Cosmin Medar

Urinary tract infections (UTIs) sit at a leading edge of the antimicrobial-resistance (AMR) crisis, and Romania ranks among the most affected European countries, yet large single-center series of urinary isolates with complete antibiograms remain scarce. We describe the full resistance landscape of urinary pathogens at a Romanian tertiary hospital, with particular attention to multidrug-resistant Klebsiella and to the often-overlooked urease-positive organisms. We analyzed all positive urine cultures processed over 13 consecutive months (February 2025–February 2026). The first isolate per patient was the primary unit of analysis (n = 2331); the full isolate-level dataset (n = 3348) was analyzed secondarily. Resistance phenotypes (putative ESBL, carbapenem resistance, MDR, VRE) were derived from per-agent susceptibility data following Magiorakos et al., and resistance rates were calculated only on tested isolates. Associations were tested with chi-square/Fisher tests and Benjamini–Hochberg correction. A total of 3348 isolates from 2331 unique patients were analyzed (median age 68 years; 55% male). Escherichia coli predominated (40.5%), followed by Enterococcus spp. (22.2%) and Klebsiella spp. (18.9%). Among the first isolates, 15.3% were MDR; 16.7% of E. coli, Klebsiella and Proteus showed a putative ESBL phenotype; and 3.9% of Enterobacterales were carbapenem-resistant. Resistance was concentrated in Klebsiella (36.2% MDR, 22.8% putative ESBL, 14.7% carbapenem-resistant; all higher than E. coli, p < 0.001) and in Pseudomonas spp. (48.1% carbapenem-resistant). E. coli retained excellent activity to nitrofurantoin (1.2%), fosfomycin (1.2%; tested in a subset) and ertapenem (0.1%) despite high trimethoprim–sulfamethoxazole (31.5%) and fluoroquinolone (24.3%) resistance. Urease-positive organisms formed a distinct subgroup with high trimethoprim–sulfamethoxazole resistance but preserved carbapenem activity. The urinary resistance burden was substantial but uneven, concentrated in Klebsiella and Pseudomonas. These findings support a stratified, locally guided empirical strategy and combined stewardship–infection control efforts.

Radiomic feature stability is a necessary condition for clinical translation, yet the impact of inter-observer segmentation variability remains insufficiently characterized for [18F]FDG PET/CT in breast carcinoma. We evaluated the inter-observer reproducibility of 107 original radiomic features extracted from [18F]FDG PET/CT images of 42 patients with biopsy-proven, treatment-naive primary breast carcinoma, using an IBSI-aligned PyRadiomics workflow. Two nuclear medicine physicians independently segmented each tumor using semi-automatic Otsu thresholding to generate an initial tumor mask, followed by manual correction. Reproducibility was quantified using ICC(A,1) with bootstrap-derived 95% confidence intervals. A two-stage reproducibility and redundancy-based feature reduction strategy, combining an ICC threshold with Spearman correlation-based redundancy removal, was applied across nine threshold combinations, and features were classified into three pre-specified stability categories. The segmentation agreement was good, with a mean Dice coefficient of 0.847. Most features showed excellent reproducibility (81/107, 75.7% with ICC ≥ 0.90; median ICC 0.972), whereas shape features based on maximum lesion extension showed poor reproducibility (ICC 0.10–0.25). The reduction strategy resulted in 19 stable non-redundant features, with eight retained across all threshold combinations; 79 features (73.8%) met high-stability criteria. These results and the proposed stability classification framework provide a methodological basis for future predictive PET radiomics studies in breast carcinoma.

Background/Objectives: Autism Spectrum Disorder (ASD) is associated with structural brain alterations, particularly involving white matter and connectivity. Artificial intelligence (AI) enhances the detection of subtle neuroanatomical changes. This study aimed to characterize structural abnormalities and volumetric patterns in children with ASD using AI-assisted MRI. Methods: This retrospective study included 90 children diagnosed with ASD. Brain MRI scans were analyzed using the CE-certified AI platform mdbrain. Structural findings were classified into corpus callosum anomalies, white matter signal abnormalities (WMSA), ventriculomegaly, other abnormalities, or no detectable changes. Group differences were assessed using ANOVA and Kruskal–Wallis tests with Tukey post hoc analysis. Logistic regression, principal component analysis (PCA), and linear discriminant analysis (LDA) were applied. Results: WMSA were identified in 23.3% of patients, followed by other anomalies (27.8%), corpus callosum anomalies (8.9%), ventriculomegaly (8.9%), and no abnormalities (31.1%). Total white matter volume was significantly reduced in pathological groups and was the only independent predictor. PCA identified three principal components reflecting shared temporo-parietal covariance, hemispheric asymmetry, and a white matter-related axis. Exploratory LDA demonstrated partial separation among anomaly categories. Conclusions: Children with ASD in this cohort showed heterogeneous but partially structured MRI alterations involving both focal and global volumetric changes. Reduced total white matter volume was the most consistent multivariable association with structural abnormalities. AI-assisted morphometric analysis may support structural phenotyping in ASD. These findings are exploratory and require confirmation in larger, prospectively validated cohorts before biomarker applications can be considered.

