Background: Accurate brain tumor detection is critical for early diagnosis and effective treatment planning in neuro-oncology. Magnetic resonance imaging (MRI) is a cornerstone for identifying and localizing brain tumors, guiding clinical interventions, and enhancing patient outcome. Advanced deep learning models, such as YOLOv8, offer promising solutions for automated and precise tumor detection in MRI. Objectives: This study aimed to develop and evaluate a YOLOv8-based model for the accurate identification and localization of brain tumors, including pituitary tumors, meningiomas, and gliomas, using a publicly available Kaggle dataset of annotated MRI images. Materials and methods: The YOLOv8 pretrained model was employed for brain tumor detection on a Kaggle dataset comprising annotated MRI images, including cases with pituitary, meningioma, and glioma, and no tumor. The dataset was pre-processed and split into training and validation sets for further analysis. The YOLOv8 model was fine-tuned to optimize the tumor detection and localization. Performance metrics, including precision, recall, F-1 score, and mean average precision (mAP), were calculated, and loss values were analyzed to evaluate the model’s learning efficiency. Results: The YOLOv8 model achieved a precision of 98.9%, recall of 98.9%, and accuracy of 99.5%. The mean average precision (mAP) reached 97.6%, indicating a high accuracy in detecting and localizing brain tumors. Loss value analysis demonstrated stable convergence during training, reflecting a robust model performance. Conclusion: The YOLOv8-based approach provides a highly accurate and reliable method for detecting and localizing brain tumors in MRI. With exceptional precision, recall, and mAP, this model demonstrates significant potential for clinical applications, enabling faster and more precise neurooncological diagnosis and treatment planning.

Robust non-rigid image-to-patient registration for contactless dynamic thoracic tumor localization using recursive deformable diffusion models.

This study aims to evaluate health-related quality of life (HRQoL) in patients who underwent surgical treatment for upper extremity bone tumors, using validated patient-reported outcome measures, and to explore clinical factors associated with better or worse outcomes. Between January 2015 and January 2024, a total of 55 patients (26 males, 29 females; mean age: 42.4 ± 19.0 years; range, 16 to 81 years) who were treated surgically for upper extremity bone tumors and evaluated at least six months postoperatively were included. The patients completed the EuroQol EQ-5D-5L and the Short Form-12 (SF-12) questionnaires. Clinical variables included age, sex, tumor site, pathology, treatment modality, and adjuvant or neoadjuvant therapies. The humerus was the most frequent tumor site (60%), and joint involvement was present in 80% of cases. Benign tumors accounted for two-thirds of patients, while malignant tumors represented one-third. Curettage-based procedures predominated, and only two patients required amputation. Postoperative complications and recurrences were both observed in 7.3% of patients. On the EQ-5D-5L, 65.5% of patients reported no difficulty with mobility or self-care, but half experienced limitations in usual activities. Pain and discomfort were reported by 74.5%, and anxiety or depression by 67.3%. The mean EQ-5D-5L index was 0.55 ± 0.49 and the mean EQ-VAS was 70.9 ± 20.1. The mean SF-12 physical component score (PCS) and mental component score (MCS) were 42.6 ± 11.3 and 47.0 ± 11.2, respectively, indicating that physical functioning was more impaired than mental well-being. Correlation analyses demonstrated strong associations between EQ-5D-5L, EQ-VAS, and SF-12 scores, supporting convergent validity. Subgroup analyses revealed that younger patients (< 50 years), those with benign pathology, and individuals without adjuvant or neoadjuvant therapies reported significantly higher HRQoL scores. Patients undergoing surgery for upper extremity bone tumors frequently experience pain and psychological distress despite preserved independence in mobility and self-care. Age, pathology, treatment intensity and tumor location were associated with HRQoL. These findings highlight the importance of rehabilitation and psychosocial support in postoperative care, alongside oncological and surgical management.

The Impact of Ultracentral Tumor Location on Outcomes in Patients with Pulmonary Oligometastases: A Secondary Analysis of the Single-Arm Phase 2 SABR-5 Trial.

Sterotactic ablative radiotherapy vs. thermal ablation of localized renal cell carcinoma: Is there a preferred second-line management option?

Redefining the role of carbon nanoparticles in colorectal cancer surgery: From lymph node yield to diagnostic paradox.

αvβ3 integrin-targeted IR-820 dye-based small molecule probe for fluorescence imaging of tumor.

RF78. Feasibility of Panitumumab-IRDye800 molecular imaging for intraoperative tumor localization of lung cancer

One Year After a Cyberattack: Lessons Learned and Dosimetric Analysis of Contingency Radiotherapy Plans.

