Carbon monoxide (CO) poisoning poses a significant threat to the geriatric population, who experience disproportionately high mortality rates. This study aimed to evaluate the prognostic utility of systemic inflammatory indices and novel composite biomarkers for predicting mortality in elderly patients with CO poisoning. A retrospective cohort study was conducted on 36 geriatric patients (≥ 65 years) admitted to the emergency department with acute CO poisoning. Demographic, clinical, and laboratory data were collected, including complete blood count, biochemistry, arterial blood gas values, and calculated inflammatory indices (NLR, PLR, SII, SIRI, CAR, MAR, RDW/lactate, COHb/lactate). The primary outcome was in-hospital all-cause mortality. While traditional inflammatory indices (SII, SIRI, NLR, PLR) showed no significant association with mortality, the novel COHb/lactate ratio demonstrated exceptional discriminatory power. Non-survivors exhibited significantly higher median creatinine, WBC, hemoglobin, COHb, and lactate levels, but significantly lower RDW/lactate, pH, HCO₃-, and COHb/lactate ratios. ROC analysis revealed an AUC of 0.956 (p = 0.032) for the COHb/lactate ratio in predicting survival. An optimal cut-off value of < 3.61 was identified, yielding a sensitivity of 94.12% and specificity of 100%. The results indicate that conventional inflammatory indices may lack prognostic specificity in geriatric CO poisoning, likely due to the dominant pathophysiology of acute hypoxic-ischemic injury and age-related alterations in inflammatory response. In contrast, the COHb/lactate ratio, which integrates direct markers of exposure and metabolic consequence, emerges as a superior, highly specific biomarker for risk stratification and mortality prediction in this vulnerable population.
Evidence regarding first-trimester umbilical venous hemodynamics and subsequent perinatal morbidity is limited. We evaluated the associations of first-trimester umbilical venous Doppler parameters with neonatal intensive care unit (NICU) admission and preterm birth in singleton pregnancies. This prospective cohort study included 90 pregnancies examined at 11+0 to 13+6 weeks. Time-averaged maximum velocity (TAMXV), umbilical vein diameter, vessel cross-sectional area, and umbilical venous blood flow (UVBF) were measured during routine first-trimester ultrasonography; pregnancy-associated plasma protein-A, free beta-human chorionic gonadotropin (free β-hCG), and nuchal translucency were also recorded. Group comparisons, logistic regression, and receiver operating characteristic analyses were performed. NICU admission occurred in 11 of 87 neonates with available outcome data, and preterm birth occurred in 10 of 90 pregnancies. First-trimester biochemical and umbilical venous Doppler parameters did not differ significantly by NICU admission or preterm birth status. In univariable analysis, higher free β-hCG multiples of the median (MoM) (odds ratio [OR], 2.64; 95% confidence interval [CI], 1.21–5.76; p=.015) and higher TAMXV (OR, 1.16; 95% CI, 1.00–1.33; p=.043) were associated with NICU admission. Neither association remained significant in the primary exploratory multivariable model. In a sensitivity analysis adjusted for gestational age at delivery, TAMXV was attenuated, whereas free β-hCG MoM remained associated with NICU admission (adjusted OR, 2.23; 95% CI, 1.03–4.81; p=.041). TAMXV and UVBF showed limited discrimination for NICU admission and preterm birth. No significant adjusted associations were observed between first-trimester umbilical venous Doppler parameters and either outcome. These findings are exploratory and require confirmation in larger prospective studies.
This retrospective observational study aimed to evaluate the complication rates and predictors of post-procedural complications following liquid polidocanol sclerotherapy in patients with superficial venous disorders. Demographic characteristics, treatment-related variables, and post-procedural adverse events were analyzed in 153 patients who underwent a total of 779 sclerotherapy sessions. Any complication occurred in 55 patients (35.9%), corresponding to a per-session complication rate of 7.1%. The most frequently observed adverse events were erythema, ecchymosis, and localized pain, all of which were self-limiting and clinically manageable. No cases of deep vein thrombosis, anaphylactic reaction, or neurological complications were identified. Multivariable logistic regression analysis demonstrated that the number of treatment sessions was the only independent predictor of complication development. Overall, liquid polidocanol sclerotherapy exhibited a favorable safety profile, with only minor transient complications and no major systemic adverse events, while an increasing number of treatment sessions was associated with a higher risk of post-procedural complications.
