Lung cancer is a leading cause of cancer-related morbidity and mortality worldwide. Coupled with the substantial workload, the clinical management of lung cancer is challenged by the critical need to efficiently and accurately process increasingly complex medical information. In recent years, large language models (LLMs) technology has undergone explosive development, demonstrating unique advantages in handling complex medical data by leveraging its powerful natural language processing capabilities, and its application value in the field of lung cancer diagnosis and treatment is continuously increasing. The paper systematically analyzes that the exceptional potential of LLMs in lung cancer auxiliary diagnosis, tumor feature extraction, automatic staging, progression/outcome analysis, treatment recommendations, medical documentation generation, and patient education. However, they face critical technical and ethical challenges including inconsistent performance in complex integrated decision-making (e.g., TNM staging, personalized treatment suggestions) and "black box" opacity issues, along with dilemmas such as training data biases, model hallucinations, data privacy concerns, and cross-lingual adaptation challenges ("data colonization"). Future directions should prioritize constructing high-quality multimodal corpora specific to lung cancer, developing interpretable and compliant specialized models, and achieving seamless integration with existing clinical workflows. Through dual drivers of technological innovation and ethical standardization, LLMs should be prudently advanced for holistic lung cancer management processes, ultimately promoting efficient, standardized, and personalized diagnosis and treatment practices.

Objective:To develop an artificial intelligence (AI)-guided intravitreal injection robot system to accurately detect the injection point on the ocular surface and guide the robotic arm to complete the intravitreal injection positioning task through 3D position calculation.Methods:The Dikablis subset of the TEyeD dataset was used.Training set, testing set, and validation set were constructed by using equal interval sampling strategy.The system read the ocular surface color RGB image with an RGBD camera, then used a PatchCrop-Transformer-based injection point detection algorithm to detect and locate key points such as the pupil, iris, and eyelid in the image.Next, it extracted the local 3D point cloud data near the injection point based on the depth information obtained by the camera.Through principal component analysis (PCA) of the local area point cloud data, the injection point and injection direction were determined.The key information was then passed to the robotic arm system.The end of the robotic arm adopted a remote center of motion (RCM) mechanism.After solving the forward and inverse kinematics, the joint movement path was obtained, and the robotic arm was controlled to move to 2 cm above the injection point.After confirmation by the doctor, the insertion, injection, and withdrawal operations were completed to ensure the stability and repeatability of the injection process.The mean square error (MSE) of key points localization and the success detection rate (SDR) within different pixel error ranges (2, 5, and 10 pixels) of the study method were compared with those of the NFDP, SLPT, and StarLoss methods, and the effects of random weight enhancement, fixed weight enhancement, and no enhancement methods on the MSE of key points localization were evaluated.The repeatability and absolute positioning accuracy of the robotic arm system were also evaluated.Results:After adding random weight enhancement, the model of this study outperformed the fixed weight enhancement and no enhancement methods in both MSE and SDR.The MSEs of the model proposed in this study for overall eye, pupil, and iris localization were 4.25, 2.41, and 1.54, respectively, which were lower than those of the NFDP, StarLoss, and SLPT methods.Within the error ranges of 5 and 10 pixels, the SDRs of the model proposed in this study were 72.09% and 92.68%, respectively, which were higher than those of the NFDP, StarLoss, and SLPT methods.The single-axis repeatability errors and absolute positioning errors of the robotic arm were within ±5 μm.Conclusions:The AI-guided intravitreal injection robot system integrates RGBD images to achieve automatic recognition of the ocular injection point and high-precision motion control through RCM mechanism design and corresponding kinematic solution methods.