BACKGROUND Primary hepatocellular carcinoma (HCC) is a common malignancy worldwide, with surgical resection being the most effective treatment for long-term survival. For complex HCC (large diameter, central location, multiple lesions, or proximity to major vessels), traditional surgery relying on surgeon experience and palpation has limitations including difficulty identifying deep lesions and high positive margin rates. Intraoperative ultrasound (IOUS) provides real-time hepatic structural visualization but has limited capability for isoechoic lesions, while indocyanine green (ICG) fluorescence imaging enables real-time tumor visualization but lacks deep anatomical information. Combined application may offer complementary advantages, yet systematic evaluation studies in complex HCC resection remain scarce. AIM To evaluate the clinical application value of IOUS combined with ICG fluorescence imaging technology in radical resection of complex HCC. METHODS Clinical data of 200 patients with complex HCC who underwent radical hepatectomy from January 2019 to August 2024 were retrospectively analyzed. Patients were divided into a combined navigation group (n = 103) and a conventional surgery group (n = 97) based on whether IOUS and ICG fluorescence imaging technology were used in combination. Intraoperative indicators, oncological indicators, postoperative recovery indicators, and long-term prognosis were compared between the two groups. Logistic regression analysis was used to analyze factors influencing postoperative complications, and Cox regression analysis was used to analyze factors influencing survival prognosis. RESULTS Baseline characteristics were balanced between the two groups. The combined navigation group had shorter tumor localization time (P &lt; 0.001), less intraoperative blood loss (P = 0.004), lower intraoperative transfusion rate (P = 0.021), higher detection rate of occult lesions (23.3% vs 6.2%, P &lt; 0.001), and higher anatomical resection rate (P = 0.040). The combined navigation group had lower positive margin rate (2.9% vs 13.4%, P = 0.006), higher R0 resection rate (97.1% vs 86.6%, P = 0.006), and greater margin distance (P &lt; 0.001). The combined navigation group had lower overall postoperative complication rate (18.4% vs 28.9%, P = 0.042) and severe complication rate (5.8% vs 12.4%, P = 0.042), and shorter postoperative hospital stay (P = 0.003). With a median follow-up of 28.6 months, the combined navigation group had higher 2-year overall survival rate (76.8% vs 65.2%, P = 0.033) and 2-year disease-free survival rate (58.4% vs 45.7%, P = 0.022), and lower postoperative recurrence rate (36.9% vs 50.5%, P = 0.048). Multivariate analysis showed that application of combined navigation technology was an independent protective factor for postoperative complications (odds ratio = 0.498, P = 0.027), and was also an independent protective factor for overall survival (hazard ratio = 0.584, P = 0.028) and disease-free survival (hazard ratio = 0.631, P = 0.025). CONCLUSION IOUS combined with ICG fluorescence imaging technology can improve tumor localization accuracy in complex HCC surgery, improve margin control, reduce intraoperative blood loss, decrease the incidence of postoperative complications, and improve long-term survival prognosis in patients, demonstrating important clinical application value.

Background and objective: Spontaneous rupture of hepatocellular carcinoma (HCC) is a life-threatening complication requiring urgent intervention. Existing studies have identified prognostic factors for ruptured HCC, but a gap exists in understanding these factors in Northeast Thailand. This research aims to investigate survival outcomes and prognostic factors in patients with spontaneous rupture of HCC treated with transarterial embolization (TAE) in the Northeastern region. Methods: This was a retrospective study including 163 patients with spontaneous rupture of HCC who were treated with TAE at Srinagarind Hospital from January 2018 to December 2022. The diagnosis of spontaneous rupture of HCC was based on clinical presentation and dynamic liver computed tomography findings. Survival outcomes were evaluated using the Kaplan-Meier method, and Cox regression analysis was used to identify associated prognostic factors. Results: The median overall survival was 143 days. The 3-month, 6-month, and 12-month cumulative overall survival rates were 60.14%, 40.06%, and 29.37%, respectively. In multivariable analysis, factors significantly associated with improved survival included absence of comorbidity (adjusted HR 0.64; 95% CI 0.44–0.93; p=0.018), higher estimated glomerular filtration rate (eGFR) (adjusted HR 0.98; 95% CI 0.98–0.99; p&lt;0.001), lower Model for End-Stage Liver Disease (MELD) score (adjusted HR 0.96; 95% CI 0.93–0.99; p=0.034), and tumor location at hepatic segment 7 (adjusted HR 0.63; 95% CI 0.40–0.99; p=0.043). Tumor location at segment 6 was associated with poorer survival (adjusted HR 1.60; 95% CI 1.05–2.44; p=0.028). The absence of inotropic or vasopressor drug use was not significantly associated with survival in multivariable analysis. Conclusion: TAE is an effective initial treatment for patients with spontaneous rupture of HCC. Favorable prognostic factors identified in this study included absence of comorbidity, higher eGFR, lower MELD score, and tumor location in hepatic segment 7, whereas tumor location in segment 6 was associated with poorer survival. Early TAE (≤24 hours) demonstrated a trend toward improved survival.

