Clinical artificial intelligence: teaching a large language model to generate recommendations that align with guidelines for the surgical management of GERD.

BACKGROUND: Large Language Models (LLMs) provide clinical guidance with inconsistent accuracy due to limitations with their training dataset. LLMs are "teachable" through customization. We compared the ability of the generic ChatGPT-4 model and a customized version of ChatGPT-4 to provide recommendations for the surgical management of gastroesophageal reflux disease (GERD) to both surgeons and patients. METHODS: Sixty patient cases were developed using eligibility criteria from the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) & United European Gastroenterology (UEG)-European Association of Endoscopic. Surgery (EAES) guidelines for the surgical management of GERD. Standardized prompts were engineered for physicians as the end-user, with separate layperson prompts for patients. A customized GPT was developed to generate recommendations based on guidelines, called the GERD Tool for Surgery (GTS). Both the GTS and generic ChatGPT-4 were queried July 21st, 2024. Model performance was evaluated by comparing responses to SAGES & UEG-EAES guideline recommendations. Outcome data was presented using descriptive statistics including counts and percentages. RESULTS: The GTS provided accurate recommendations for the surgical management of GERD for 60/60 (100.0%) surgeon inquiries and 40/40 (100.0%) patient inquiries based on guideline recommendations. The Generic ChatGPT-4 model generated accurate guidance for 40/60 (66.7%) surgeon inquiries and 19/40 (47.5%) patient inquiries. The GTS produced recommendations based on the 2021 SAGES & UEG-EAES guidelines on the surgical management of GERD, while the generic ChatGPT-4 model generated guidance without citing evidence to support its recommendations. CONCLUSION: ChatGPT-4 can be customized to overcome limitations with its training dataset to provide recommendations for the surgical management of GERD with reliable accuracy and consistency. The training of LLM models can be used to help integrate this efficient technology into the creation of robust and accurate information for both surgeons and patients. Prospective data is needed to assess its effectiveness in a pragmatic clinical environment.

Nanoindentation Response of Structural Self-Healing Epoxy Resin: A Hybrid Experimental-Simulation Approach.

In recent years, self-healing polymers have emerged as a topic of considerable interest owing to their capability to partially restore material properties and thereby extend the product's lifespan. The main purpose of this study is to investigate the nanoindentation response in terms of hardness, reduced modulus, contact depth, and coefficient of friction of a self-healing resin developed for use in aeronautical and aerospace contexts. To achieve this, the bifunctional epoxy precursor underwent tailored functionalization to improve its toughness, facilitating effective compatibilization with a rubber phase dispersed within the host epoxy resin. This approach aimed to highlight the significant impact of the quantity and distribution of rubber domains within the resin on enhancing its mechanical properties. The main results are that pure resin (EP sample) exhibits a higher hardness (about 36.7% more) and reduced modulus (about 7% more), consequently leading to a lower contact depth and coefficient of friction (11.4% less) compared to other formulations that, conversely, are well-suited for preserving damage from mechanical stresses due to their capabilities in absorbing mechanical energy. Furthermore, finite element method (FEM) simulations of the nanoindentation process were conducted. The numerical results were meticulously compared with experimental data, demonstrating good agreement. The simulation study confirms that the EP sample with higher hardness and reduced modulus shows less penetration depth under the same applied load with respect to the other analyzed samples. Values of 877 nm (close to the experimental result of 876.1 nm) and 1010 nm (close to the experimental result of 1008.8 nm) were calculated for EP and the toughened self-healing sample (EP-R-160-T), respectively. The numerical results of the hardness provide a value of 0.42 GPa and 0.32 GPa for EP and EP-R-160-T, respectively, which match the experimental data of 0.41 GPa and 0.30 GPa. This validation of the FEM model underscores its efficacy in predicting the mechanical behavior of nanocomposite materials under nanoindentation. The proposed investigation aims to contribute knowledge and optimization tips about self-healing resins.

Verification of the Self-Healing Ability of PP-co-HUPy Copolymers in Epoxy Systems.

