Objective This study aims to explore the application value of spectral CT and perfusion CT parameters in the pathological classification and prognostic assessment of lung cancer.Methods A total of 94 lung cancer patients confirmed by pathology at Shanghai Chest Hospital from January 2023 to November 2024 were included in the study,including 49 cases of lung adenocarcinoma(LUAD),30 cases of lung squamous cell carci-noma(LUSC),and 15 cases of small cell lung cancer(SCLC).All patients underwent spectral CT combined with perfusion scanning using a 256-slice Revolution Apex from GE.Two radiologists independently measured the spectral and perfusion parameters of the three groups of images,including spectral curve slope(K),iodine concentration in the lesion area(ICL),effective atomic number(Zeff),surface permeability(PS),and perfusion index(PI),and established a lung cancer pathological subtype discrimination prediction model based on spectral CT radiomics features.All subjects were randomly divided into a training group and a validation group at a ratio of 3∶1.The discrimination efficacy of the spectral discrimination model between different pathological subtypes and the discrimination efficacy of arterial and venous phase images were compared in multiple dimensions.The performance of the model was evaluated using the receiver operating characteristic(ROC)curve.Results Statistical analysis showed that the spectral curve slope,ICL,NIC,and Zeff of LUAD patients were significantly higher than those of LUSC and SCLC patients(P<0.05),while there were no significant differences in these parameters between LUSC and SCLC patients(P>0.05).Among the perfusion CT parameters,surface permeability(PS)showed significant differences among the three groups(P<0.05),while blood volume(BV),blood flow(BF),perfusion index(PI),time to peak(TTP),and mean transit time(MTT)did not show statistical differences.The multi-factor logistic regression model based on spectral parameters showed strong discriminatory performance:the area under the curve(AUC)of the LUAD and LUSC discrimination model was 0.806/0.77(training group/test group)in the arterial phase and 0.867/0.9(training group/test group)in the venous phase;the AUC of the LUAD and SCLC discrimination model was 0.885/0.883(training group/test group)in the arterial phase and 0.851/0.776(training group/test group)in the venous phase.Conclusion This study indicates that the multi-dimensional functional metabolic analysis indicators of spectral and perfusion CT imaging have significant value in the differential diagnosis of lung cancer pathological subtypes.The diagnostic model constructed by combining multiple spectral parameters can significantly improve the discrimination efficacy of lung adenocarcinoma,squamous cell carcinoma,and small cell lung cancer,providing precise imaging evidence for the formulation of individualized treatment plans.

Purpose To explore the feasibility and clinical value of deep learning-based image reconstruction technology in 5.0T MRI for nasopharyngeal carcinoma.Materials and Methods A prospective study was conducted on 50 newly diagnosed nasopharyngeal carcinoma patients from August to December 2024 at Sun Yat-sen University Cancer Center.5.0T MRI was performed to scan the nasopharynx region.Routine scanning protocols included transverse T2WI,transverse T1WI,transverse contrast-enhanced T1WI and coronal fat-suppressed contrast-enhanced T1WI sequences.Based on these standard scanning protocols,DeepRecon deep learning reconstruction technology with different levels(grade 1-5)was applied,generating a total of 24 sets of images.Qualitative evaluation employed a Likert scale(5-point system)for subjective scoring on lesion detection,lesion edge clarity,artifacts and overall image quality.Quantitative evaluation was performed using the signal-to-noise ratio and contrast-to-noise ratio to objectively assess the quality of the 24 image sets.Differences in qualitative and quantitative indicators between different groups were compared,while the Kappa coefficient was used to analyze the consistency of subjective evaluations by two radiologists.Results In the qualitative assessment of 24 image sets from four MRI sequences(with and without DeepRecon reconstruction),DeepRecon images(grade 2-4)significantly outperformed traditional images in all features except for artifact reduction(Z=-12.11--6.23,all P<0.001).Images reconstructed at DeepRecon grade 3 had the highest overall score and the best image quality.Furthermore,compared with traditional images,DeepRecon images(grade 2-5)demonstrated significantly improved signal-to-noise ratio for both lesions and the lateral pterygoid muscle(t=-15.67--3.44,Z=-6.09--4.63,all P<0.01).In addition,in the transverse T2WI,transverse contrast-enhanced T1WI and coronal fat-suppressed contrast-enhanced T1WI images with DeepRecon reconstruction(grade 2-5),the contrast-to-noise ratio(lesion/lateral pterygoid muscle)also showed significant improvement compared to traditional images(t=-12.71--3.19,Z=-6.08--4.47,all P<0.001).The inter-observer agreement for the overall subjective quality score between the two radiologists was good(Kappa=0.75-0.82,all P<0.01).Conclusion DeepRecon deep learning reconstruction technology significantly increases the signal-to-noise ratio and resolution of traditional magnetic resonance images of nasopharyngeal cancer,improving image clarity and bringing more possibilities for the advancement of imaging diagnosis.