The patellofemoral joint is affected by biomechanical alterations to the knee joint following Anterior cruciate ligament (ACL) reconstruction. This study aimed to investigate the relationship between MRI-based Anterior Tibial Translation (ATT) measurements and patellofemoral joint chondral damage following ACL reconstruction. This retrospective study included 40 patients who underwent ACL reconstruction using an autograft between 2020 and 2024 were included in this study. The mean postoperative MRI acquisition time was 11.97 ± 3.23 months. Patellofemoral cartilage lesions were primarily evaluated using the MRI Osteoarthritis Knee Score (MOAKS) scoring system. Patients were classified into rotation groups according to the difference between their ATT-L and ATT-M values. The ATT-L-ATT-M difference was calculated and rotational groups defined: ± 2 mm, neutral; > 2 mm, internal; and < -2 mm, external rotation. The mean age of the patients was 28.93 ± 9.62 years, with 80% of the patients were male. Patellar cartilage damage was detected in 11 (27.5%) patients. The mean ATT-L was 6.075 ± 5.47 mm, whereas the mean ATT-M was 1.84 ± 4.39 mm. No significant association was observed between patellofemoral cartilage damage and either ATT-L or ATT-M (p = 0.575 and p = 0.228, respectively). The cartilage damage rate was 17.9% in the internal rotation group, compared to 50.0% in the neutral and external rotation groups. Cartilage damage was observed significantly less frequently in the internal rotation group than in the neutral and external rotation groups (p = 0.037). Early patellofemoral cartilage damage may be associated with patellofemoral morphology and tibial rotational profile rather than with isolated medial or lateral ATT measurements one year after ACL reconstruction.
Central venous catheter (CVC) placement is a vital procedure in emergency medicine that requires technical expertise to reduce potentially serious complications. As YouTube has become an increasingly popular educational resource, concerns persist regarding the scientific accuracy and reliability of freely accessible procedural training videos. This study aimed to assess the educational quality and reliability of YouTube videos on central venous catheter training. A cross-sectional evaluation of YouTube videos was performed between November 22 and December 22, 2025, using the terms “subclavian,” “jugular,” and “femoral catheter.” After screening the first 60 videos per keyword, 72 eligible videos were included. Video metrics and uploader type were recorded. Educational quality was assessed using DISCERN instrument, Journal of the American Medical Association (JAMA) benchmark criteria, Global Quality Scale (GQS), and Central Venous Catheter-Objective Structured Assessment of Technical Skills (cvcOSATS). Three experts independently scored all videos, and inter-rater reliability was evaluated with Cronbach’s alpha and Intraclass Correlation Coefficient (ICC). Among the analyzed videos, 42 (58.3%) were uploaded by individuals and 30 (41.7%) by institutional sources. Median scores were GQS 3 (Interquartile Range [IQR] 3-4), JAMA 2 (IQR 2-3), DISCERN 44 (IQR 32-60), and cvcOSATS 3 (IQR 3-4), reflecting an overall moderate educational quality. Institutional videos achieved significantly higher scores across all evaluation tools (p < 0.05). Inter-rater agreement was good to excellent, with ICC values ranging from 0.690 to 0.939. Video duration correlated positively with quality scores, whereas popularity indicators such as views and likes did not consistently predict educational value. YouTube videos on CVC training demonstrate only moderate educational quality and should be considered supplementary resources rather than primary training tools. Institutional videos appear more reliable and comprehensive, highlighting the need for critical appraisal and guideline-based education in procedural training. Popularity indicators did not consistently predict educational value.
