Search

Abstract Background The task of writing multiple choice question examinations for medical students is complex, timely and requires significant efforts from clinical staff and faculty. Applying artificial intelligence algorithms in this field of medical education may be advisable. Methods During March to April 2023, we utilized GPT-4, an OpenAI application, to write a 210 multi choice questions-MCQs examination based on an existing exam template and thoroughly investigated the output by specialist physicians who were blinded to the source of the questions. Algorithm mistakes and inaccuracies, as identified by specialists were classified as stemming from age, gender or geographical insensitivities. Results After inputting a detailed prompt, GPT-4 produced the test rapidly and effectively. Only 1 question (0.5%) was defined as false; 15% of questions necessitated revisions. Errors in the AI-generated questions included: the use of outdated or inaccurate terminology, age-sensitive inaccuracies, gender-sensitive inaccuracies, and geographically sensitive inaccuracies. Questions that were disqualified due to flawed methodology basis included elimination-based questions and questions that did not include elements of integrating knowledge with clinical reasoning. Conclusion GPT-4 can be used as an adjunctive tool in creating multi-choice question medical examinations yet rigorous inspection by specialist physicians remains pivotal.

D Portal,1 L Hofstetter,2 I Eshed,3 C Dan-Lantsman,3 T Sella,4,5 D Urban,5 A Onn,5,6 J Bar,5 G Segal21Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel; 2Department of Internal Medicine ‘T’, Chaim Sheba Medical Center, affiliated to the Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel; 3Department of Diagnostic Imaging, Chaim Sheba Medical Center, affiliated to the Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel; 4The Dr. Pinchas Borenstein Talpiot Medical Leadership Program, Sheba Medical Center, Ramat Gan, Israel; 5Institute of Oncology, Chaim Sheba Medical Center, affiliated to the Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel; 6Institute of Pulmonary Medicine, Chaim Sheba Medical Center, affiliated to the Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, IsraelBackground: Non-small cell lung cancer (NSCLC) is a common and highly lethal disease. As advanced treatment modalities are being developed, improved prognostication methods are sought. L3 skeletal muscle index (L3SMI) and alanine aminotransferase (ALT) levels are accepted surrogate markers of sarcopenia and related frailty. We aimed to evaluate the potential association of these markers with NSCLC patients’ survival.Methods: A retrospective, single-center study of an NSCLC patients’ cohort. L3SMI was calculated based on skeletal muscle area on computed tomography scans at the level of the L3 vertebra. Clinical data were extracted from clinical charts.Results: A total of 140 patients (56.4% males, median age 66 [range 37–86]) were included in this study, 32% were diagnosed at stage 3 and 45% at stage 4. During the follow-up duration (median of 1.9 years; range 1 month to 6.4 years), 102 patients (72.8%) died. Patients’ characteristics that were found to be associated with increased mortality were performance status, albumin and tumor stage at diagnosis. Sarcopenia, defined as low L3SMI (lower than 41 cm2,/m2, for women and lower than 53 cm2,/m2, for men) was significantly associated with higher risk of mortality compared with patients with normal L3SMI values (77.2%, vs 64.6%, p=0.013) in univariate analysis, but not in a multiple regression analysis.Conclusion: Low L3SMI could serve as a surrogate marker for sarcopenia and frailty and, as such, facilitate the prognostication process of NSCLC patients.Keywords: sarcopenia, skeletal muscle index, lung cancer, tomography, ALT, prognosis

Background Sarcopenia and frailty influence clinical patients' outcomes. Low alanine aminotransferase (ALT) serum activity is a surrogate marker for sarcopenia and frailty. In-hospital hypoglycemia is associated, also with worse clinical outcomes. Aim We evaluated the association between low ALT, risk of in-hospital hypoglycemia and subsequent mortality. Design This was a retrospective cohort analysis. Methods We included patients hospitalized in a tertiary hospital between 2007 and 2019. Patients' data were retrieved from their electronic medical records. Results The cohort included 51 831 patients (average age 70.88). The rate of hypoglycemia was 10.8% (amongst diabetics 19.4% whereas in non-diabetics 8.3%). The rate of hypoglycemia was higher amongst patients with ALT < 10 IU/l in the whole cohort (14.3% vs. 10.4%, P  < 0.001) as well as amongst diabetics (24.6% vs. 18.8%, P  < 0.001). Both the overall and in-hospital mortality were higher in the low ALT group (57.7% vs. 39.1% P  < 0.001 and 4.3% vs. 3.2%, P  < 0.001). A propensity score matching, after which a regression model was performed, showed that patients with ALT levels < 10 IU/l had higher risk of overall mortality (HR = 1.21, CI 1.13–1.29, P  < 0.001). Conclusions Low ALT values amongst hospitalized patients are associated with increased risk of in-hospital hypoglycemia and overall mortality. [ABSTRACT FROM AUTHOR]

Ab initio configuration interaction calculations with full geometric optimization along the Cl–CN dissociation coordinate have been carried out for several electronic states of ClCN. The calculations treat all low-lying singlet and triplet states and yield the vertical transition probabilities and oscillator strengths, as well as angular dependences near the Franck–Condon (FC) region and least energy paths for several electronic states as a function of rCCl. We find that the low-lying excited states derive from three electronic configurations: π3σ*, π3π*, and σπ4σ*. The lowest excited triplet and singlet states derive from the π3σ* configuration and give rise to bent 1,3A’ and 1,3A‘ states. States arising from the π3π* configuration are linear (Σ and Δ states). There is evidence of surface crossings along the reaction coordinate between triplet states arising from the π3π* configuration and those arising from the π3σ* and σπ4σ* configurations. These crossings can be induced by slight bending of ClCN causing lowering of the symmetry. The calculated vertical excitation energies are in good agreement with features of the absorption spectrum, and suggest that the A continuum of ClCN involves transitions to the 2 1A’ and 1 1A‘ states. The implications to the photodissociation dynamics of the cyanogen halides are discussed. [ABSTRACT FROM AUTHOR]

Large scale configuration interaction calculations have been carried out for the potential energy surfaces of the four lowest singlet states of nitric acid along the dissociation paths leading to OH (X 2Π) and NO2 in its ground X 2A1 and B 2B1 states. In order to gain a detailed picture of the dissociation dynamics, correlations are drawn between the forces experienced by these species upon photodissociation and observed oscillator strengths, vector coupling parameters, relative Λ doublet intensities and available energy distribution where this data is available. The calculations are in good agreement with experiment and support the conclusion that the transition to the lowest such excited state, 1 1A‘, is a vibronic transition whose electric dipole transition polarization is in plane, i.e., that the overall vibronic state observed is of A’ symmetry. Dissociation from the ground state is shown to be associated with a tumbling motion of the OH due to the predominance of quadrupolar forces at intermediate distances. [ABSTRACT FROM AUTHOR]