Purpose To explore the feasibility and clinical value of deep learning-based image reconstruction technology in 5.0T MRI for nasopharyngeal carcinoma.Materials and Methods A prospective study was conducted on 50 newly diagnosed nasopharyngeal carcinoma patients from August to December 2024 at Sun Yat-sen University Cancer Center.5.0T MRI was performed to scan the nasopharynx region.Routine scanning protocols included transverse T2WI,transverse T1WI,transverse contrast-enhanced T1WI and coronal fat-suppressed contrast-enhanced T1WI sequences.Based on these standard scanning protocols,DeepRecon deep learning reconstruction technology with different levels(grade 1-5)was applied,generating a total of 24 sets of images.Qualitative evaluation employed a Likert scale(5-point system)for subjective scoring on lesion detection,lesion edge clarity,artifacts and overall image quality.Quantitative evaluation was performed using the signal-to-noise ratio and contrast-to-noise ratio to objectively assess the quality of the 24 image sets.Differences in qualitative and quantitative indicators between different groups were compared,while the Kappa coefficient was used to analyze the consistency of subjective evaluations by two radiologists.Results In the qualitative assessment of 24 image sets from four MRI sequences(with and without DeepRecon reconstruction),DeepRecon images(grade 2-4)significantly outperformed traditional images in all features except for artifact reduction(Z=-12.11--6.23,all P<0.001).Images reconstructed at DeepRecon grade 3 had the highest overall score and the best image quality.Furthermore,compared with traditional images,DeepRecon images(grade 2-5)demonstrated significantly improved signal-to-noise ratio for both lesions and the lateral pterygoid muscle(t=-15.67--3.44,Z=-6.09--4.63,all P<0.01).In addition,in the transverse T2WI,transverse contrast-enhanced T1WI and coronal fat-suppressed contrast-enhanced T1WI images with DeepRecon reconstruction(grade 2-5),the contrast-to-noise ratio(lesion/lateral pterygoid muscle)also showed significant improvement compared to traditional images(t=-12.71--3.19,Z=-6.08--4.47,all P<0.001).The inter-observer agreement for the overall subjective quality score between the two radiologists was good(Kappa=0.75-0.82,all P<0.01).Conclusion DeepRecon deep learning reconstruction technology significantly increases the signal-to-noise ratio and resolution of traditional magnetic resonance images of nasopharyngeal cancer,improving image clarity and bringing more possibilities for the advancement of imaging diagnosis.

