Diagnostics最新文献

Background: This study evaluated the prognostic impact of Trop-2, CD47, and CD163 expression on clinical outcomes in triple-negative breast cancer (TNBC) and investigated their interactions with tumor progression. Methods: A retrospective cohort of 92 patients with TNBC was analyzed. The expression scores for Trop-2, CD47, and CD163 were categorized as negative/low (0-3 points) or high (4-6 points). The primary endpoint was overall survival (OS). Results: The median age of the cohort was 50 years old. High Trop-2 expression was observed in 55.4% of the patients and was significantly associated with advanced disease stage (p < 0.001). High CD47 expression (44.6%) was correlated with advanced stage (p = 0.044), whereas high CD163 expression (45.7%) was associated with advanced stage (p = 0.021), absence of comorbidities (p = 0.022), and lower pT stage (p = 0.023). Moderate positive correlations were found between Trop-2 and CD47 (p = 0.022), Trop-2 and CD163 (p = 0.037), and CD47 and CD163 (p < 0.001), respectively. Kaplan-Meier survival analysis revealed that patients with low Trop-2 expression exhibited significantly prolonged OS (p = 0.021) and progression-free survival (PFS) (p = 0.026) compared to those with high Trop-2 expression. Univariate and multivariate analyses revealed significant associations between OS and PFS for Trop-2, lymphovascular invasion, and BRCA status. Conclusions: Trop-2 expression is a significant prognostic factor for TNBC and is correlated with worse outcomes. Although CD47 and CD163 showed trends for poorer prognosis, their significance was not confirmed. These findings offer promising prospects for future studies on combined antibody-drug conjugates (ADCs), as they may present opportunities to address multiple resistance mechanisms in the management of TNBC and enhance clinical outcomes.

Objective: Ultrasound is a valuable diagnostic tool in the diagnostic work-up of dyspnea and can identify even small pleural effusions. The incorporation of shear wave elastography (SWE) represents a possible tool in stratifying pleural effusions by the risk of underlying malignancy. No previous studies on ultrasound with the incorporation of SWE have been conducted in an emergency department (ED), where such stratification might have a clinical impact by hastening referrals for the diagnostic work-up of underlying malignancy. The objective of this study was to appraise the diagnostic accuracy of ultrasonographic findings associated with thoracic malignancy as well as to calculate the optimal cutoff values for SWE in this regard. Methods: Patients with a unilateral pleural effusion of unknown origin were included in the ED and subjected to a thoracic ultrasound (TUS) scan during their first 48 h after admittance. Two index tests were applied: (i) traditional B-mode TUS examination registering the presence of diaphragmatic nodules, pleural thickenings and other findings associated with malignancy and (ii) an SWE examination of different regions of interest. The reference test was defined as the subsequent diagnosis of malignant pleural effusion (MPE) in the three months following inclusion. Results: In total, 39 patients were included. The B-mode TUS index test yielded a sensitivity of 28.57% (95%CI 3.67-70.96%) and a specificity of 90.62% (95%CI 74.98-98.02%). The SWE max of the intercostal space yielded a sensitivity of 100% (95%CI 47.82-100%) and a specificity of 59.09% (95%CI 36.35-79.29%). Conclusions: A TUS with integrated SWE may aid in identifying MPEs and improving referrals for the diagnostic work-up of underlying malignancy. Larger, adequately powered studies are warranted.

