Search Results (247)

Objectives: This study aimed to develop and validate a biparametric MRI (bpMRI)-based radiomics model for the noninvasive prediction of tumor–stroma ratio (TSR) in prostate cancer (PCa). Additionally, we sought to explore lesion distribution patterns in the peripheral zone (PZ) and transition zone (TZ) to provide deeper insights into the biological behavior of PCa. Methods: This multicenter retrospective study included 223 pathologically confirmed PCa patients, with 146 for training and 39 for internal validation at Center 1, and 38 for external testing at Center 2. All patients underwent preoperative bpMRI (T2WI, DWI acquired with a b-value of 1400 s/mm², and ADC maps）, with TSR histopathologically quantified. Regions of interest (ROIs) were manually segmented on bpMRI images using ITK-SNAP software (version 4.0.1), followed by high-throughput radiomic feature extraction. Redundant features were eliminated via Spearman correlation analysis and least absolute shrinkage and selection operator (LASSO) regression. Five machine learning (ML) classifiers—Logistic Regression (LR), Support Vector Machine (SVM), BernoulliNBBayes, Ridge, and Stochastic Gradient Descent (SGD)—were trained and optimized. Model performance was rigorously evaluated using receiver operating characteristic (ROC) curves and decision curve analysis (DCA). Results: The Ridge demonstrated superior diagnostic performance, achieving AUCs of 0.846, 0.789, and 0.745 in the training, validation, and test cohorts, respectively. Lesion distribution analysis revealed no significant differences between High-TSR and Low-TSR groups (p = 0.867), suggesting that TSR may not be strongly associated with zonal localization. Conclusions: This exploratory study suggests that a bpMRI-based radiomic model holds promise for noninvasive TSR estimation in prostate cancer and may provide complementary insights into tumor aggressiveness beyond conventional pathology. Full article

In histopathological image segmentation, existing methods often show low sensitivity to small lesions and indistinct boundaries, leading to missed detections. Since, in clinical diagnosis, the consequences of missed detection are more serious than false alarms, this study proposes ClinSegNet, a recall-oriented and human-centred framework for reliable histopathology screening. ClinSegNet employs a composite optimisation strategy, termed HistoLoss, which balances stability and boundary refinement while prioritising recall. An uncertainty-driven refinement mechanism is further introduced to target high-uncertainty cases with limited fine-tuning cost. In addition, a clinical data processing pipeline was developed, where pixel-level annotations were automatically derived from IHC-to-H&E mapping and combined with public datasets, enabling effective training under limited clinical data conditions. Experiments on the NuInsSeg and NuInsSeg-UHS datasets showed that ClinSegNet achieved recall scores of 0.8803 and 0.8917, further improved to 0.8983 and 0.9053 with HITL refinement, while maintaining competitive Dice and IoU. Comparative and ablation studies confirmed the complementary design of the framework and its advantage in capturing small or complex lesions. In conclusion, ClinSegNet provides a clinically oriented, recall-prioritised framework that enhances lesion coverage, reduces the risk of missed diagnosis, and offers both a methodological basis for future human-in-the-loop systems and a feasible pipeline for leveraging limited clinical data. Full article