Despite rapid proliferation of AI applications in prostate MRI showing impressive technical performance, clinical adoption remains limited. We conducted a comprehensive narrative review of literature from January 2018 to December 2024, examining AI applications in prostate MRI with emphasis on real-world performance and implementation challenges. Among 200+ studies reviewed, AI systems achieve 87% sensitivity and 72% specificity for cancer detection in research settings. However, external validation reveals average performance drops of 12%, with some implementations showing degradation up to 31%. Only 31% of studies follow reporting guidelines, 11% share code, and 4% provide model weights. Seven real-world implementation studies demonstrate integration times of 3–14 months, with one major center terminating deployment due to unacceptable false positive rates. The translation gap between artificial and clinical intelligence remains substantial. Success requires shifting focus from accuracy metrics to patient outcomes, establishing transparent reporting standards, developing realistic economic models, and creating appropriate regulatory frameworks. The field must combine methodological rigor, clinical relevance, and implementation science to realize AI’s transformative potential in prostate cancer care.

Background/Objectives: [18F]FDG PET/CT radiomics could improve risk stratification in untreated breast carcinoma. Methods: PubMed Central was accessed for full-text English articles (2015–2025) evaluating radiomic features from pretreatment [18F]FDG PET/CT. The Newcastle-Ottawa Scale (NOS) was used to evaluate the risk of bias. Results: Seven studies (1394 patients with a median cohort of about 150 patients) met the inclusion criteria. Radiomics outperformed conventional metabolic measures at predicting pCR to NAC (with the best AUC 0.83 when combining intra- and peritumoral features); differentiating molecular subtypes (AUC 0.856 luminal vs. non-luminal; 0.818 HER2+ vs. HER2−, and 0.888 triple negative vs. others); and assessing androgen receptor (AR) expression. No additional value was found for ER/PR status. Age influenced SUV and texture metrics, especially in triple-negative lesions. Methodological variation was notable: all studies were retrospective, the majority were single-center, only two provided external validation with different protocols of acquisition and segmentation, and at least four distinct software platforms were used for feature extraction and statistical analysis. Conclusions: [18F]FDG PET/CT radiomics shows good potential for predicting neoadjuvant response and molecular profile in breast cancer. However, small, diverse cohorts and non-standardized methodologies limit the evidence. Prospective multicenter studies with standardized acquisition, segmentation and feature extraction are required before clinical use.

The discovery of a resident urinary microbiome has significantly altered the understanding of urolithiasis, expanding its etiology beyond metabolic and dietary factors to include microbial contributions. This review highlights how specific uropathogens—particularly Proteus mirabilis, Klebsiella pneumoniae, and Escherichia coli—facilitate stone formation through mechanisms such as urease activity, citrate degradation, urine alkalinization, biofilm development, and inflammatory signaling. We critically examine how antibiotic therapies, while essential for treating urinary tract infections (UTIs), disrupt microbial homeostasis by depleting beneficial taxa like Lactobacillus and enabling colonization by lithogenic and resistant strains. Recurrent or broad-spectrum antibiotic use is linked to persistent dysbiosis and increased risk of stone recurrence. Additionally, this paper explores emerging microbiome-targeted strategies—such as probiotics, prebiotics, bacteriotherapy, and precision antimicrobials—as potential interventions to restore microbial balance and mitigate stone risk. Recognizing the urinary microbiome as a therapeutic target opens new avenues for personalized, microbiota-conscious strategies in the prevention and management of kidney stone disease.

Adrenal glands are vital endocrine organs whose accurate segmentation on CT imaging presents significant challenges due to their small size and variable morphology. This study evaluates the efficacy of deep learning approaches for automatic adrenal gland segmentation from multiphase CT scans. We implemented three convolutional neural network architectures (U-Net, SegNet, and NablaNet) and assessed their performance on a dataset comprising 868 adrenal glands from contrast-enhanced abdominal CT scans. Performance was evaluated using the Dice similarity coefficient (DSC), alongside practical implementation metrics including training and deployment time. U-Net demonstrated superior segmentation performance (DSC: 0.630 ± 0.05 for right, 0.660 ± 0.06 for left adrenal glands) compared to NablaNet (DSC: 0.552 ± 0.08 for right, 0.550 ± 0.07 for left) and SegNet (DSC: 0.320 ± 0.10 for right, 0.335 ± 0.09 for left). While all models achieved high specificity, boundary delineation accuracy remained challenging. Our findings demonstrate the feasibility of deep learning-based adrenal gland segmentation while highlighting the persistent challenges in achieving the segmentation quality observed with larger abdominal organs. U-Net provides the optimal balance between accuracy and computational requirements, establishing a foundation for further refinement of AI-assisted adrenal imaging tools.

