacity, increases CD8+ T cell effectiveness, and reduces tumor burden. Thus, microbial consumption of arginine alters anti-tumor immunity, offering potential therapeutic strategies for tumors in visceral adipose tissue.

Health economic analysis of polygenic risk score use in primary prevention of coronary artery disease - A system dynamics model.

Background: Primary prevention programs utilising traditional risk scores fail to identify all individuals who suffer acute cardiovascular events. We aimed to model the impact and cost effectiveness of incorporating a Polygenic risk scores (PRS) into the cardiovascular disease CVD primary prevention program in Australia, using a whole-of-system model. Methods: System dynamics models, encompassing acute and chronic CVD care in the Australian healthcare setting, assessing the cost-effectiveness of incorporating a CAD-PRS in the primary prevention setting. The time horizon was 10-years. Results: Pragmatically incorporating a CAD-PRS in the Australian primary prevention setting in middle-aged individuals already attending a Heart Health Check (HHC) who are determined to be at low or moderate risk based on the 5-year Framingham risk score (FRS), with conservative assumptions regarding uptake of PRS, could have prevented 2, 052 deaths over 10-years, and resulted in 24, 085 QALYs gained at a cost of $19, 945 per QALY with a net benefit of $724 million. If all Australians overs the age of 35 years old had their FRS and PRS performed, and acted upon, 12, 374 deaths and 60, 284 acute coronary events would be prevented, with 183, 682 QALYs gained at a cost of $18, 531 per QALY, with a net benefit of $5, 780 million. Conclusions: Incorporating a CAD-PRS in a contemporary primary prevention setting in Australia would result in substantial health and societal benefits and is cost-effective. The broader the uptake of CAD-PRS in the primary prevention setting in middle-aged Australians, the greater the impact and the more cost-effective the strategy.

C-terminal Poly-histidine Tags Alter Escherichia coli Polyphosphate Kinase Activity and Susceptibility to Inhibition.

In Escherichia coli, many environmental stressors trigger polyphosphate (polyP) synthesis by polyphosphate kinase (PPK1), including heat, nutrient restriction, toxic compounds, and osmotic imbalances. PPK1 is essential for virulence in many pathogens and has been the target of multiple screens for small molecule inhibitors that might serve as new anti-virulence drugs. However, the mechanisms by which PPK1 activity and polyP synthesis are regulated are poorly understood. Our previous attempts to uncover PPK1 regulatory elements resulted in the discovery of PPK1* mutants, which accumulate more polyP in vivo, but do not produce more in vitro. In attempting to further characterize these mutant enzymes, we discovered that the most commonly-used PPK1 purification method - Ni-affinity chromatography using a C-terminal poly-histidine tag - altered intrinsic aspects of the PPK1 enzyme, including specific activity, oligomeric state, and kinetic values. We developed an alternative purification strategy using a C-terminal C-tag which did not have these effects. Using this strategy, we were able to demonstrate major differences in the in vitro response of PPK1 to 5-aminosalicylic acid, a known PPK1 inhibitor, and observed several key differences between the wild-type and PPK1* enzymes, including changes in oligomeric distribution, increased enzymatic activity, and increased resistance to both product (ADP) and substrate (ATP) inhibition, that help to explain their in vivo effects. Importantly, our results indicate that the C-terminal poly-histidine tag is inappropriate for purification of PPK1, and that any in vitro studies or inhibitor screens performed with such tags need to be reconsidered in that light.

Determination of Bovine Lactoferrin in Powdered Infant Formula and Adult Nutritionals by Heparin Affinity Extraction and Reverse-Phase High-Performance Liquid Chromatography/Ultraviolet Detection (HPLC/UV): Single-Laboratory Validation, First Action 2021.10.

BACKGROUND: Infant formulas, and pediatric and adult nutritional products, are being fortified with bovine lactoferrin (bLF) due to its beneficial impacts on immune development and gut health. Lactoferrin supplementation into these products requires an analytical method to accurately quantify the concentrations of bLF to meet global regulatory and quality standards. OBJECTIVE: To develop and validate a lactoferrin method capable of meeting the AOAC INTERNATIONAL Standard Method Performance Requirements (SMPR®) 2020.005. METHODS: Powder formula samples are extracted using warm dibasic phosphate buffer, pH 8, then centrifuged at 4°C to remove insoluble proteins, fat, and other solids. The soluble fraction is further purified on a HiTrap heparin solid-phase extraction (SPE) column to isolate bLF from interferences. Samples are filtered, then analyzed by LC-UV using a protein BEH C4 analytical column and quantitated using an external calibrant. RESULTS: The LOQ (2 mg/100 g), repeatability (RSD: 2.0-4.8%), recovery (92.1-97.7%), and analytical range (4-193 mg/100 g) all meet the method requirements as stated in SMPR 2020.005 for lactoferrin. CONCLUSION: The reported single-laboratory validation (SLV) results demonstrate the ability of this lactoferrin method to meet or exceed the method performance requirements to measure soluble, intact, non-denatured bLF in infant and adult nutritional powder formulas. HIGHLIGHTS: The use of a heparin affinity column to isolate lactoferrin from bovine milk products combined with a selective analytical chromatographic column provides suitable analyte specificity without requiring proprietary equipment or reagents.

Quantifying lung fissure integrity using a three-dimensional patch-based convolutional neural network on CT images for emphysema treatment planning.