Background/Objectives: Despite multimodal therapeutic concepts, treatment of recurrent malignant gliomas remains challenging. Stereotactic radiosurgery (SRS) may be a possible safe and effective non-invasive salvage treatment. In this study, we aim to investigate the SRS treatment outcomes using partly 18F-Fluorethylthyrosine (FET)-PET-imaging sequences for SRS treatment planning focusing on overall survival, event-free survival, and the incidence and factors influencing radiation necrosis (RN) occurrence. Additionally, we evaluated the potential application of AI-based tumor segmentation. Methods: We conducted a retrospective analysis of patients with recurrent malignant glioma treated with single-fraction or hypofractionated SRS at our institution. The outcomes assessed included local control, overall survival (OS), and local event-free survival (LEFS, defined as the interval until tumor recurrence or the onset of RN). We also performed a simulation analysis to assess the potential of AI-based tumor segmentation. Results: The study included 27 patients with a median age of 57 years and 41 lesions. The median OS post-SRS was 9.6 months and an LEFS of 5.2 months. Factors positively influencing OS and LEFS included the gross tumor volume (GTV) of the lesions before SRS therapy, presence of an IDH mutation, and lomustine treatment post-SRS. The incidence of RN post-SRS was 31.7%. RN was confirmed histopathologically in 15.4%, based on MRI in 46.2% and by FET-PET in 38.5% of lesions. In a simulation analysis, AI-based tumor segmentation reliably delineated all lesions, requiring only minimal manual adjustments to define target volumes. Conclusions: High-dose SRS is a feasible salvage treatment for small-volume recurrent high-grade gliomas, achieving local control and survival outcomes comparable to other re-irradiation strategies. IDH mutation, smaller tumor volume, and lomustine therapy were associated with improved survival. RN occurred frequently, particularly in periventricular lesions. AI-based tumor segmentation showed promise in well-defined satellite recurrences, but remains limited in cavity-adjacent lesions, underlining the need for expert review and ¹⁸FET-PET imaging. Full article

The intensity and prevalence of different gastrointestinal nematode species vary across regions worldwide. Oesophagostomum radiatum commonly shows a high occurrence in young cattle. O. radiatum causes anaemia, hypoproteinaemia, and immunopathological changes in the large intestine wall, impairing calves’ body weight gain. This study aimed to assess the impact of natural O. radiatum infection on haematological parameters and immune responses in 23 Nellore calves, considering sex-based differences. Assessments included Oesophagostomum egg count (EPG), worm count, packed cell volume (PCV), total plasma protein, histopathological and immunohistochemistry analyses. A large number of parasites attached to the colon mucosa were observed, along with massive nodule formation and haemorrhagic lesions, mainly within a 20–30 cm-long segment adjacent to the nodules. The maximum mean egg shedding was approximately 165 EPG for males and 173 EPG for female calves; however, males presented a significantly higher worm count (969 ± 200.5) than females (460 ± 99.5). There were significant positive correlations between the total O. radiatum worm count and O. radiatum EPG for both female and male calves. Significant negative correlations were observed between the total O. radiatum worm count and PCV in female calves. Our results demonstrated that natural O. radiatum infection in Nellore calves induced marked immunopathological alterations, including chronic inflammatory responses that impaired intestinal function. Sex-related differences suggested that female calves may develop more effective tissue responses. These findings emphasise the economic impact of subclinical infections and reinforce the importance of control strategies to minimise productivity losses in cattle. Full article

Objective: This study aims to evaluate the management and outcomes over 14 years at a tertiary maxillofacial surgery unit. Methods: Retrospective cohort study of patients from a Portuguese tertiary center of maxillofacial surgery with histopathologically confirmed diagnoses of chronic osteomyelitis of the jaws between January 2010 and December 2023. Demographic and clinical characteristics, treatment, and progression of the disease were evaluated. Results: Fifty-three patients were included—28 women (52.8%), mean age 55 (95% CI 5–90) years. The mandible was affected in 84.9% (n = 45) of cases. Secondary chronic osteomyelitis was diagnosed in 88.7% (n = 47), with medication-related osteonecrosis of the jaw (MRONJ) being the most common etiology (38.3%). Bacteriological samples were contributory in 52% (n = 13) and 46.1% (n = 6) were resistant to amoxicillin. All received antibiotics for a median time of 27.3 days. Surgical treatment included sequestrectomy (n = 40, 75.5%), marginal (n = 5, 9.4%), and segmental mandibulectomy (n = 8, 15.1%). Clinical remission was achieved in 77.4% (n = 41) of cases with higher success in MRONJ (n = 15, 83.3%) than ORN (n = 4, 57.1%). Conclusions: Almost half of the isolates were amoxicillin-resistant, reinforcing the need for susceptibility testing. Surgical management guided by etiology and disease stage remains essential, with more extensive resection needed in MRONJ and ORN. Full article