Pregnancy is a transformative stage in a woman’s life, marked by significant physical, emotional, and social changes. This study had three main objectives: (1) to assess the quality of life (QoL) of pregnant women in Romania, (2) to identify the sociodemographic, health, and obstetric factors influencing their QoL and (3) to examine the psychometric properties of the WHOQOL-BREF questionnaire within the Romanian context, determining its effectiveness in evaluating QoL during pregnancy. Methods: A cross-sectional analytical survey was conducted between January and July 2023 among pregnant women in Romania, targeting a geographically diverse sample from urban and rural areas. Eligible participants were Romanian citizens aged 18 or older. Data were collected through an online self-administered questionnaire using Google Forms, with informed consent obtained electronically. The survey included demographic, obstetric, and health-related variables alongside the WHOQOL-BREF tool, which evaluates QoL across four dimensions: Physical, Psychological, Social Relationships, and Environment. Statistical analysis involved confirmatory factor analysis, reliability testing (Cronbach’s α and McDonald’s ω), and comparisons using Welch’s t-tests and ANOVA. Results: A total of 1550 valid responses were analyzed. The WHOQOL-BREF demonstrated excellent internal consistency (Cronbach’s α > 0.9 across all dimensions). Women with higher education and stable employment reported significantly higher QoL scores in physical and psychological dimensions. No significant differences were found based on pregnancy trimester, previous births, or participation in prenatal classes, although trends suggested slight advantages for participants in prenatal education. Women delivering in private hospitals or non-hospital settings reported better psychological and physical QoL than those delivering in public hospitals. Support from partners and urban residency positively influenced perceived QoL. Conclusions: The WHOQOL-BREF is a reliable tool for assessing QoL in pregnant women in Romania. The study highlights the role of education, employment, and delivery location in influencing QoL, emphasizing the need for targeted support for vulnerable groups during pregnancy.

Persistent Müllerian duct syndrome (PMDS) is a rare kind of internal male pseudohermaphroditism. The patient, who has a male karyotype and phenotypic characteristics, exhibits Müllerian duct derivatives such as the uterus, cervix, fallopian tubes, and upper two-thirds of the vagina. This article provides a comprehensive analysis of the CT and MRI characteristics of a case of PMDS in a 35-year-old male patient who sought medical attention at our clinic due to pain in the left inguinal region and the presence of undescended testes on both sides. The imaging results showed a pelvic mass with a bicornuate appearance, situated adjacent to the bladder on the left side. The diagnosis of compensated hypergonadotropic hypogonadism with a normal male karyotype is confirmed through biological and genetic studies. The final diagnosis was confirmed through histopathological examination following laparoscopic transperitoneal surgical removal. The examination revealed a left lateral vesical pelvic tumor with a firm-elastic, bicornuate appearance, along with a thickened endometrium. Microscopic findings included simple glandular hyperplasia with edema in the endometrium, a small adenomatous polyp at the uterine fundus, and bilateral rigid cords consistent with vas deferens histology. The primary issue with PMDS is in its rarity, which consequently limits the availability of comprehensive case series and prospective research. As a result, radiologists and surgeons must possess knowledge of this ailment, as there is a scarcity of defined treatment guidelines and long-term care strategies.

Background: Midwives are strong advocates for vaginal births. However, their visibility and accessibility are poorly perceived by women in Romania. Consequently, the women’s options are limited to a single direction when pregnancy occurs, involving the family doctor, the obstetrician, and often an interventional technical approach at the time of birth. The aim of this research is to identify specific variables that affect the perceptions and attitudes of pregnant women towards the care provided by midwives. This knowledge could contribute to the development of more effective education and information strategies within maternal health services. Methods: A cross-sectional observational analytical survey was conducted in Romania among pregnant women from the general population. Data were collected through a self-administered questionnaire, with informed consent obtained from each participating pregnant woman. The questionnaire was administered online using the cloud-based Google Forms platform and was available on the internet for seven months, from January to July 2023. The questionnaire was distributed through various media channels, both individually and in communication groups, in the form of a link. All questions were mandatory, and the questionnaire could only be submitted after answering all questions. Results: A total of 1301 individual responses were collected. The analysis of the socio-demographic and obstetrical profile of the pregnant women revealed that approximately half, 689 (52.95%), of the participants were aged between 18–29 years, and 1060 (81.47%) of the participants were married. Among our group of 1301 pregnant women, 973 (74.78%) had higher education, and 987 (75.86%) had a regular job. A majority of the survey participants, 936 (71.94%), lived in an urban geographic area, while 476 (36.58%) had attended childbirth education courses, and 791 (60.79%) were in the third trimester of pregnancy. A total of 298 (22.9%) respondents did not want to give birth in a hospital, and one-third, 347 (26.67%), did not place significant importance on control over the childbirth process. Conclusions: The main factors influencing women’s decisions regarding perinatal care and the importance of midwives as a component of the maternal-infant care team are modifiable, and thorough educational and psychological preparation would reduce the increasing predominance of preference for cesarean section, thereby promoting healthier and more woman- and child-centered perinatal care.