Intramedullary spinal cord tumors (IMSCTs) are typically treated with maximal safe resection, during which neurosurgeons often monitor for neurological injury using muscle motor evoked potential (mMEP) and direct wave (D-wave) neuromonitoring. The predictive value of changes in D-waves for identifying motor outcomes is underexplored. This study evaluated the utility of D-waves for predicting postoperative motor deficits. Patients who underwent resection of a primary IMSCT with mMEP neuromonitoring from 2003 to 2023 at a tertiary care hospital were identified. Patients who underwent D-wave monitoring in addition to mMEP monitoring were compared to those who underwent mMEP monitoring alone using the Mann-Whitney U-test, chi-square test, and Fisher's exact test. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of D-wave and mMEP monitoring for identifying new neurological deficits immediately postoperatively and at 1 month, 6 months, and last follow-up were calculated. After matching, 125 patients were included (median age 42.0 years; 57.6% male; median follow-up 34.0 months), of whom 88 had both mMEP and D-wave data. The most common pathologies were ependymoma (64.0%) and astrocytoma (17.6%). Patients who did and did not undergo D-wave neuromonitoring had similar preoperative neurological function, primary pathology, tumor grade, and tumor location. D-wave use was associated with increased gross-total resection (88.6% vs 64.9%, p = 0.002) and reduced mortality (5.7% vs 24.3%, p = 0.007), length of stay (5.0 vs 6.0 days, p = 0.033), and 30-day readmission (2.3% vs 13.5%, p = 0.013) and reoperation (1.1% vs 10.8%, p = 0.012). At the 6-month follow-up, D-wave monitoring alone was superior to mMEP and combination monitoring for detecting new motor deficits. D-wave monitoring had peak sensitivity (77.8%) and NPV (96.5%) at 6 months and peak specificity (95.8%) and PPV (76.9%) in the immediate postoperative period. D-wave monitoring was associated with reduced mortality and was more accurate than mMEP monitoring alone or combination monitoring for detecting new postoperative neurological deficits. Further prospective studies are needed to validate these results.

Accurate preoperative assessment of bone invasion is crucial in oral squamous cell carcinoma (OSCC), because it directly influences staging and surgical planning. Computed tomography (CT) is widely used, but its diagnostic performance may vary with tumor localization and image quality. In this retrospective study, patients with OSCC who underwent preoperative CT imaging and subsequent surgical resection with histopathological evaluation were identified. Bone invasion was assessed on CT using a graded and dichotomous classification. Histopathology served as reference standard. Diagnostic performance was analyzed overall and stratified by tumor localization, radiological severity, and image quality. 572 patients were included. Histologically confirmed bone invasion was present in 134 cases (23.6%). Overall, CT demonstrated a sensitivity of 63.4% and a specificity of 90.8%, with an overall diagnostic accuracy of 84.3%. The probability of bone invasion increased stepwise with increasing radiological severity (p < 0.001). Diagnostic performance varied by tumor localization, with significant differences in specificity and overall accuracy (p < 0.001), while sensitivity did not differ significantly (p = 0.597). Radiological grading correlated with pathological stage; nevertheless, 12.3% of pT2 tumors were interpreted as showing bone involvement, while 36.6% of histologically confirmed pT4 tumors remained radiologically occult. Exclusion of cases with relevant imaging artifacts resulted in improved diagnostic performance (AUC 0.80 vs. 0.78). Preoperative CT provides clinically relevant information for assessing bone invasion in OSCC. However, diagnostic performance varies across anatomical subsites, as well as radiological severity and image quality. A localization-aware interpretation may help to avoid under- and overtreatment.