Accurate preoperative staging of bladder cancer on MRI remains challenging because visual reads vary across observers. We investigated a multiparametric MRI (mpMRI) radiomics approach to predict muscle invasion (≥T2) and prospectively tested it on a validation cohort. Eighty-four patients with urothelial carcinoma underwent 1.5-T mpMRI per VI-RADS (T2-weighted imaging and DWI-derived ADC maps). Two blinded radiologists performed 3D tumor segmentation; 37 features per sequence were extracted (LifeX) using absolute resampling. In the training cohort (n = 40), features that differed between non-muscle-invasive and muscle-invasive tumors (Mann–Whitney p < 0.05) underwent ROC analysis with cut-offs defined by the Youden index. A compact descriptor combining GLRLM-LRLGE from T2 and GLRLM-SRLGE from ADC was then fixed and applied without re-selection to a prospective validation cohort (n = 44). Histopathology within 6 weeks—TURBT or cystectomy—served as the reference. Eleven T2-based and fifteen ADC-based features pointed to invasion; DWI texture features were not informative. The descriptor yielded AUCs of 0.934 (training) and 0.871 (validation) with 85.7% sensitivity and 96.2% specificity in validation. Collectively, these findings indicate that combined T2/ADC radiomics can provide high diagnostic accuracy and may serve as a useful decision support tool, after multicenter, multi-vendor validation. Full article

Background/Objectives: Colorectal cancer (CRC) is a major global health concern, with rising incidence across age groups. Early detection via colonoscopy and identification of precancerous polyps are crucial for prevention and improved outcomes. The objectives were to evaluate the epidemiology, anatomical distribution, morphology, and histopathology of CRC, and its association with synchronous colorectal polyps. Methods: In 2023, a retrospective study was conducted on 1973 patients undergoing colonoscopy due to symptoms like blood in the stool, changes in bowel habits, abdominal pain, weight loss, anemia, or as CRC follow-up. Complete colonoscopies were performed, and suspicious lesions were biopsied or resected for histological evaluation. Statistical analysis was performed using SPSS 11.0. Results: CRC was diagnosed in 78 patients (3.95%), with a male predominance (70.51%, p < 0.05) and a mean age of 65.1 ± 8.9 years. The most affected age group was 61–70 years (43.58%). Tumors were most commonly located in the rectum (32.05%) and sigmoid colon (26.92%). Polypoid morphology was observed in 67.95% of cases. Adenocarcinoma was the predominant histological type (93.59%), followed by mucinous adenocarcinoma (6.41%), with significant differences between right and left colon (p < 0.001). Synchronous polyps were detected in 47.43% of CRC cases, primarily adenomas (60.22%). In 37.84%, the tumor and polyp were in the same colon segment. Men had a higher rate of synchronous polyps than women (p < 0.05). Conclusions: CRC is more common in older males and typically affects the rectosigmoid region. Adenocarcinoma is the leading type. Nearly half of patients had synchronous adenomas, highlighting the importance of full colonoscopy for early CRC detection and prevention. Full article