Background: The aim of our study was to evaluate the retinal nerve fiber layer (RNFL) and macular and choroidal thicknesses in women with different phenotypes of polycystic ovary syndrome (PCOS), and compare these measurements with those of healthy women of reproductive age. Materials and Methods: This prospective case-control study included 120 eyes of 120 women with PCOS, with each of the four distinct phenotypes comprising 30 eyes of 30 women. Additionally, 30 eyes from 30 healthy women were included in the control group. All participants underwent comprehensive ophthalmologic examinations, and RNFL thickness, macular thickness (MT), and choroidal thickness (CT) in each eye were measured via spectral-domain optical coherence tomography. The body mass index (BMI) of the patients was recorded and compared with the retinal changes. Results: The average mean and nasal segments of the RNFL were significantly greater in the PCOS group than in the healthy control group (p < 0.001). There was a statistically significant difference in foveal retinal thickness between the groups (p < 0.001). Our study revealed significant choroidal tissue thickening subfoveally and at locations 500 μm temporal, 500 μm nasal, 1500 μm nasal, and 1500 μm temporal to the fovea in all phenotypes of the PCOS group (p < 0.001). Additionally, there was a positive correlation between BMI and all CT changes. Conclusions: Our findings indicate that the retinal layers and choroid are affected by all phenotypes of PCOS, one of the most common reproductive abnormalities, albeit to varying degrees. Furthermore, these changes were found to be correlated with BMI.

Background/Objectives: Osteoporotic fractures, particularly subcapital hip fractures (SCF), pose a significant healthcare and economic burden. The routine pathological examination of resected femoral heads in such cases is common practice, aimed at identifying malignancies that may have contributed to bone fragility. This study evaluated the cost-effectiveness and clinical utility of routine femoral head pathology in patients undergoing surgical treatment for SCF. Methods: A retrospective cohort study was conducted at a university-affiliated, tertiary care hospital. Patients undergoing surgical treatment for SCF between 2015 and 2018, with available femoral head pathology reports, were included. Data on demographics, prior or active malignancies, surgical procedures, and pathology results were analyzed. Results: The study included 230 patients with a mean age of 82.4 ± 14.1 years, of whom 57% were female. A total of 72 (31%) patients had a history of malignancy at the time of surgery. Pathological examination identified malignancies in eight patients (3.4%), all of whom had active malignancies at the time of admission. The most common malignancies detected were breast cancer and multiple myeloma (three cases each). None of the findings led to changes in patient management. Conclusions: The routine pathological examination of femoral heads following SCF provided a limited diagnostic yield and did not alter clinical management in this cohort. These findings suggest that routine pathology may not be cost-effective and support the adoption of selective screening approaches based on clinical risk factors such as a history of malignancy or atypical fracture presentations.

Background: Oral diseases such as caries, gingivitis, and periodontitis are highly prevalent worldwide and often arise from plaque. This study focuses on detecting three plaque stages-new, mature, and over-mature-using state-of-the-art YOLO architectures to enhance early intervention and reduce reliance on manual visual assessments. Methods: We compiled a dataset of 531 RGB images from 177 individuals, captured via multiple mobile devices. Each sample was treated with disclosing gel to highlight plaque types, then preprocessed for lighting and color normalization. YOLOv9, YOLOv10, and YOLOv11, in various scales, were trained to detect plaque categories, and their performance was evaluated using precision, recall, and mean Average Precision (mAP@50). Results: Among the tested models, YOLOv11m achieved the highest mAP@50 (0.713), displaying superior detection of over-mature plaque. Across all YOLO variants, older plaque was generally easier to detect than newer plaque, which can blend with gingival tissue. Applying the O'Leary index indicated that over half of the study population exhibited severe plaque levels. Conclusions: Our findings demonstrate the feasibility of automated plaque detection with advanced YOLO models in varied imaging conditions. This approach offers potential to optimize clinical workflows, support early diagnoses, and mitigate oral health burdens in low-resource communities.