Objective This study aims to explore the application value of spectral CT and perfusion CT parameters in the pathological classification and prognostic assessment of lung cancer.Methods A total of 94 lung cancer patients confirmed by pathology at Shanghai Chest Hospital from January 2023 to November 2024 were included in the study,including 49 cases of lung adenocarcinoma(LUAD),30 cases of lung squamous cell carci-noma(LUSC),and 15 cases of small cell lung cancer(SCLC).All patients underwent spectral CT combined with perfusion scanning using a 256-slice Revolution Apex from GE.Two radiologists independently measured the spectral and perfusion parameters of the three groups of images,including spectral curve slope(K),iodine concentration in the lesion area(ICL),effective atomic number(Zeff),surface permeability(PS),and perfusion index(PI),and established a lung cancer pathological subtype discrimination prediction model based on spectral CT radiomics features.All subjects were randomly divided into a training group and a validation group at a ratio of 3∶1.The discrimination efficacy of the spectral discrimination model between different pathological subtypes and the discrimination efficacy of arterial and venous phase images were compared in multiple dimensions.The performance of the model was evaluated using the receiver operating characteristic(ROC)curve.Results Statistical analysis showed that the spectral curve slope,ICL,NIC,and Zeff of LUAD patients were significantly higher than those of LUSC and SCLC patients(P<0.05),while there were no significant differences in these parameters between LUSC and SCLC patients(P>0.05).Among the perfusion CT parameters,surface permeability(PS)showed significant differences among the three groups(P<0.05),while blood volume(BV),blood flow(BF),perfusion index(PI),time to peak(TTP),and mean transit time(MTT)did not show statistical differences.The multi-factor logistic regression model based on spectral parameters showed strong discriminatory performance:the area under the curve(AUC)of the LUAD and LUSC discrimination model was 0.806/0.77(training group/test group)in the arterial phase and 0.867/0.9(training group/test group)in the venous phase;the AUC of the LUAD and SCLC discrimination model was 0.885/0.883(training group/test group)in the arterial phase and 0.851/0.776(training group/test group)in the venous phase.Conclusion This study indicates that the multi-dimensional functional metabolic analysis indicators of spectral and perfusion CT imaging have significant value in the differential diagnosis of lung cancer pathological subtypes.The diagnostic model constructed by combining multiple spectral parameters can significantly improve the discrimination efficacy of lung adenocarcinoma,squamous cell carcinoma,and small cell lung cancer,providing precise imaging evidence for the formulation of individualized treatment plans.

Objective To observe the value of deep learning(DL)models for automatic classification of echocardiographic views.Methods Totally 100 patients after heart transplantation were retrospectively enrolled and divided into training set,validation set and test set at a ratio of 7∶2∶1.ResNet18,ResNet34,Swin Transformer and Swin Transformer V2 models were established based on 2D apical two chamber view,2D apical three chamber view,2D apical four chamber view,2D subcostal view,parasternal long-axis view of left ventricle,short-axis view of great arteries,short-axis view of apex of left ventricle,short-axis view of papillary muscle of left ventricle,short-axis view of mitral valve of left ventricle,also 3D and CDFI views of echocardiography.The accuracy,precision,recall,F1 score and confusion matrix were used to evaluate the performance of each model for automatically classifying echocardiographic views.The interactive interface was designed based on Qt Designer software and deployed on the desktop.Results The performance of models for automatically classifying echocardiographic views in test set were all good,with relatively poor performance for 2D short-axis view of left ventricle and superior performance for 3D and CDFI views.Swin Transformer V2 was the optimal model for automatically classifying echocardiographic views,with high accuracy,precision,recall and F1 score was 92.56％,89.01％,89.97％and 89.31％,respectively,which also had the highest diagonal value in confusion matrix and showed the best classification effect on various views in t-SNE figure.Conclusion DL model had good performance for automatically classifying echocardiographic views,especially Swin Transformer V2 model had the best performance.Using interactive classification interface could improve the interpretability of prediction results to some extent.