Precise segmentation of glands in histopathological images is essential for the diagnosis of colorectal cancer, as the changes in gland morphology are associated with pathological progression. Conventional computer-assisted methods rely on dense pixel-level annotations, which are costly and labor-intensive to obtain. The present study proposes a two-stage weakly supervised segmentation framework named Multi-Level Attention and Affinity (MAA). The MAA framework utilizes the image-level labels and combines the Multi-Level Attention Fusion (MAF) and Affinity Refinement (AR) modules. The MAF module extracts the hierarchical features from multiple transformer layers to grasp global semantic context, and generates more comprehensive initial class activation maps. By modeling inter-pixel semantic consistency, the AR module refines pseudo-labels, which can sharpen the boundary delineation and reduce label noise. The experiments on the GlaS dataset showed that the proposed MAA framework achieves the Intersection over Union (IoU) of 81.99% and Dice coefficient of 90.10%, which outperformed the state-of-the-art Online Easy Example Mining (OEEM) method with an improvement of 4.43% in IoU. Such experimental results demonstrated the effectiveness of integrating hierarchical attention mechanisms with affinity-guided refinement for annotation-efficient and robust gland segmentation. Full article

Brain gliomas are highly infiltrative and heterogenous tumors, whose early and accurate detection as well as therapeutic management are challenging. Artificial intelligence (AI) has the potential to redefine the landscape in neuro-oncology and can enhance glioma detection, imaging segmentation, and non-invasive molecular characterization better than conventional diagnostic modalities through deep learning-driven radiomics and radiogenomics. AI algorithms have been shown to predict genotypic and phenotypic glioma traits with remarkable accuracy and facilitate patient-tailored therapeutic decision-making. Such algorithms can be incorporated into surgical planning to optimize resection extent while preserving eloquent cortical structures through preoperative imaging fusion and intraoperative augmented reality-assisted navigation. Beyond resection, AI may assist in radiotherapy dose distribution optimization, thus ensuring maximal tumor control while minimizing surrounding tissue collateral damage. AI-guided molecular profiling and treatment response prediction models can facilitate individualized chemotherapy regimen tailoring, especially for glioblastomas with MGMT promoter methylation. Applications in immunotherapy are emerging, and research is focusing on AI to identify tumor microenvironment signatures predictive of immune checkpoint inhibition responsiveness. AI-integrated prognostic models incorporating radiomic, histopathologic, and clinical variables can additionally improve survival stratification and recurrence risk prediction remarkably, to refine follow-up strategies in high-risk patients. However, data heterogeneity, algorithmic transparency concerns, and regulatory challenges hamstring AI implementation in neuro-oncology despite its transformative potential. It is therefore imperative for clinical translation to develop interpretable AI frameworks, integrate multimodal datasets, and robustly validate externally. Future research should prioritize the creation of generalizable AI models, combine larger and more diverse datasets, and integrate multimodal imaging and molecular data to overcome these obstacles and revolutionize AI-assisted patient-specific glioma management. Full article

Objectives: This study examines the influence of the uterine wall region from which a tumor originates on the prognosis of endometrial cancers (ECs) originating from the uterine corpus without lower uterine segment involvement (LUSI). Methods: This study included 68 EC patients who were operated on between January 2015 and January 2022 and met the following criteria: no LUSI noted in the final pathology report; endometrioid adenocarcinoma originating from the uterine corpus, as noted in the histopathology report; and Stages 1A and 1B tumors, classified according to the FIGO 2009 staging system. From the final pathology results after the operation, these patients were divided into four groups according to the uterine wall region from which the tumor originated: the anterior wall, posterior wall, fundus, and lateral wall. Results: No statistically significant difference was recorded between the groups regarding age, BMI, comorbidity, nulliparity, stage, grade, tumor size, DFS, and OS (p > 0.05). LVSI was detected at a higher rate in the anterior and posterior wall groups than in the fundus and lateral wall groups (p = 0.038). Similarly, the rates of recurrence and death were statistically notably higher in the anterior and posterior wall groups in comparison with those in the other two groups (p < 0.001). Conclusions: In endometrioid adenocarcinomas of the uterine corpus, tumors originating from the anterior and posterior walls have a worse prognosis than those originating from the fundus and lateral wall. Closer follow-up may be necessary for endometrial cancers originating from the anterior and posterior uterine walls. Full article