Cosmetic surgeries prioritize individual expectations and utilize optimal enhancement strategies, and aim to maintain functionality. The plastic surgeon must thoroughly analyze the individual\'s body structure before surgery and make the most appropriate choices. The impact of combined cosmetic surgeries, where several cosmetic body procedures are performed simultaneously under a single anesthesia, on patients\' body image perception and satisfaction compared to single cosmetic surgeries remains poorly understood. This study examined patient satisfaction and body image perception after single and combined aesthetic body contouring procedures. This study is descriptive and cross-sectional. A sample of 96 individuals who underwent single and combined cosmetic body procedures participated in the study. Data were collected via an online platform using an introductory information form, the Body Image Scale, and the Visual Analog Scale for Satisfaction (VAS-S). Data analysis involved the descriptive statistics and nonparametric tests. All individuals who underwent cosmetic body procedures in this study were female, with a mean age of 34.15 ± 6.48, and had between one to four children. 96.9% had a spouse or partner, and 70.8% had a good financial status. Of the individuals who underwent cosmetic body procedures, 56.2% underwent single procedures, and 43.8% underwent combined cosmetic body procedures. In this study, those with a high-income level had a better body image perception (p < 0.001); however, combined surgery resulted in both higher body image and satisfaction regardless of income level (p < 0.001). There was a positive, moderate, and statistically significant relationship between body image perception and satisfaction among individuals who underwent cosmetic body procedures (r = 0.341, p = 0.034). The results of this study showed that individuals who underwent combined cosmetic surgery had better body image perception and satisfaction than those who underwent single cosmetic surgery. We recommend a detailed analysis of individuals\' expectations before undergoing cosmetic surgery.
A retrospective analysis was conducted to assess hypophosphatemia and associated hormonal alterations in patients with iron deficiency anemia treated with ferric carboxymaltose (FCM). Overall, 163 patients were evaluated. Hematological and biochemical parameters, including phosphorus, calcium, magnesium, parathyroid hormone (PTH), and 1,25-dihydroxy vitamin D, were evaluated before and after FCM administration. For the purposes of this study, serum phosphorus levels below 2.5 mg/dL were considered indicative of hypophosphatemia, whereas levels below 1 mg/dL were classified as severe hypophosphatemia. Of the patients, 20.9% were male (n = 34) and 79.1% were female (n = 129). The median age was 56 years (25th–75th percentile: 45–71). Fifty-four patients (33%) received 500 mg and 109 patients (67%) received 1000 mg of FCM. Serum phosphorus levels measured on day 14 after treatment were significantly lower compared to baseline values (3.4 ± 0.6 vs. 2.4 ± 0.9; p < 0.001). Hypophosphatemia developed in 91 patients (55.8%). While serum phosphorus levels decreased significantly, PTH levels increased (72.7 [50.7–105.3] vs. 84.4 [60.7–122.6]; p < 0.001), and 1,25-dihydroxy vitamin D levels showed a significant decline (35.8 ± 15.5 to 26.7 ± 16.2). A significantly higher risk of hypophosphatemia was observed among patients receiving 1000 mg FCM compared with those receiving 500 mg (3.7-fold increase, p = 0.001). Moreover, the rate of hypophosphatemia detected in our study population was comparable to previously published data. Although FCM provides rapid improvement in hematological parameters, it is associated with transient, dose-dependent hypophosphatemia. Hypophosphatemia occurred in 55.8% of patients and was significantly more frequent in patients receiving 1000 mg ferric carboxymaltose.
Breast cancer is the most common cancer in women and the leading cause of cancer-related mortality. Its incidence and early diagnosis rates vary between countries, ethnic groups, and regions. This study examines the impact of educational status on tumor size, axillary involvement, and disease stage in breast cancer patients from Eastern and Southeastern Türkiye. A total of 51 breast cancer patients undergoing surgery at our hospital between August 2023 and January 2025 were evaluated. Four with recurrent cancer were excluded, leaving 47 patients. A significant relationship was found between education level and axillary involvement (p = 0.021). The highest rate of negative axillary involvement was in primary school graduates (89.47%), while the lowest was in high school graduates (37.50%). No significant relationship was found between literacy status and axillary involvement (p = 0.147). Educational level significantly affected axillary involvement in breast cancer, with the lowest rates observed in primary school graduates and the highest in high school graduates. The lack of a significant difference between literate and illiterate groups suggests that visual and auditory media-based awareness campaigns are effective throughout society. Countries should emphasize visual and auditory breast cancer education to maximize outreach.
This study evaluates cerebrospinal fluid (CSF) flow parameters in patients with Primary Open-Angle Glaucoma (POAG) and healthy individuals using Phase Contrast Cine Magnetic Resonance Imaging (PC Cine MRI). The study included 16 patients diagnosed with POAG and 16 healthy volunteers who had normal ophthalmologic examinations. aqueduct area, peak velocity, forward and reverse volumes, net forward volume, aqueductal stroke volume, and flow per minute were among the CSF parameters computed and compared between the POAG and healthy groups. PC Cine MRI parameters of peak velocity, average velocity, forward volume, reverse volume, net forward volume, aqueductal stroke volume, and average flow rate per minute did not differ significantly between the POAG patient and the healthy groups. However, the mean aqueduct area was significantly larger in the POAG patient group compared to the control group, and this difference remained statistically significant even after controlling for the age variable, which can affect CSF flow parameters.