Background: Peritonsillar abscess (PTA) is relatively common but challenging to diagnose clinically. Several clinical signs may be used, with unknown performances. We evaluated and compared the diagnostic performance of individual and combined clinical signs (trismus, edema, pharynx immobility, uvula deviation, hot potato voice, and overall clinical impression) assessed by an otolaryngologist and of contrast-enhanced computed tomography (CT) to detect acute PTA. Methods: Prospective study in 133 consecutive adult patients (77 males, mean age = 33 years) with suspected clinical PTA and CT obtained in the emergency setting of a tertiary care hospital between November 2020 and October 2022. The standard of reference consisted of surgically proven pus within 24 h of CT or a favorable clinical evolution at 48 h without surgical intervention. Results: PTA was present in 117/133 (88%) patients, with no difference between mean age or sex distribution between the groups with and without PTA. None of the evaluated clinical signs were associated with PTA (OR = 1.26-5.43, p > 0.05), whereas the CT finding "abscess" was significantly associated with PTA (OR = 67.2, p < 0.0001). The sensitivity of individual clinical signs varied between 19.7% and 73.5%, and the sensitivity of CT was significantly higher for all clinical signs (95.7%, p < 0.0001) except for overall clinical impression (97.4%, p = 0.7266). The specificity of clinical signs varied between 12.5% and 93.8%, and the specificity of CT was significantly higher (75%, p < 0.05) for overall clinical impression and edema. All clinical signs together yielded an area under the curve (AUC) = 0.677. Conclusions: In adults, clinical assessment alone using independent clinical signs and overall clinical impression does not allow a reliable diagnosis of PTA, even when performed by an otolaryngologist. CT is reliable in diagnosing PTA and, whenever available, should be the examination method of choice for diagnosing PTA, especially by a non-specialist.

Background/Objectives: Although preeclampsia (PE) and small for gestational age (SGA) are known to come from impaired placentation during the first trimester, prior studies have focused mostly on Doppler findings in the second trimester. Methods: In this retrospective pilot study, we enrolled 628 singleton pregnant women who underwent ultrasound in both the first and second trimesters and blood test. For SGA correlation, we further excluded 12 subjects with PE because PE may be the cause of SGA. We first presented the reference range of parameters of uterine artery Doppler in the first trimester and then grouped the subjects according to the presence of SGA (presence = 104, absence = 512) or PE (presence = 12, absence = 616) and investigated the association of uterine artery Doppler findings and serum markers in the first trimester with the occurrence of SGA or PE. Results: The uterine artery pulsatility index and the resistance index and the proportion of uterine artery notch decreased progressively in the first trimester. A lower serum beta-hCG level in the first trimester predicted the occurrence of SGA (adjusted odds ratio [AOR] = 0.53, p = 0.019), while the presence of the uterine artery notch in the first trimester predicted the development of PE (notch at least on one side: AOR = 8.65, p = 0.045 and notch on both sides: AOR = 8.91, p = 0.047). Regardless of whether a notch was present in the second trimester, a uterine artery notch in the first trimester was associated with an excellent negative predictive value (99.6%) for PE. Conclusions: This study suggests the clinical importance of assessing serum beta-hCG and the uterine artery notch in the first trimester to predict SGA and PE.

Background/Objectives: Artificial intelligence (AI) is gaining an increasing amount of influence in various fields, including medicine. In radiology, where diagnoses are based on collaboration between diagnostic devices and the professional experience of radiologists, AI intervention seems much easier than in other fields, but this is often not the case. Many times, the patients orient themselves according to the doctor, which is not applicable in the case of AI. Another limitation rests in the controversies regarding medico-legal liability. In the field of radio-imaging in Romania, the implementation of AI systems in diagnosis is at its beginning. An important aspect of this is raising awareness among the population about these assistive AI systems and, also, awareness of the technological evolution of AI among medical staff. This narrative review manuscript analyzes the existing literature data regarding the medico-legal aspects of AI application in radiology, highlighting the controversial aspects and the lack of statutory legislative regulations in Romania. Methods: A detailed search was conducted across three electronic databases including MEDLINE/PubMed, Scopus, and Web of Science, with 53 papers serving as the literature corpus of our review. Results: General requirements for artificial intelligence systems used in radiology have been established. In the radiological diagnostic process, there are five levels of AI system implication. Until now, completely autonomous AI systems have not been used. Regarding liability in the case of malpractice, at the currently accepted legislative level, most of the time, the radiologist is liable for their own fault or non-compliant use of diagnostic AI systems. Controversies arise in the case of radio-imaging diagnosis in which AI systems act autonomously. Conclusions: In order for AI diagnostic radio-imaging systems to be implemented, they must meet certain quality standards and be approved. The radiologist must know these systems, accept them, know their limits, and validate them in accordance with their degree of involvement in radiological diagnosis. Considering the evolution of technology in the Romanian medical system, including radiology, in the future, an alignment with the legal standards established/proposed at the European level is desired.