Objective:To compare the clinical efficacies between arthroscopic anatomic single bundle reconstruction of posterior cruciate ligament (PCL) assisted by internal tension relieving technique combined with rapid rehabilitation and anatomic single bundle reconstruction combined with conventional rehabilitation in the treatment of PCL rupture.Methods:A retrospective cohort study was used to analyze the clinical data of 88 patients with PCL rupture admitted to First Affiliated Hospital of Kunming Medical University from September 2016 to September 2020. The patients included 65 males and 23 females, aged 18-55 years [(39.3±10.8)years]. Forty-four patients underwent arthroscopic anatomic single bundle reconstruction of PCL assisted by internal tension relieving technique combined with rapid rehabilitation therapy (tension-relieving group), and 44 patients underwent arthroscopic routine anatomic single bundle reconstruction of PCL combined with conventional rehabilitation (traditional group). The two groups were compared before and at 3, 12, and 24 months after surgery regarding the following items: International Knee Literature Committee (IKDC) score, Hospital for Special Surgery (HSS) score, Lysholm score, knee motion cycle (maximum stride length, minimum stride length, and stride frequency) and 6 kinematic indicators (angle of forward and back extension, angle of internal and external rotation, angle of internal and external rotation, up and down displacement, internal and external displacement, and forward and back displacement). The Marburger arthroscopy score (MAS) and Professor Ao Yingfang′s improved score by secondary arthroscopy were compared between the two groups at 12 months after surgery. The perioperative complications were observed.Results:All patients were followed for 24-36 months [(25.5±6.3)months]. In tension-relieving group and the traditional group, the values of IKDC score were (71.8±9.8)points and (68.5±6.5)points at 3 months after surgery, (87.6±6.0)points and (87.6±5.5)points at 12 months after surgery, and (95.5±3.1)points and (92.8±11.6)points at 24 months after surgery, respectively. The values were gradually increased, significantly higher than those before surgery [(48.1±16.9)points and (47.1±15.0)points] (all P<0.05). There were no significant differences between the two groups at each time point (all P>0.05). In tension-relieving group and the traditional group, the values of HSS score were (74.2±6.2)points and (68.4±9.5)points at 3 months after surgery, (91.9±5.4)points and (88.4±4.7)points at 12 months after surgery, and (97.1±2.0)points and (96.2±2.8)points at 24 months after surgery, respectively. The values of HSS score gradually increased, significantly higher than those before surgery [(57.5±17.7)points and (56.8±14.3)points] (all P<0.05). At 3 and 12 months after surgery, the values of HSS score in the tension-relieving group were significantly higher than those in the traditional group (all P<0.05), but others were not significantly different between the two groups (all P>0.05). In tension-relieving group and the traditional group, the values of Lysholm score were (74.2±14.9)points and (70.3±7.5)points at 3 months after surgery, (90.9±6.1)points and (88.7±4.7)points at 12 months after surgery, and (96.9±3.0)points and (96.3±2.8)points at 24 months after surgery, respectively. The values of Lysholm score were gradually increased, significantly higher than those before operation [(48.7±20.7)points and (48.2±19.9)points] (all P<0.05). There were no significant differences between the two groups at any time points (all P>0.05). At 3, 12, and 24 months after surgery, the motion cycle (maximum stride length, minimum stride length and stride frequency) and 6 kinematic indicators (angle of forward bending and backward extension, angle of internal and external rotation, angle of internal and external rotation, internal and external displacement, up and down displacement, and forward and backward displacement) of knee joint were significantly improved in both groups compared with those before surgery (all P<0.05). At 3, 12, and 24 months after surgery, the forward and backward displacement in the tension-relieving group was significantly decreased than that in the traditional group (all P<0.05), but others were not significantly different between the two groups (all P>0.05). The MAS was rated as excellent to good in 14 patients and fair to poor in 2, with the excellent and good rate of 87.5% (14/16) in the tension-relieving group, while the score was rated as excellent to good in 11 patients and fair to poor in 3, with the excellent and good rate of 78.6% (11/14) in the traditional group ( P>0.05). The Professor Ao Yingfang′s improved score was (10.6±1.5)points in the tension-relieving group, markedly higher than that in the traditional group [(9.6±2.3)points] ( P<0.05). No perioperative complications were observed. Conclusion:Compared with anatomic single bundle reconstruction combined with conventional rehabilitation, arthroscopic anatomic single bundle reconstruction of PCL assisted by internal tension relieving technique combined with rapid rehabilitation have higher subjective function score, better anteroposteric stability during knee movement, and better results of secondary microscopy.