The aim of this study was to evaluate the relationship between macroscopic measurements (diameter and length) of the appendix and factors such as lumen status, gender, and histopathological diagnosis subgroups in patients diagnosed with acute appendicitis. A total of 134 patients who underwent surgery for acute appendicitis were included in the study. Demographic data such as age and gender, and clinical parameters such as lumen status and histopathological diagnosis were recorded. Appendix diameter and length were measured macroscopically and recorded in millimeters (mm). The mean age of the patients was 14.03 ± 2.78 years, with 44% being female and 56% male. The appendix lumen was found to be patent in 36 cases (26.9%) and obliterated in 98 cases (73.1%). Appendix diameter was higher in the obliterated lumen group compared to the patent lumen group (p = 0.005), with a significant increase in Acute Phlegmenous Appendicitis cases compared to other cases. Our findings suggest that macroscopic measurements while not alone, can significantly contribute to the evaluation of acute appendicitis and their interpretation when used in conjunction with clinical and histopathological data.
Dermatologic, cosmetic, and aesthetic procedures have become increasingly widespread, largely driven by rising aesthetic expectations and the expanding influence of social media. Although these interventions are primarily performed to enhance physical appearance rather than to treat disease, they inherently carry risks of complications and unsatisfactory outcomes, which may give rise to medicolegal disputes between patients and physicians. In this retrospective, descriptive study, high court decisions concerning dermatologic, cosmetic, and aesthetic procedures were analyzed to evaluate judicial approaches to physician liability, informed consent, and the distinction between complications and malpractice. Judicial decisions issued within the last ten years were retrieved from the official websites of the High Courts of the Republic of Türkiye and the Lexpera jurisprudence database using predefined keywords related to cosmetic and dermatologic interventions. A total of 54 high court decisions were included, encompassing cases involving both physicians and unauthorized non-physician practitioners. The decisions were examined with respect to violations of advertising and promotion regulations, deficiencies in informed consent, breach of the duty of medical care, dissatisfaction with aesthetic outcomes, and the imposition of legal or administrative sanctions. The analysis revealed that violations of advertising regulations and inadequate informed consent were the most frequently identified legal deficiencies. Procedures performed by unauthorized individuals were associated with substantially higher rates of legal violations, claims for non-pecuniary damages, and administrative penalties compared with those performed by physicians. In physician-related cases, the scope and adequacy of informed consent emerged as the primary determinant of liability, even when adverse outcomes were accepted as medically recognized complications. High courts consistently characterized dermatologic and cosmetic procedures as contracts for work (contracts of result), imposing a stricter liability framework on physicians. These findings highlight the importance of comprehensive, procedure-specific informed consent, adherence to professional standards, and the performance of aesthetic procedures exclusively by authorized physicians to enhance patient safety and reduce legal risk.
Coparenting refers to how caregivers coordinate, support, and manage disagreement/conflict in raising a child, and may be particularly salient during bedtime routines that require predictability and shared coordination. This cross-sectional exploratory pilot study examined associations between coparenting dimensions and infant/toddler sleep outcomes in a clinical sample from Türkiye. Families with infants/toddlers aged 0–36 months were recruited consecutively from a general child and adolescent psychiatry outpatient clinic at Eskişehir City Hospital between 01 October 2025 and 10 January 2026. Mothers completed the Expanded Brief Infant Sleep Questionnaire (BISQ-E), and both parents completed the Coparenting Relationship Scale (CRS). Primary analyses were prespecified age-adjusted partial correlations (partial r) controlling for child age, focusing on parent-mean CRS scores and three bedtime/settling outcomes. Of 40 consenting families, 30 completed all measures and comprised the analytic sample (mean child age 23.6 ± 8.8 months; 60% male). Most children room-shared with parents (76.7%), and 63.3% of parents perceived their child’s sleep as at least some degree of a problem despite generally favorable night sleep quality ratings. In the age-adjusted parent-mean model, later bedtime routine start time was associated with higher coparenting conflict (partial r = 0.590; q = 0.011) and undermining (partial r = 0.469; q = 0.039), and with lower coparenting support (partial r = −0.485; q = 0.038). More frequent sleep-onset difficulties were associated with higher coparenting conflict (partial r = 0.499; q = 0.038). Greater coparenting disagreement was moderately associated with greater bedtime difficulty (partial r = −0.438) but did not survive BH-FDR correction (q = 0.052). In exploratory parent-specific models, paternal exposure to conflict was additionally associated with more frequent night wakings (partial r = 0.436). These pilot findings suggest that coparenting functioning may be more consistently linked to bedtime/settling processes than to sleep maintenance in early childhood, supporting larger longitudinal, dyadic studies incorporating child temperament and parental distress indicators.