Insulinoma is a rare pancreatic neuroendocrine tumour, usually benign and responsible for organic hypoglycaemia. This retrospective study, conducted at the endocrinology department of Mohammed VI University Hospital between 2017 and 2025, reports 13 cases of insulinoma. The average age at diagnosis was 46 years, with a predominance of females. The clinical manifestations were dominated by sweating (69%) and loss of consciousness (61%), with an average diagnostic delay of 28 months. Laboratory tests confirmed hyperinsulinemic hypoglycaemia, and pancreatic MRI allowed tumour localisation in 90% of cases. The majority of lesions were single, small (median 15 mm) and located in the tail of the pancreas. Surgical enucleation was the most commonly performed procedure (91%), with generally uncomplicated postoperative outcomes. Histological analysis confirmed well-differentiated neuroendocrine tumours, classified as G1 or G2 according to the Ki-67 index. The diagnosis of insulinoma is based on the combination of documented hypoglycaemia and high-performance imaging, particularly MRI. Conservative surgery remains the standard treatment. Genetic testing should be considered if MEN1 is suspected

Phaeochromocytomas (PCCs) and paragangliomas (PGLs), are rare neuroendocrine tumours that arise in the neural crest (NC)-derived adrenal medulla and the paraganglia, respectively. Approximately 10%-15% of patients with PCCs and 35%-40% with PGLs go on to develop metastatic disease, leading to a reported median overall survival of 7 years. The development of prognostic markers and subsequent personal therapeutic strategies are hindered by a lack of understanding of tumourigenesis. In other organs, cells with stem-like properties are at the root of tumour initiation and maintenance, due to their ability to self-renew and give rise to differentiated cells. We have recently shown that, in the human adrenal, a subset of sustentacular cells, endowed with a support role, are in fact SOX2+ postnatal adrenomedullary stem cells, that are specified along the neural crest migratory route. In this study, we intended to determine if SOX2+ cells in PCCs and PGLs can behave as tumour-initiating stem cells. Using expression and transcriptomic studies, we demonstrate the presence of SOX2/SOX2-expressing cells across a broad range of PCCs and PGLs, irrespective of tumour aggressiveness, location, and causative mutation. In silico analyses reveal the co-expression of SOX2 and chromaffin cell markers in the tumour, and the active proliferation of these double-positive cells. Isolation of these cells in vitro in stem cell-promoting media, and their xenotransplantation on chicken chorioallantoic membranes, demonstrates that they have the potential to expand and metastasise in ovo, supporting their potential as tumour-initiating cells.

Gliomas are among the most severe types of brain tumors and can be life-threatening without early detection. Accurate and timely segmentation of brain tumors from MRI scans is crucial for effective treatment planning; however, it remains challenging due to significant variation in tumor shape, size, and location. This paper proposes a 2D Patch-wise Deep Residual U-Net with 102 convolutional layers for automatic tumor segmentation. The approach divides MRI scans into uniform, non-overlapping patches to achieve precise localization and better preserve local features. Residual blocks with identity mapping help mitigate vanishing gradient issues, while dropout layers reduce overfitting during training. T1, T2, and FLAIR modalities from the BraTS 2019 and 2020 datasets were used to evaluate the model. Experimental results show high segmentation accuracy on BraTS 2020 and the Dice Similarity Coefficients (DSC) achieved were 0.9136 (WT), 0.7143 (TC), and 0.7028 (ET). The paper demonstrates that patch-wise deep residual architectures, even with limited training data, can deliver reliable and robust brain tumor segmentation.

Nanoparticle-based photothermal therapy (PTT) provides localized tumor ablation but remains limited by off-target accumulation and the need for high systemic doses. To address these challenges, we developed gold nanorods (AuNRs) coated with a multivalent glucose ligand (mvGlu-AuNR) that engages glucose transporter type 1 (GLUT1) for selective tumor delivery. This design leverages the Warburg effect, using GLUT1 as a metabolic Trojan horse to enter glycolytic cancer cells. In 4T1 breast cancer models, mvGlu-AuNR showed an 8-fold increase in gold content and a 3-fold rise in photoacoustic signal compared to nontargeted controls. Notably, mvGlu-AuNRs converted light to heat more efficiently than mPEG-AuNRs under identical irradiation conditions. ICP-MS analysis confirmed tumor-to-liver ratios ranging from 1.56 to 4.88, which is consistent with strong tumor localization and minimal hepatic uptake. At a systemic dose of 1 mg/kg, mvGlu-AuNRs enabled efficient tumor heating and slowed tumor growth without signs of off-target toxicity. These findings establish metabolic targeting as an effective strategy to enhance PTT specificity, reduce off-target exposure, and enable markedly lower gold dosing.