This study presents the development of a quantitative evaluation method utilizing machine vision technology to characterize the extent of body surface damage in largemouth bass (Micropterus salmoides) infected with largemouth bass ranavirus (LMBV). High-resolution, multi-angle images (6000 × 4000 pixels) of the body surface from 239 infected specimens were acquired at a fixed distance of 40 cm using a SONY ILCE-7RM3 digital camera within a GODOX-LST60 softbox. Key parameters, including the number of segmented injury areas, the count of body surface lesions, and the total lesion area, were analyzed. These parameters were integrated through principal component analysis (PCA) to construct a comprehensive damage scoring model. The severity of viral-induced body surface damage was categorized into four grades: uninjured (0), minor injury (1), moderate injury (2), and severe injury (3). Histopathological examination revealed that early-stage infection (grade 1) predominantly exhibited localized hemorrhagic spots in the muscular region of the body side (B/E region) with limited lesion area. In contrast, moderate to severe infections (grades 2–3) were characterized by extensive ulceration, muscle necrosis, and visceral lesions, including hepatic fibrosis and splenic granulomatous formations. Quantitative real-time PCR (qRT-PCR) analysis demonstrated a progressive upregulation of pro-inflammatory cytokines (IL-6, IL-8, TNF-α, CXCL2) in immune organs, concomitant with increased expression of apoptosis-related genes (CASP8, CYC). This study successfully established a rapid and objective quantitative grading system for ranavirus infection, offering a novel technical approach for early diagnosis and precise prevention and control strategies against largemouth bass ranavirus. Full article

Focal segmental glomerulosclerosis (FSGS) is a histopathological pattern of segmental glomerulosclerosis that arises from diverse primary and secondary causes. Primary (idiopathic) FSGS is rare and is often linked to intrinsic podocyte injury, while secondary forms are more prevalent and may reflect adaptative, toxic, genetic, or viral etiologies. This pattern of injury can lead to progressive renal dysfunction and, in some cases, end-stage kidney disease. The pathophysiology is multifactorial and includes direct podocyte injury (e.g., genetic defects, mechanical or toxic injury), immune-mediated processes (e.g., circulating permeability factors, inflammatory mediators), and metabolic disturbances. In particular, disturbance of lipid metabolism, including intracellular cholesterol accumulation in podocytes, have been implicated as a contributory mechanism in podocyte dysfunction and progression of disease in proteinuric/nephrotic presentations and in specific disease subtypes. Diagnosis relies on clinical assessment, laboratory testing, and histological examination, with kidney biopsy remaining the gold standard. Conventional treatments include corticosteroids, and other immunosuppressants when indicated, and measures to reduce proteinuria and control blood pressure, but the therapeutic response is variable and many patients show progression, highlighting the need for more effective and novel therapeutic approaches. Cyclodextrins (CDs), widely used as drug carriers to enhance solubility, can also mobilize and promote efflux of cholesterol from cells. Preclinical studies show that CDs reduce renal lipid accumulation and ameliorate podocyte injury in experimental models, supporting the idea that CDs could have a dual role as drug carriers and as direct modulators of lipid-related podocyte injury in lipid-associated forms of FSGS. Given the limited direct clinical data in FSGS, in this article we discuss the biological rationale, preclinical evidence, and remaining knowledge gaps for exploring CDs as an innovative therapeutic strategy. Full article