Sleep disturbances are common during infancy and may adversely affect parental sleep and overall well-being. This study aimed to evaluate the relationship between sleep characteristics of infants aged 0–36 months and parental sleep quality. This cross-sectional study included 753 parents (96.7% mothers) of infants aged 0–36 months. Data were collected using an online questionnaire assessing infant sleep patterns (sleep duration, nocturnal awakenings, and sleep onset latency) and parental sleep quality using a composite index based on self-reported sleep parameters. Frequent infant nocturnal awakenings were associated with lower parental sleep quality (r = −0.36, p < 0.001). Longer infant nighttime sleep duration was positively correlated with better parental sleep quality (r = 0.42, p < 0.001). Parents of infants who fell asleep within 10 minutes reported significantly higher sleep quality scores compared to those whose infants required longer to fall asleep (p < 0.001). Infant sleep fragmentation and delayed sleep onset are significantly associated with poorer parental sleep quality. These findings highlight the importance of promoting healthy infant sleep routines as part of family-centered interventions to improve parental well-being and overall sleep health.
Absolute lymphocyte count has often been proposed as a practical prognostic variable in critical illness. In older adults, however, immunosenescence and long-term immune remodeling may lower baseline lymphocyte levels and weaken their relationship with outcomes. We therefore examined whether lymphopenia at ICU admission predicts 28-day mortality in geriatric patients and whether it improves commonly used prognostic models. We undertook a retrospective cohort analysis of 202 consecutive ICU admissions of patients aged ≥ 65 years in a tertiary university hospital between January and December 2019. Admission lymphopenia was defined as an absolute lymphocyte count < 1100 × 10³/μL. The primary outcome was 28-day mortality. Independent predictors were assessed using multivariable logistic regression, and discriminative performance was evaluated using ROC curve analysis. Lymphopenia was present in 76.2% of patients and showed no significant association with 28-day mortality (p = 0.209). Mortality was independently related to ward admission, higher APACHE II and Charlson Comorbidity Index scores, lower Glasgow Coma Scale score, and the need for mechanical ventilation and renal replacement therapy. Adding lymphopenia to multivariable models did not improve discrimination (AUROC 0.752 – 0.897; all p > 0.9). In geriatric ICU patients, lymphopenia at admission is frequent but does not independently predict short-term mortality and does not provide incremental value beyond established severity indicators. These findings support prioritizing validated clinical severity scores and organ support variables over isolated immune-cell counts when stratifying mortality risk in older ICU populations.
This study aimed to assess the diagnostic yield of follow‑up diffusion‑weighted magnetic resonance imaging (DWI‑MRI) in emergency department patients with suspected acute ischemic stroke who had an initial negative scan, and to identify associated risk factors. In this retrospective cross‑sectional analysis, we reviewed records of patients presenting between January 2018 and January 2023 who underwent two DWI‑MRI scans for suspected stroke. Inclusion required an initial scan without acute ischemic lesions and a follow‑up scan within 48 hours. Demographics, symptoms, comorbidities, laboratory data, and imaging intervals were analyzed. Among 171 included patients (mean age 66.0 ± 14.7 years; 58.5% female), follow‑up DWI‑MRI revealed new ischemic lesions in 25 cases (14.6%). Patients with positive follow‑up imaging were significantly older (p = 0.047) and more frequently had hyperlipidemia (12.0% vs. 2.1%; p = 0.041). In multivariate analysis, hyperlipidemia remained an independent predictor of positive findings (OR = 7.434, 95% CI: 1.369–40.357, p = 0.020), whereas age did not retain significance. Approximately one in seven patients with suspected ischemic stroke and an initially normal DWI‑MRI show detectable infarction on early follow‑up imaging. Hyperlipidemia is a strong independent risk factor for such delayed‑appearing lesions, suggesting that repeat MRI should be considered in high‑risk patients, particularly those with dyslipidemia, even after an initial negative scan.