Objective:To investigate the risk factors for delirium after free flap reconstruction surgery in elderly patients with head and neck cancer.Methods:This study retrospectively collected clinical data from 309 elderly patients with head and neck cancer who underwent radical resection and free flap reconstruction surgery in the Department of Operation & Anesthesiology, College & Hospital of Stomatology, Guangxi Medical University from January 2018 to December 2022. Among them, there were 197 males (63.8%) and 112 females (36.2%), with an average age of (66.8±5.5) years old. Based on the delirium diagnostic criteria outlined in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders, patients were evaluated for postoperative delirium within one week after surgery. They were then divided into a delirium group ( n=75) and a non-delirium group ( n=234). Perioperative indicators potentially related to postoperative delirium, including age, gender, past medical history, tumor characteristics, hematological tests, perioperative sleep disturbances, surgical procedures, intraoperative fluid intake and output, anesthetic dosage, postoperative pain, etc, were collected from the relevant medical record system.Variables with P<0.05 were included in a multivariate Logistic regression model to screen for independent risk factors for delirium. Results:The incidence of delirium after free flap reconstruction surgery for head and neck cancer in elderly patients was 24.3% (75/309). Multivariate Logistic regression analysis showed that male gender ( OR=2.802, P=0.005), perioperative sleep disturbances ( OR=7.104, P<0.001), and moderate-to-severe postoperative pain ( OR=6.903, P<0.001) were risk factors for postoperative delirium. Conclusions:Male gender, perioperative sleep disturbances, and moderate-to-severe postoperative pain are independent risk factors for delirium in these patients.

Mental disorders pose a significant challenge to global public health,profoundly affecting the quality of life of a vast number of individuals and imposing a heavy health burden on society.Nonetheless,there remains a substantial gap between the current societal capacity to provide prevention,diagnosis,and treatment for mental disorders and the existing demand for such services.In recent years,the development and application of artificial intelligence(AI)technologies have provided unprecedented opportunities to enhance mental healthcare services.As one of the fastest-growing fields of AI,generative AI has played a pivotal role in analyzing diverse forms of data,including medical image processing,protein structure prediction,clinical document generation,auxiliary diagnostic discrimination,and clinical decision support.These advancements have significantly strengthened capabilities in clinical diagnosis,data reconstruction,and adjunctive therapeutic interventions.This review highlights the potential applications of generative AI in advancing fundamental psychiatric research,identifying early risk factors for mental disorders,and assisting clinicians in diagnosis and treatment.Additionally,it addresses the challenges and limitations currently facing the application of generative AI to mental healthcare,including biases,privacy breaches,and insufficient interpretability.Finally,the review summarizes strategies to enhance AI's capacity to deliver mental health services,aiming to leverage new technologies to reduce the global burden of mental disorders and improve the quality of life of affected individuals.

Intraluminal thrombus(ILT)at the vertebral artery origin is rare and often missed due to atypical symptoms.However,it poses a high risk of artery-to-artery embolism and severe posterior circulation ischemia.This article reports a case of missed diagnosis of ILT to improve the diagnostic ability.The patient,a 49-year-old male was admitted to emergency department with sudden dizziness accompanied by nausea for 12 hours.Head CT suggested posterior circulation ischemia.Patient was given aspirin 0.2 g/d and atorvastatin 20 mg/d orally and discharged from the hospital.Head MRI in the outpatient department showed a recent infarct in the right cerebellar hemisphere next day.On the 28th,dizziness recurred,accompanied by unsteady walking and right-sided hemisensory numbness.The symptoms persisted and did not improve.He was admitted to the hospital on the 29th.Cervical CTA showed severe stenosis at the origin of the right vertebral artery,filling defect in the distal lumen combined with the Donut sign.Thus,ILT was diagnosed.The drug treatment regimen was adjusted to aspirin 100 mg/d combined with clopidogrel 75 mg/d for a total of 15 weeks.High-resolution MRI of the cervical blood vessels in the outpatient department showed occlusion of the V1-V2 segment of the right vertebral artery.On November 2,endovascular treatment was performed under local anesthesia.At 90 days and 6 months of follow-up after the operation,no recurrence of cerebral ischemia events occurred,and the mRS score was 0.This case highlights the need for vigilance toward ILT in posterior circulation ischemia when initial imaging is inconclusive.Early dynamic CTA(5-7 days)combined with dual antiplatelets may prevent thrombus progression and improve outcomes.