COVID-19 has caused widespread morbidity and mortality, with its effects extending to multiple organ systems. Despite known risk factors for severe disease, including advanced age and underlying comorbidities, patient outcomes can vary significantly. This variability complicates efforts to predict disease progression and tailor treatment strategies. While diagnostic and therapeutic approaches are still under debate, RNA sequencing (RNAseq) has emerged as a promising tool to provide deeper insights into the pathophysiology of COVID-19 and guide personalized treatment. A comprehensive literature review was conducted using PubMed, Scopus, Web of Science, and Google Scholar. We employed Medical Subject Headings (MeSH) terms and relevant keywords to identify studies that explored the role of RNAseq in COVID-19 diagnostics, prognostics, and therapeutics. RNAseq has proven instrumental in identifying molecular biomarkers associated with disease severity in patients with COVID-19. It allows for the differentiation between asymptomatic and symptomatic individuals and sheds light on the immune response mechanisms that contribute to disease progression. In critically ill patients, RNAseq has been crucial for identifying key genes that may predict patient outcomes, guiding therapeutic decisions, and assessing the long-term effects of the virus. Additionally, RNAseq has helped in understanding the persistence of viral RNA after recovery, offering new insights into the management of post-acute sequelae, including long COVID. RNA sequencing significantly improves COVID-19 management, particularly for critically ill patients, by enhancing diagnostic accuracy, personalizing treatment, and predicting therapeutic responses. It refines patient stratification, improving outcomes, and holds promise for targeted interventions in both acute and long COVID.

Background/Objectives: Earlier detection of severe immune-related hematological adverse events (irHAEs) in cancer patients treated with a PD-1 or PD-L1 inhibitor is critical to improving treatment outcomes. The study aimed to develop a simple machine learning (ML) model for predicting irHAEs associated with PD-1/PD-L1 inhibitors. Methods: We utilized the Observational Medical Outcomes Partnership-Common Data Model based on electronic medical records from a tertiary (KHMC) and a secondary (KHNMC) hospital in South Korea. Severe irHAEs were defined as Grades 3-5 by the Common Terminology Criteria for Adverse Events (version 5.0). The predictive model was developed using the KHMC dataset, and then cross-validated against an independent cohort (KHNMC). The full ML models were then simplified by selecting critical features based on the feature importance values (FIVs). Results: Overall, 397 and 255 patients were included in the primary (KHMC) and cross-validation (KHNMC) cohort, respectively. Among the tested ML algorithms, random forest achieved the highest accuracy (area under the receiver operating characteristic curve [AUROC] 0.88 for both cohorts). Parsimonious models reduced to 50% FIVs of the full models showed comparable performance to the full models (AUROC 0.83-0.86, p > 0.05). The KHMC and KHNMC parsimonious models shared common predictive features including furosemide, oxygen gas, piperacillin/tazobactam, and acetylcysteine. Conclusions: Considering the simplicity and adequate predictive performance, our simplified ML models might be easily implemented in clinical practice with broad applicability. Our model might enhance early diagnostic screening of irHAEs induced by PD-1/PD-L1 inhibitors, contributing to minimizing the risk of severe irHAEs and improving the effectiveness of cancer immunotherapy.