Although previous studies have assessed the adequacy of ChatGPT (Generative Pre-Trained Transformer) in psychiatry, no research, to our knowledge, has focused specifically on child psychiatry. This study aimed to evaluate the performance of ChatGPT in child psychiatry by testing it with board-level questions from the question bank of a reference textbook in the field. Seven tests from the textbook were randomized using random.org, and Tests 1,2, and 5 were selected. A total of 600 questions were presented to two versions of ChatGPT: GPT-4o (July 2024) and GPT-5.2 (December 2025). Each response was recorded as correct or incorrect. The questions were further categorized into two groups based on length (short vs.long) and question type (multiple-choice questions–MCQ vs. true/false questions-TFQ). ChatGPT-4o answered 522 (87.0%) of the 600 questions correctly, while GPT-5.2 provided correct answers to 546 (91.0%). The improvement in ChatGPT-4o\'s and ChatGPT-5.2’s accuracy was statistically significant (p < 0.001). No significant differences were found between short and long questions for either version (ChatGPT-4o: p = 0.61; ChatGPT-5.2: p = 0.60). Both versions achieved higher accuracy on MCQs compared to TFQs, this difference was borderline significant for ChatGPT-4o (p < 0.047) and clearly significant for ChatGPT-5.2 (p < 0.009). Both versions of ChatGPT performed successfully, with the latest version, ChatGPT-5.2, showing significantly improved accuracy. As large language models continue to enhance, improvements in accuracy may contribute to reducing certain risks in psychiatry-related tasks.
This study investigated the correlations between biochemical parameters, hepatic steatosis grade, and visceral adipose tissue (VAT) thickness in seborrheic dermatitis (SD) patients without cardiovascular disease, aiming to enhance understanding of SD\'s systemic metabolic implications.Fifty-four SD patients and 54 age- and sex-matched healthy controls were included in this retrospective study (January 2019 - February 2024). Exclusion criteria involved known cardiovascular, metabolic, or chronic inflammatory conditions. Demographic and laboratory data (AST, ALT, GGT, ESR, CRP, complete blood count) were recorded. Ultrasonography assessed hepatic steatosis grade, carotid intima-media thickness (IMT), and subcutaneous, preperitoneal, and VAT thicknesses. SD patients showed a significantly higher prevalence of hepatic steatosis (p = 0.028), with Grade 2-3 steatosis in 18.5% of SD patients versus 3.7% of controls. Aorta-VAT thickness (37.61 ± 17.46 mm vs. 26.63 ± 9.56 mm, p < 0.001) and ALT levels (17.15 ± 8.86 U/L vs. 13.30 ± 3.55 U/L, p = 0.036) were significantly elevated in the SD group. IMT, ESR, and CRP values did not differ significantly. Multiple linear regression identified SD presence, male sex, and increasing age as robust independent predictors for increased Aorta-VAT thickness. This research demonstrates that SD patients, even without cardiovascular comorbidities, exhibit increased hepatic steatosis and visceral adiposity. These findings support SD\'s strong association with systemic metabolic dysregulation and emphasize the critical need for thorough metabolic risk factor assessment in this patient population.
Forensic document examination depends heavily on expert opinion, often leading to subjective reports. This review systematically evaluated image processing and artificial intelligence (AI)-based for offline signature and author verification. The evaluation encompasses model architectures, explainability components, and compliance with forensic reporting standards. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines, a systematic review was conducted across PubMed, Web of Science, IEEE Xplore, and Google Scholar for records published until February 1, 2026. Inclusion criteria covered offline signature verification, document preprocessing, and explainability approaches. Online/dynamic signature studies or those with insufficient methodology were excluded. From 200 initial records, strict eligibility assessments yielded 22 studies. Due to methodological diversity, findings were evaluated using a narrative synthesis. Post-2020 studies predominantly rely on AI and transfer learning. Common limitations include non-writer-independent designs, lack of subject-disjoint training/test sets, data leakage risks, and limited external validation. Performance metrics varied significantly, and reporting based on interpretability and likelihood ratio (LR) frameworks remains notably scarce. The synthesis also showed that similarity scores require calibration before they can be interpreted as evidential strength, and that explainability outputs are valuable only when integrated with expert-based forensic reasoning. While AI-based methods show promise for handwriting analysis, significant gaps remain in external validation and reporting standardization. Establishing quantifiable, standardized approaches is urgently required to support expert opinions and ensure the robust forensic validity of evidence in legal proceedings. Future forensic applications should therefore prioritize subject-disjoint validation, calibrated score interpretation, uncertainty reporting, and transparent integration of AI outputs into expert opinion.