Objective:To investigate short-term safety, efficacy and the learning curve of this self-developed novel transcatheter valve repair system (Neonova?) in patients with mitral regurgitation, and explore the role of perioperative echocardiography.Methods:Ten patients who visited the Union Hospital of Tongji Medical College, Huazhong University of Science and Technology from June 2021 to March 2022 and met the inclusive criteria were prospectively enrolled. All the patients were at high risk of surgery with moderate to severe or severe mitral regurgitation (MR). Clamps of Neonova? were implanted under guidance of transesophageal echocardiography and digital subtraction angiography. Clinical outcomes, echocardiography indexes and learning curves of this technique were evaluated immediately after intervention, 7 d, 1 month and 3 months post-intervention.Results:The technical success rate was 100% with MR relieved in all patients immediately after intervention. The device and procedural success rates were both 90.0% with 1 patient received surgical replacement at 37 days post-intervention while the others′ reduced to mild (8/9) and moderate (1/9) MR. New York Heart Association class and the Kansas City Cardiomyopathy Questionnaire improved significantly (all P<0.001). Mean mitral valve pressure gradient didn′t increase significantly after intervention when compared with that before intervention( P=0.324), and no mitral stenosis was observed. Left ventricular end-diastolic diameter decreased significantly ( P=0.008) during follow up.Procedure duration ranged from 60 to 300 (175.8±75.2)minutes. The simple linear regression model between procedure volume and duration showed that procedure duration decreased significantly with the increase of procedure volume ( F=15.857, P=0.004). Conclusions:Neonova? implantation can improve MR severity and clinical symptoms safely and effectively. Transthoracic echocardiography and transesophageal echocardiography are essential for perioperative management of transcatheter mitral valve repair.

Objective:To evaluate the efficacy of the domestic D-Shant device for the treatment of patients with chronic heart failure (CHF) using echocardiography.Methods:Twenty-four CHF patients who were treated with domestic D-Shant device in Union Hospital, Tongji Medical College, Huazhong University of Science and Technology from September 2020 to December 2021 were enrolled in the study. Pulmonary capillary wedge pressure (PCWP)/ left atrial pressure (LAP), right atrial pressure (RAP), pulmonary artery pressure, interatrial septal gradient pressure, cardiac index and pulmonary/systemic blood flow ratio (Qp/Qs) were measured before and after implantation using right heart catheterization.Left atrial end-diastolic area index (LAEDAI), left atrial end-diastolic volume index (LAEDVI), left ventricular end-diastolic volume index (LVEDVI), left ventricular end-systolic volume index (LVESVI), left ventricular ejection fraction (LVEF), right atrial end-diastolic diameter, right ventricular end-diastolic diameter, tricuspid annular plane systolic excursion (TAPSE), right ventricular tractional area change (RVFAC), device shunt aperture, velocity and pressure, together with mitral and tricuspid regurgitation severity were measured using echocardiography before, and 1 month as well as 3 months after D-Shant device implantation. Clinical data were collected and analyzed including 6-minute walking test (6MWT), New York Heart Association (NYHA) classification and the Kansas City Cardiomyopathy Questionnaire (KCCQ). Spearman correlation analysis was used to determine the relation between the changes in PCWP/LAP as well as echocardiographic parameters before and 3 months after implantation and NYHA classification. Binary Logistic regression analysis was performed to determine the predictive factors of NYHA classification improvement at 3-month follow-up after D-Shant device implantation.Results:①D-Shant devices were successfully implanted in all patients. ②Compared with preoperative values, invasive PCWP/LAP systolic, diastolic and mean pressures, transatrial septal gradient, and pulmonary systolic, diastolic and