Introduction: Small Bowel Obstruction (SBO) accounts for 15% of emergency department (ED) admissions. While conservative management is recommended, surgery becomes necessary when strangulation is suspected. Identifying which patients need surgery remains a challenge, as traditional imaging lacks sufficient sensitivity and specificity. This study aimed to explore radiomic features to identify potential predictors of strangulation. Methods: This retrospective study included patients admitted to a tertiary referral hospital ED between 2019 and 2023, diagnosed with Adhesion Small Bowel Obstruction (aSBO) via contrast-enhanced abdominal CT scans. Two patient groups were examined: those who underwent surgery with bowel resection and ischemic changes confirmed histologically (operative management—OM) and those successfully treated with conservative management (CM). All CT scans were reviewed blindly by a general surgeon and an experienced radiologist. Pre-obstructive loop segmentation was performed using 3D Slicer software, with slice-by-slice contouring of intestinal borders on images of suspected strangulated bowel. Radiomic features were extracted, followed by univariate and multivariate regression analysis. Results: A total of 55 patients were included: 27 CM and 28 OM. Significant differences emerged in GLCM (Gray Level Co-occurrence Matrix), GLDM (Gray Level Dependence Matrix), GLRLM (Gray Level Run Length Matrix), and GLSZM (Gray Level Size Zone Matrix), particularly involving entropy and uniformity. These metrics reflect subtle variations in gray levels not visible to the naked eye. Conclusions: Differences in entropy, uniformity, and energy align with imaging and histopathological findings, supporting the development of radiomic models and future AI-based prediction tools. Full article

Background/Objectives: Glioblastomas (GBMs) and brain metastases (BMs) are both frequent brain lesions. Distinguishing between them is crucial for suitable therapeutic and follow-up decisions, but this distinction is difficult to achieve, as it includes clinical, radiological and histopathological correlation. However, non-invasive AI examination of conventional and advanced MRI techniques can overcome this issue. Methods: We retrospectively selected 78 patients with confirmed GBM (39) and single BM (39), with conventional MRI investigations, consisting of T2W FLAIR and CE T1W acquisitions. The MRI images (DICOM) were evaluated by an AI segmentation tool, comparatively evaluating tumor heterogeneity and peripheral edema. Results: We found that GBMs are less edematous than BMs (p = 0.04) but have more internal necrosis (p = 0.002). Of the BM primary cancer molecular subtypes, NSCCL showed the highest grade of edema (p = 0.01). Compared with the ellipsoidal method of volume calculation, the AI machine obtained greater values when measuring lesions of the occipital and temporal lobes (p = 0.01). Conclusions: Although extremely useful in radiomics analysis, automated segmentation applied alone could effectively differentiate GBM and BM on a conventional MRI, calculating the ratio between their variable components (solid, necrotic and peripheral edema). Other studies applied to a broader set of participants are necessary to further evaluate the efficacy of automated segmentation. Full article

Canine mammary tumors (CMTs) are among the most frequently diagnosed neoplasms in female dogs, with complex etiologies involving age, breed, body size, and reproductive status. This retrospective study analyzed 250 cases comprising 361 mammary tumors to evaluate demographic patterns, histopathological subtypes, and their topographical distribution. The majority of tumors (88.64%) were malignant, with complex carcinoma being the most common subtype (24%), followed by intraductal papillary carcinoma (14.95%). The mean age at diagnosis was 9.36 years, with 60.8% of cases occurring in dogs older than 8 years. Mixed-breed dogs were most frequently affected, followed by Bichon-type breeds and German Shepherds. Small-sized dogs (<10 kg) accounted for 43.5% of cases. Regarding the reproductive status, 35.6% of dogs were unspayed, while the remainder were either intact or had an undetermined status, reflecting regional spaying practices. A statistically significant association was identified between histological subtype and mammary gland location, revealing non-random, region-specific tumor distribution. Complex carcinomas were frequently located in the left canial abdominal mammary segment (L.A2), right canial abdominal mammary segment (R.A2), and left inguinal mammary segment (L.ING); mixed and tubular carcinomas were more prevalent in R.A2 and R.ING; and tubulopapillary carcinomas appeared predominantly in L.ING. While no significant overall association was found between tumor grade and gland location, grade III carcinomas showed a possible predilection for the left inguinal gland (L.INGH), indicating a regional tendency that warrants further investigation. These findings underscore the importance of anatomical factors in the development of CMTs and support the integration of gland-specific tumor patterns into diagnostic assessment and surgical decision-making. Full article