We aimed to describe the prevalence, frequency, and purposes of artificial intelligence (AI) tool use among child and adolescent psychiatrists and residents in Türkiye, and to examine perceived benefits, risks, child/adolescent-specific concerns, and approval of patient/parent mental health–related use outside clinical settings. We conducted an anonymous, cross-sectional online survey (November–December 2025) distributed through national professional WhatsApp groups (convenience sample). The survey assessed sociodemographic/professional variables, technology readiness, AI tool use (ever/regular use, frequency), purposes of use across common tasks, and attitudes/perceptions rated on Likert scales. Composite scores were created for perceived benefits/risks, child/adolescent-specific concerns, and approval of out-of-clinic mental health uses; associations were examined with Spearman correlations. Of 100 initiated responses, 88 complete responses were analyzed (mean age 32.5 years; 76.1% female). ChatGPT was the most commonly used AI tool in our sample: nearly all participants (98.9%) had used it before. 93.2% reported regular professional/educational use of at least one AI tool (ChatGPT: 90.9%). Formal AI training was uncommon (5.7%). Most respondents used AI at least weekly (84.1%), and nearly half used it daily (46.6%). The most frequent purposes were preparing educational materials, accessing up-to-date information, and literature scanning, whereas clinical decision support and medicolegal documentation were less common. Participants reported moderate endorsement of both benefits and risks, high child/adolescent-specific concerns, and comparatively low approval of out-of-clinic mental health uses, with the highest endorsement for therapist-supervised adjunctive use and the lowest for medication or treatment advice. More frequent AI use was associated with higher self-rated AI knowledge, greater perceived benefits and approval, and lower child/adolescent-specific concerns. AI tool use is widespread among child and adolescent psychiatrists in Türkiye despite minimal formal training, suggesting a training-and-governance gap. Use is primarily oriented toward education and knowledge-support tasks, while substantial child/adolescent-specific concerns persist. Patient-facing applications appear most acceptable when supervised; guidance and competency-based training are needed to support safe, bounded implementation.
Gliomas are the most common primary brain tumors, characterized by substantial heterogeneity in clinical course and prognosis. Accurate tumor grading remains essential for therapeutic decision-making and survival prediction. The integration of molecular and clinical variables into predictive models may improve the diagnostic accuracy beyond histology alone. This retrospective study analyzed 856 patients with glioblastoma (GBM) and lower-grade glioma (LGG) from The Cancer Genome Atlas (TCGA) database. Clinical and mutational data were processed using IBM SPSS Neural Networks (v29.0). A Multilayer Perceptron (MLP) artificial neural network was developed to classify glioma grade. Twenty-two genetic predictors (including IDH1, IDH2, TP53, EGFR, PDGFRA, and others) and demographic variables (age, sex, race) were incorporated. Model performance was assessed on an independent test set using accuracy, sensitivity, specificity, precision, F1-score, Matthews correlation coefficient (MCC), and area under the ROC curve (AUC). The model achieved an overall accuracy of 88.2%, with a sensitivity of 91.5% for GBM and specificity of 85.4% for LGG in the test set. The AUC was 0.945, indicating excellent discriminatory power. Variable importance analysis identified IDH1 mutation as the strongest predictor (normalized importance = 100%), followed by age at diagnosis (70.5%) and IDH2 mutation (54.4%). In contrast, RB1 and BCOR mutations had minimal contributions. The proposed ANN model demonstrated robust accuracy and strong discriminatory performance in glioma grading. The integration of molecular markers such as IDH1/2 with clinical variables enhances diagnostic precision and supports the potential role of artificial intelligence in clinical decision support. Future validation in prospective and multi-center cohorts, as well as multimodal data integration, may facilitate personalized treatment strategies for glioma patients.