This work concerns the verification of the self-healing ability of PP-co-HUPy copolymers dispersed in epoxy systems. PP is the acronym for the Poly-PEGMA polymer, and HUPy refers to the HEMA-UPy copolymers based on ureidopyrimidinone (UPy) moieties. In particular, this work aims to verify whether this elastomer characterized by an intrinsic self-healing ability can activate supramolecular interactions among polymer chains of an epoxy resin, as in the elastomer alone. The elastomer includes a class of polyethylene glycol monomethyl ether methacrylate-based copolymers, with different percentages of urea-N-2-amino-4-hydroxy-6-methyl pyrimidine-N'-(hexamethylene-n-carboxyethyl methacrylate) (HEMA-UPy) co-monomers. The self-healing capability of these copolymers based on possible quadruple hydrogen bond interactions between polymer chains has been verified. The formulated epoxy samples did not show self-healing efficiency. This can be attributed to the formation of phase segregation that originates during the curing process of the samples, although the PP-co-HUPy copolymers are completely soluble in the liquid epoxy matrix EP. The morphological investigation highlighted the presence of crystals of PP-co-HUPy copolymers, which are in greater quantity in the sample containing the highest weight percentage (7.8 wt%) of HUPy units. Furthermore, the crystals act as promotors for increasing the curing degree (DC) of the epoxy systems containing HUPy units. DC goes from 91.6% for EP to 96.1% and 95.4% for the samples containing weight percentages of 2.5 and 7.8 wt% of HUPy units, respectively. Dynamic mechanical analysis (DMA) shows storage modulus values for epoxy systems containing PP-co-HUPy units lower than that of the unfilled resin EP. The values of maximum in Tan δ (Tg), representing the temperature at which the glass transition occurs, are 220 for the unfilled resin EP, 228 for the sample containing 2.5 wt% of HEMA-UPy units, and 211 for the sample containing 7.8 wt% of HEMA-UPy units.

The performance of artificial intelligence large language model-linked chatbots in surgical decision-making for gastroesophageal reflux disease.

BACKGROUND: Large language model (LLM)-linked chatbots may be an efficient source of clinical recommendations for healthcare providers and patients. This study evaluated the performance of LLM-linked chatbots in providing recommendations for the surgical management of gastroesophageal reflux disease (GERD). METHODS: Nine patient cases were created based on key questions addressed by the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) guidelines for the surgical treatment of GERD. ChatGPT-3.5, ChatGPT-4, Copilot, Google Bard, and Perplexity AI were queried on November 16th, 2023, for recommendations regarding the surgical management of GERD. Accurate chatbot performance was defined as the number of responses aligning with SAGES guideline recommendations. Outcomes were reported with counts and percentages. RESULTS: Surgeons were given accurate recommendations for the surgical management of GERD in an adult patient for 5/7 (71.4%) KQs by ChatGPT-4, 3/7 (42.9%) KQs by Copilot, 6/7 (85.7%) KQs by Google Bard, and 3/7 (42.9%) KQs by Perplexity according to the SAGES guidelines. Patients were given accurate recommendations for 3/5 (60.0%) KQs by ChatGPT-4, 2/5 (40.0%) KQs by Copilot, 4/5 (80.0%) KQs by Google Bard, and 1/5 (20.0%) KQs by Perplexity, respectively. In a pediatric patient, surgeons were given accurate recommendations for 2/3 (66.7%) KQs by ChatGPT-4, 3/3 (100.0%) KQs by Copilot, 3/3 (100.0%) KQs by Google Bard, and 2/3 (66.7%) KQs by Perplexity. Patients were given appropriate guidance for 2/2 (100.0%) KQs by ChatGPT-4, 2/2 (100.0%) KQs by Copilot, 1/2 (50.0%) KQs by Google Bard, and 1/2 (50.0%) KQs by Perplexity. CONCLUSIONS: Gastrointestinal surgeons, gastroenterologists, and patients should recognize both the promise and pitfalls of LLM's when utilized for advice on surgical management of GERD. Additional training of LLM's using evidence-based health information is needed.

Hybrid Hemp Particles as Functional Fillers for the Manufacturing of Hydrophobic and Anti-icing Epoxy Composite Coatings.

The development of hydrophobic composite coatings is of great interest for several applications in the aerospace industry. Functionalized microparticles can be obtained from waste fabrics and employed as fillers to prepare sustainable hydrophobic epoxy-based coatings. Following a waste-to-wealth approach, a novel hydrophobic epoxy-based composite including hemp microparticles (HMPs) functionalized with waterglass solution, 3-aminopropyl triethoxysilane, polypropylene-graft-maleic anhydride, and either hexadecyltrimethoxysilane or 1H,1H,2H,2H-perfluorooctyltriethoxysilane is presented. The resulting epoxy coatings based on hydrophobic HMPs were cast on aeronautical carbon fiber-reinforced panels to improve their anti-icing performance. Wettability and anti-icing behavior of the prepared composites were investigated at 25 °C and -30 °C (complete icing time), respectively. Samples cast with the composite coating can achieve up to 30 °C higher water contact angle and doubled icing time than aeronautical panels treated with unfilled epoxy resin. A low content (2 wt %) of tailored HMPs causes an increase of ∼26% in the glass transition temperature of the coatings compared to pristine resin, confirming the good interaction between the hemp filler and epoxy matrix at the interphase. Finally, atomic force microscopy reveals that the HMPs can induce the formation of a hierarchical structure on the surface of casted panels. This rough morphology, combined with the silane activity, allows the preparation of aeronautical substrates with enhanced hydrophobicity, anti-icing capability, and thermal stability.