Objective:To develop an artificial intelligence (AI)-guided intravitreal injection robot system to accurately detect the injection point on the ocular surface and guide the robotic arm to complete the intravitreal injection positioning task through 3D position calculation.Methods:The Dikablis subset of the TEyeD dataset was used.Training set, testing set, and validation set were constructed by using equal interval sampling strategy.The system read the ocular surface color RGB image with an RGBD camera, then used a PatchCrop-Transformer-based injection point detection algorithm to detect and locate key points such as the pupil, iris, and eyelid in the image.Next, it extracted the local 3D point cloud data near the injection point based on the depth information obtained by the camera.Through principal component analysis (PCA) of the local area point cloud data, the injection point and injection direction were determined.The key information was then passed to the robotic arm system.The end of the robotic arm adopted a remote center of motion (RCM) mechanism.After solving the forward and inverse kinematics, the joint movement path was obtained, and the robotic arm was controlled to move to 2 cm above the injection point.After confirmation by the doctor, the insertion, injection, and withdrawal operations were completed to ensure the stability and repeatability of the injection process.The mean square error (MSE) of key points localization and the success detection rate (SDR) within different pixel error ranges (2, 5, and 10 pixels) of the study method were compared with those of the NFDP, SLPT, and StarLoss methods, and the effects of random weight enhancement, fixed weight enhancement, and no enhancement methods on the MSE of key points localization were evaluated.The repeatability and absolute positioning accuracy of the robotic arm system were also evaluated.Results:After adding random weight enhancement, the model of this study outperformed the fixed weight enhancement and no enhancement methods in both MSE and SDR.The MSEs of the model proposed in this study for overall eye, pupil, and iris localization were 4.25, 2.41, and 1.54, respectively, which were lower than those of the NFDP, StarLoss, and SLPT methods.Within the error ranges of 5 and 10 pixels, the SDRs of the model proposed in this study were 72.09% and 92.68%, respectively, which were higher than those of the NFDP, StarLoss, and SLPT methods.The single-axis repeatability errors and absolute positioning errors of the robotic arm were within ±5 μm.Conclusions:The AI-guided intravitreal injection robot system integrates RGBD images to achieve automatic recognition of the ocular injection point and high-precision motion control through RCM mechanism design and corresponding kinematic solution methods.

Objective:To explore the differences between clinical-pathological-ultrasound features in hepatocellular carcinoma（HCC）with negative and positive des-gamma-carboxy prothrombin（DCP）.Methods:A retrospective analysis was conducted on 649 patients with pathologically confirmed HCC at Zhongshan Hospital，Fudan University from April 2020 to May 2024. Patients were stratified into DCP-negative（177 cases，<40 mAU/ml）and DCP-positive（472 cases，≥40 mAU/ml）groups. Clinical data，pathological features，and ultrasound findings were collected. Conventional ultrasound and contrast-enhanced ultrasound（CEUS）imaging characteristics were analyzed and compared between the two groups，and the correlation between ultrasound features and pathological characteristics were analyzed.Results:The DCP-negative group exhibited a lower incidence of microvascular invasion（10.17% vs. 34.75%， P<0.001）and smaller median tumor diameter（23 mm vs. 40 mm， P<0.001）. Heterogeneous internal echogenicity was less frequent in DCP-negative tumors［48.59%（86/177） vs. 74.58%（352/472）， P<0.001］. CEUS revealed higher rates of arterial-phase iso-enhancement（6.78% vs. 1.69%）and absence of washout（13.56% vs. 4.45%）in DCP-negative HCC（both P<0.001）. CEUS LI-RADS classification showed fewer LR-5 lesions［50.85%（90/177） vs. 59.53%（281/472）］in DCP-negative group（ P<0.001）. Conclusions:HCC with different DCP states has different clinical-pathological-ultrasound features. DCP-negative HCCs are more likely to show atypical enhancement patterns characteristic of HCC.