To evaluate the role of total neoadjuvant therapy (TNT) in locally advanced rectal cancer, focusing on its impact on compliance, tumor downstaging, and organ preservation when combined with watch-and-wait (W&W) protocols and to assess whether radiotherapy dose escalation may increase clinical complete response (cCR) rates. We retrospectively analyzed 135 patients with locally advanced rectal cancer treated with TNT and radiotherapy dose escalation (57.5 Gy to the tumor and nodes in 23fractions) between 2015 and 2024. Patients achieving cCR were considered for the W&W strategy. Outcomes included tumor response, local control, disease-free survival (DFS), and toxicity. Prognostic factors were explored through univariate and multivariate analyses. Atotal of 48.9% achieved cCR and were managed nonoperatively, while 51.1% underwent surgery. Local control at 3years was 83.5%, with local regrowth significantly higher in the W&W group (33.1%) than the rate of local recurrences in the surgical cohort (6.5%, p = 0.001). Poorer local control was associated with low rectal tumors, local regrowth, and the development of metachronous metastases. Disease-free survival at 3years was 69.2%, with better outcomes in patients who underwent surgery or were included in the W&W strategy based on multidisciplinary consensus. No acute or chronic radiotherapy-related toxicity greater than grade3 was observed. In multivariate analysis, local regrowth and nonconsensual W&W inclusion remained independent predictors of impaired DFS and metastasis-free survival. Total neoadjuvant therapy with radiotherapy dose escalation is effective and well tolerated. Tumor location, local regrowth, and the role of multidisciplinary decision-making are critical prognostic factors. Careful patient selection and structured follow-up are essential to optimize outcomes in the W&W strategy. Although functional and quality of life outcomes were not assessed, future prospective studies should incorporate these measures to further strengthen organ-preservation approaches.

Colorectal cancer (CRC) exhibits considerable molecular heterogeneity. This study aimed to delineate the mutational landscape and investigate the associations between frequently mutated genes and key clinicopathological features in a single-center cohort of CRC patients using next-generation sequencing (NGS). This study included 381 patients with pathologically confirmed colorectal cancer. Tumor tissue samples were collected and subjected to targeted sequencing and variant analysis of 40 cancer-related genes using an NGS platform. Associations between gene mutation status and clinicopathological features-including tumor location, clinical stage, and MSI status-were assessed using the χ2 test. Sequencing analysis revealed that 12 patients harbored no detectable mutations in the targeted genes. Among the remaining 369 patients, somatic variants were identified across 30 genes. Regarding mutational patterns, 115 cases (30.2%) exhibited single-gene mutations, 158 cases (41.5%) showed two co-occurring mutations, and 96 cases (25.2%) carried alterations in three or more genes. The most frequently mutated genes were TP53 (76.9%), KRAS (47.8%), and PIK3CA (18.9%). TP53 mutations were significantly enriched in left-sided colon cancers (p < 0.0001). In contrast, both KRAS (p = 0.010) and PIK3CA (p = 0.001) mutations were significantly associated with right-sided colon cancers. Furthermore, the frequency of PIK3CA mutations was significantly higher in MSI-high tumors compared to MSS tumors. This study demonstrates significant associations between specific gene mutations and distinct clinicopathological characteristics. The findings underscore the importance of integrating molecular profiling with conventional clinicopathological parameters for precise stratification.

ABSTRACT Early detection of breast abnormalities remains challenging: manual palpation is subjective and operator‐dependent, while imaging modalities may miss small or subtle stiffness anomalies. This paper presents a biomimetic multifinger robotic palpation approach intended to support early breast‐cancer screening and follow‐up assessment as a proof‐of‐concept. The system integrates tactile arrays with a contact‐regulation scheme under standardized protocols. Each fingertip produces spatial tactile heatmaps, and a human‐like multifinger pressing strategy is used to elicit finger‐wise normal‐interaction responses under controlled contact conditions. The feedback variable is a resultant tactile signal obtained by aggregating taxel readings and is treated as a proxy of normal interaction rather than an absolute force measurement. A real‐time Kalman filter is employed to improve signal fidelity during dynamic contact. The platform is validated on breast‐inspired silicone phantoms with embedded rigid inclusions at varying depths and orientations. Across the tested scenarios, the system achieves repeatable real‐time localization, with stiffness‐weighted centroid errors within 10 mm of the nominal inclusion coordinates and a low incidence of spurious detections under the standardized protocol. We clarify that this study is a proof‐of‐concept focusing on stiffness anomaly localization under controlled phantom conditions rather than clinical diagnosis or benign/malignant classification.