Nanometric Mechanical Behavior of Electrospun Membranes Loaded with Magnetic Nanoparticles.

This work analyzes on nanoscale spatial domains the mechanical features of electrospun membranes of Polycaprolactone (PCL) loaded with Functionalized Magnetite Nanoparticles (FMNs) produced via an electrospinning process. Thermal and structural analyses demonstrate that FMNs affect the PCL crystallinity and its melting temperature. HarmoniX-Atomic Force Microscopy (H-AFM), a modality suitable to map the elastic modulus on nanometric domains of the sample surface, evidences that the FMNs affect the local mechanical properties of the membranes. The mechanical modulus increases when the tip reveals the magnetite nanoparticles. That allows accurate mapping of the FMNs distribution along the nanofibers mat through the analysis of a mechanical parameter. Local mechanical modulus values are also affected by the crystallinity degree of PCL influenced by the filler content. The crystallinity increases for a low filler percentage (<5 wt.%), while, higher magnetite amounts tend to hinder the crystallization of the polymer, which manifests a lower crystallinity. H-AFM analysis confirms this trend, showing that the distribution of local mechanical values is a function of the filler amount and crystallinity of the fibers hosting the filler. The bulk mechanical properties of the membranes, evaluated through tensile tests, are strictly related to the nanometric features of the complex nanocomposite system.

Thermal and Electrical Characterization of Polyester Resins Suitable for Electric Motor Insulation.

This paper undertakes the thermal and electrical characterization of three commercial unsaturated polyester imide resins (UPIR) to identify which among them could better perform the insulation function of electric motors (high-power induction motors fed by pulse-wide modulation (PWM) inverters). The process foreseen for the motor insulation using these resins is Vacuum Pressure Impregnation (VPI). The resin formulations were specially selected because they are one-component systems; hence, before the VPI process, they do not require mixing steps with external hardeners to activate the curing process. Furthermore, they are characterized by low viscosity and a thermal class higher than 180 °C and are Volatile Organic Compound (VOC)-free. Thermal investigations using Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC) techniques prove their excellent thermal resistance up to 320 °C. Moreover, impedance spectroscopy in the frequency range of 100 Hz-1 MHz was analyzed to compare the electromagnetic performance of the considered formulations. They manifest an electrical conductivity starting from 10-10 S/m, a relative permittivity around 3, and a loss tangent value lower than 0.02, which appears almost stable in the analyzed frequency range. These values confirm their usefulness as impregnating resins in secondary insulation material applications.

Rheological, Thermal and Mechanical Characterization of Toughened Self-Healing Supramolecular Resins, Based on Hydrogen Bonding.

This paper proposes the design of toughened self-healing supramolecular resins able to fulfill functional and structural requirements for industrial applications. These new nanocomposites are based on compounds acting as promotors of reversible self-healing interactions. Electrically conductive carbon nanotubes, selected among those allowing to reach the electrical percolation threshold (EPT) with a very low amount of nanofiller, were dispersed in the self-healing polymeric matrix to contrast the electrical insulating properties of epoxy matrices, as required for many applications. The formulated supramolecular systems are thermally stable, up to 360 °C. Depending on the chemical formulation, the self-healing efficiency η, assessed by the fracture test, can reach almost the complete self-repairing efficiency (η = 99%). Studies on the complex viscosity of smart nanocomposites highlight that the effect of the nanofiller dominates over those due to the healing agents. The presence of healing compounds anchored to the hosting epoxy matrix determines a relevant increase in the glass transition temperature (Tg), which results in values higher than 200 °C. Compared to the unfilled matrix, a rise from 189 °C to 223 °C is found for two of the proposed formulations.

Tunneling Atomic Force Microscopy Analysis of Supramolecular Self-Responsive Nanocomposites.

A big step forward for composite application in the sector of structural materials is given by the use of Multi-Wall Carbon Nanotubes (MWCNTs) functionalized with hydrogen bonding moieties, such as barbiturate and thymine, to activate self-healing mechanisms and integrate additional functionalities. These materials with multiple healing properties at the same damaged site, imparted by hydrogen bonds, will also have the potential to improve material reliability, extend the service life, reduce replacement costs, and improve product safety. This revolutionary approach is obtained by integrating the non-covalent interactions coupled with the conventional covalent approach used to cross-link the polymer. The objective of this work is to characterize rubber-toughened supramolecular self-healing epoxy formulations based on unfunctionalized and functionalized MWCNTs using Tunneling Atomic Force Microscopy (TUNA). This advanced technique clearly shows the effect produced by the hydrogen bonding moieties acting as reversible healing elements by their simultaneous donor and acceptor character, and covalently linked to MWCNTs to originate self-healing nanocomposites. In particular, TUNA proved to be very effective for the morphology study of both the unfunctionalized and functionalized carbon nanotube-based conductive networks, thus providing useful insights aimed at understanding the influence of the intrinsic nature of the nanocharge on the final properties of the multifunctional composites.