Purpose: Evaluation of lung fissure integrity is required to determine whether emphysema patients have complete fissures and are candidates for endobronchial valve (EBV) therapy. We propose a deep learning (DL) approach to segment fissures using a three-dimensional patch-based convolutional neural network (CNN) and quantitatively assess fissure integrity on CT to evaluate it in subjects with severe emphysema. Approach: From an anonymized image database of patients with severe emphysema, 129 CT scans were used. Lung lobe segmentations were performed to identify lobar regions, and the boundaries among these regions were used to construct approximate interlobar regions of interest (ROIs). The interlobar ROIs were annotated by expert image analysts to identify voxels where the fissure was present and create a reference ROI that excluded non-fissure voxels (where the fissure is incomplete). A CNN configured by nnU-Net was trained using 86 CT scans and their corresponding reference ROIs to segment the ROIs of left oblique fissure (LOF), right oblique fissure (ROF), and right horizontal fissure (RHF). For an independent test set of 43 cases, fissure integrity was quantified by mapping the segmented fissure ROI along the interlobar ROI. A fissure integrity score (FIS) was then calculated as the percentage of labeled fissure voxels divided by total voxels in the interlobar ROI. Predicted FIS (p-FIS) was quantified from the CNN output, and statistical analyses were performed comparing p-FIS and reference FIS (r-FIS). Results: The absolute percent error mean (±SD) between r-FIS and p-FIS for the test set was 4.0% (±4.1%), 6.0% (±9.3%), and 12.2% (±12.5%) for the LOF, ROF, and RHF, respectively. Conclusions: A DL approach was developed to segment lung fissures on CT images and accurately quantify FIS. It has potential to assist in the identification of emphysema patients who would benefit from EBV treatment.

MIDRC-MetricTree: a decision tree-based tool for recommending performance metrics in artificial intelligence-assisted medical image analysis.

Purpose: The Medical Imaging and Data Resource Center (MIDRC) was created to facilitate medical imaging machine learning (ML) research for tasks including early detection, diagnosis, prognosis, and assessment of treatment response related to the coronavirus disease 2019 pandemic and beyond. The purpose of this work was to create a publicly available metrology resource to assist researchers in evaluating the performance of their medical image analysis ML algorithms. Approach: An interactive decision tree, called MIDRC-MetricTree, has been developed, organized by the type of task that the ML algorithm was trained to perform. The criteria for this decision tree were that (1) users can select information such as the type of task, the nature of the reference standard, and the type of the algorithm output and (2) based on the user input, recommendations are provided regarding appropriate performance evaluation approaches and metrics, including literature references and, when possible, links to publicly available software/code as well as short tutorial videos. Results: Five types of tasks were identified for the decision tree: (a) classification, (b) detection/localization, (c) segmentation, (d) time-to-event (TTE) analysis, and (e) estimation. As an example, the classification branch of the decision tree includes two-class (binary) and multiclass classification tasks and provides suggestions for methods, metrics, software/code recommendations, and literature references for situations where the algorithm produces either binary or non-binary (e.g., continuous) output and for reference standards with negligible or non-negligible variability and unreliability. Conclusions: The publicly available decision tree is a resource to assist researchers in conducting task-specific performance evaluations, including classification, detection/localization, segmentation, TTE, and estimation tasks.

Assessing variability in non-contrast CT for the evaluation of stroke: The effect of CT image reconstruction conditions on AI-based CAD measurements of ASPECTS value and hypodense volume.

Purpose: To rule out hemorrhage, non-contrast CT (NCCT) scans are used for early evaluation of patients with suspected stroke. Recently, artificial intelligence tools have been developed to assist with determining eligibility for reperfusion therapies by automating measurement of the Alberta Stroke Program Early CT Score (ASPECTS), a 10-point scale with > 7 or ≤ 7 being a threshold for change in functional outcome prediction and higher chance of symptomatic hemorrhage, and hypodense volume. The purpose of this work was to investigate the effects of CT reconstruction kernel and slice thickness on ASPECTS and hypodense volume. Methods: The NCCT series image data of 87 patients imaged with a CT stroke protocol at our institution were reconstructed with 3 kernels (H10s-smooth, H40s-medium, H70h-sharp) and 2 slice thicknesses (1.5mm and 5mm) to create a reference condition (H40s/5mm) and 5 non-reference conditions. Each reconstruction for each patient was analyzed with the Brainomix e-Stroke software (Brainomix, Oxford, England) which yields an ASPECTS value and measure of total hypodense volume (mL). Results: An ASPECTS value was returned for 74 of 87 cases in the reference condition (13 failures). ASPECTS in non-reference conditions changed from that measured in the reference condition for 59 cases, 7 of which changed above or below the clinical threshold of 7 for 3 non-reference conditions. ANOVA tests were performed to compare the differences in protocols, Dunnett's post-hoc tests were performed after ANOVA, and a significance level of p < 0.05 was defined. There was no significant effect of kernel (p = 0.91), a significant effect of slice thickness (p < 0.01) and no significant interaction between these factors (p = 0.91). Post-hoc tests indicated no significant difference between ASPECTS estimated in the reference and any non-reference conditions. There was a significant effect of kernel (p < 0.01) and slice thickness (p < 0.01) on hypodense volume, however there was no significant interaction between these factors (p = 0.79). Post-hoc tests indicated significantly different hypodense volume measurements for H10s/1.5mm (p = 0.03), H40s/1.5mm (p < 0.01), H70h/5mm (p < 0.01). No significant difference was found in hypodense volume measured in the H10s/5mm condition (p = 0.96). Conclusion: Automated ASPECTS and hypodense volume measurements can be significantly impacted by reconstruction kernel and slice thickness.