Amorphous calcium phosphate（ACP）is a component of urinary stones，primarily forming mixed stones with calcium oxalate，while pure ACP stones are relatively rare. This article reports a case of a patient with an ACP bladder stone who was admitted due to progressive dysuria over 5 years，which had worsened in the past months. Upon admission，tPSA was 29.63 ng/ml. CT and enhanced MRI revealed multiple bladder stones and prostatic hyperplasia. The patient underwent ultrasound-guided prostate biopsy and transurethral cystolithotripsy with pneumatic lithotripsy. Postoperative infrared spectroscopy confirmed the stone composition as ACP，and prostatic adenocarcinoma was diagnosed by prostate biopsy pathology. Endocrine therapy was administered postoperatively，and follow-up imaging at 3 months showed no stone. This article presents the first reported case of an ACP bladder stone coexisting with prostate cancer，providing important clinical insights into the etiology of such stones and the rare local manifestations of prostate cancer.

Intraluminal thrombus(ILT)at the vertebral artery origin is rare and often missed due to atypical symptoms.However,it poses a high risk of artery-to-artery embolism and severe posterior circulation ischemia.This article reports a case of missed diagnosis of ILT to improve the diagnostic ability.The patient,a 49-year-old male was admitted to emergency department with sudden dizziness accompanied by nausea for 12 hours.Head CT suggested posterior circulation ischemia.Patient was given aspirin 0.2 g/d and atorvastatin 20 mg/d orally and discharged from the hospital.Head MRI in the outpatient department showed a recent infarct in the right cerebellar hemisphere next day.On the 28th,dizziness recurred,accompanied by unsteady walking and right-sided hemisensory numbness.The symptoms persisted and did not improve.He was admitted to the hospital on the 29th.Cervical CTA showed severe stenosis at the origin of the right vertebral artery,filling defect in the distal lumen combined with the Donut sign.Thus,ILT was diagnosed.The drug treatment regimen was adjusted to aspirin 100 mg/d combined with clopidogrel 75 mg/d for a total of 15 weeks.High-resolution MRI of the cervical blood vessels in the outpatient department showed occlusion of the V1-V2 segment of the right vertebral artery.On November 2,endovascular treatment was performed under local anesthesia.At 90 days and 6 months of follow-up after the operation,no recurrence of cerebral ischemia events occurred,and the mRS score was 0.This case highlights the need for vigilance toward ILT in posterior circulation ischemia when initial imaging is inconclusive.Early dynamic CTA(5-7 days)combined with dual antiplatelets may prevent thrombus progression and improve outcomes.

Mental disorders pose a significant challenge to global public health,profoundly affecting the quality of life of a vast number of individuals and imposing a heavy health burden on society.Nonetheless,there remains a substantial gap between the current societal capacity to provide prevention,diagnosis,and treatment for mental disorders and the existing demand for such services.In recent years,the development and application of artificial intelligence(AI)technologies have provided unprecedented opportunities to enhance mental healthcare services.As one of the fastest-growing fields of AI,generative AI has played a pivotal role in analyzing diverse forms of data,including medical image processing,protein structure prediction,clinical document generation,auxiliary diagnostic discrimination,and clinical decision support.These advancements have significantly strengthened capabilities in clinical diagnosis,data reconstruction,and adjunctive therapeutic interventions.This review highlights the potential applications of generative AI in advancing fundamental psychiatric research,identifying early risk factors for mental disorders,and assisting clinicians in diagnosis and treatment.Additionally,it addresses the challenges and limitations currently facing the application of generative AI to mental healthcare,including biases,privacy breaches,and insufficient interpretability.Finally,the review summarizes strategies to enhance AI's capacity to deliver mental health services,aiming to leverage new technologies to reduce the global burden of mental disorders and improve the quality of life of affected individuals.