Prospective Evaluation of Health Literacy and Its Impact on Outcomes in Emergency General Surgery.

BACKGROUND: Health literacy (HL) is an important component of national health policy. The aim of our study was to assess the prevalence of low HL (LHL) and determine its impact on outcomes after emergency general surgery (EGS). METHODS: We performed a (2016-2017) prospective cohort analysis of adult EGS patients. HL was assessed using the Short Assessment of HL score. LHL was defined as Short Assessment of HL score <14. Outcomes were the prevalence of LHL, compliance with medications, wound/drain care, 30-d complications, 30-d readmission, and time to resuming activities of daily living. RESULTS: We enrolled 900 patients. The mean age was 43 ± 11 y. Overall, 22% of the patients had LHL. LHL patients were more likely to be Hispanics (59% versus 15%, P < 0.01), uninsured (50% versus 20%, P < 0.01), have lower socioeconomic status (80% versus 40%, P < 0.02), and are less likely to have completed college (5% versus 60%, P < 0.01) compared with HL patients. On regression analysis, LHL was associated with lower medication compliance (OR: 0.81, [0.4-0.9], P = 0.02), inadequate wound/drain care (OR: 0.75, [0.5-0.8], P = 0.01), 30-d complications (OR: 1.95, [1.3-2.5], P < 0.01), and 30-d readmission (OR: 1.51, [1.2-2.6], P = 0.02). The median time of resuming activities of daily living was longer in patients with LHL than HL patients (4 d versus 7 d, P < 0.01). CONCLUSIONS: One in five patients undergoing EGS has LHL. LHL is associated with decreased compliance with discharge instructions, medications, and wound/drain care. Health literacy must be taken into account when discussing the postoperative plan and better instruction is needed for patients with LHL. LEVEL OF EVIDENCE: Level III. STUDY TYPE: Prognostic.

Reversible Self-Healing Carbon-Based Nanocomposites for Structural Applications.

Reversible Hydrogen Bonds (RHB) have been explored to confer self-healing function to multifunctional nanocomposites. This study has been carried out through a sequence of different steps. Hydrogen bonding moieties, with the intrinsic ability to simultaneously perform the functions of both hydrogen donors and acceptors, have been covalently attached to the walls of carbon nanotubes. The epoxy matrix has been modified to adapt the formulation for hosting self-healing mechanisms. It has been toughened with different percentages of rubber phase covalently linked to the epoxy precursor. The most performant matrix, from the mechanical point of view, has been chosen for the incorporation of MWCNTs. Self-healing performance and electrical conductivities have been studied. The comparison of data related to the properties of nanocomposites containing incorporated functionalized and nonfunctionalized MWCNTs has been performed. The values of the electrical conductivity of the self-healing nanocomposites, containing 2.0% by weight of functionalized multiwalled carbon nanotubes (MWCNTs), range between 6.76 × 10-3 S/m and 3.77 × 10-2 S/m, depending on the nature of the functional group. Curing degrees, glass transition temperatures, and storage moduli of the formulated multifunctional nanocomposites prove their potential for application as functional structural materials.

Loss of Blood-Brain Barrier Integrity in a KCl-Induced Model of Episodic Headache Enhances CNS Drug Delivery.

Cortical spreading depression (CSD) in the CNS is suggested as a common mechanism contributing to headache. Despite strong evidence for CNS involvement in headache disorders, drug development for headache disorders remains focused on peripheral targets. Difficulty in delivering drugs across the blood-brain barrier (BBB) may partially account for this disparity. It is known, however, that BBB permeability is increased during several CNS pathologies. In this study, we investigated BBB changes in response to KCl-induced CSD events and subsequent allodynia in rats. Cortical KCl injection in awake, freely moving rats produced facial allodynia with peak intensity between 1.5 and 3 h and CSD induction within 0.5-2 h postinjection. Brain perfusion of 14C-sucrose as a marker of BBB paracellular permeability revealed increased leak in the cortex, but not brainstem, beginning 0.5 h post-KCl injection and resolving within 6 h; no changes in tight junction (TJ) proteins occludin or claudin-5 expression were observed. Acute pretreatment with topiramate to inhibit CSD did not prevent the increased BBB paracellular permeability. CNS delivery of the abortive anti-migraine agent sumatriptan was increased in the cortex 1.5 h post-KCl injection. Surprisingly, sumatriptan uptake was also increased in the brainstem following CSD induction, suggesting regulation of active transport mechanisms at the BBB. Together, these results demonstrate the ability of CSD events to produce transient, time-dependent changes in BBB permeability when allodynia is present and to mediate access of clinically relevant therapeutics (i.e., sumatriptan) to the CNS.