mean pressures decreased significantly after implantation(all P<0.001); Qp/Qs increased significantly after implantation( P<0.001). ③Compared with preoperative values, TAPSE, RVFAC and pulmonary artery flow velocity increased at 1 month after implantation(all P<0.05), whereas a significant reduction in mitral regurgitation grade, and an increase in LVEF and pulmonary artery flow velocity at 3 months after implantation(all P<0.05). Right atrial end-diastolic diameter, right ventricular end-diastolic diameter, LAEDAI, LAEDVI, LVEDVI, LVESVI, ratio of early to late diastolic peak velocities of mitral inflow(E/A), systolic peak velocity of mitral annulus at septal site(S′), ratio of early diastolic peak velocity of mitral inflow to diastolic peak velocity of mitral annulus(E/e′), pulmonary artery diameter, inferior vena cava diameter and degree of tricuspid regurgitation did not change among before, and 1 month as well as 3 months after implantation. There were no significant changes in the device shunt aperture, velocity and pressure between 1 month and 3 months after implantation(all P>0.05). ④The significant improvements in NYHA classification, KCCQ scores and 6MWT were observed at 1 and 3 months after implantation compared with preoperative values (all P<0.01). ⑤NYHA classification at 3 months after implantation was correlated with LVEF pre-post, PCWP/LAP pre-post, TAPSE pre-post and RVFAC pre-post ( rs=0.738, -0.730, 0.738, 0.723; all P<0.001). Logistic regression analysis showed that LVEF pre-post was an independent predictor for NYHA classification improvement at 3 months after implantation ( OR=0.687, 95% CI=0.475-0.992, P=0.045) . Conclusions:Domestic D-Shant device can effectively improve the cardiac function and clinical symptoms in patients with CHF. Echocardiography is a feasible and effective method to evaluate the benefits of domestic D-Shant device for the treatment of CHF.

Objective:To analyze and compare clinical characteristics and risk factors of patients with uremic encephalopathy (UE).Methods:A retrospective analysis was performed from January 2014 to January 2019 in our hospital. Seventy patients diagnosed with chronic kidney disease (CKD) at the end stage (according to diagnosis standard of CKD) complicated with UE were classified into the UE group. In principle matching with sex, age and duration of disease, seventy patients with chronic kidney disease at the end stage but without UE were classified into the non-UE group (NUE group). The demographic data, laboratory examination, CT or MRI examination were recorded and analyzed by using t or χ 2 test. In addition, independent risk factors of patients with UE were analyzed by using Logistic model. Results:A total of 70 patients in the UE group and 70 patients in the NUE group were collected. The proportion of patients with a history of alcohol comsumption, chronic obstructive pulmonary disease and polycystic kidney disease were higher in the UE group than in the NUE group ( P<0.05). There were no significant differences in hypertension, diabetes, and coronary artery disease history between the two groups ( P>0.05). The proportion of cerebral focus and lesions for brain white matter revealed by head CT or MRI in the UE group were significantly higher than that in the NUE group ( P<0.05). The serum NLR and UA levels in the UE group were higher than those in the NUE group ( P<0.05), but the serum ALB and FT3 levels in the UE group were lower than those in the NUE group ( P<0.05). Logistic regression analysis showed that serum NLR, ALB and FT3 levels were independent risk factors for UE patients. Conclusions:Cerebral focus and lesions for brain white matter revealed by CT or MRI are typical abnormal in UE patients. The serum NLR, FT3 and ALB levels are independent risk factor for UE patients.

Objective:To explore the accuracy and clinical effect of computer navigation system in locating the tunnel position in anterior cruciate ligament (ACL) reconstruction.Methods:A retrospective case-control study was conducted to analyze clinical data of 60 patients with ACL rupture treated in Daping Hospital of Army Military Medical University from March 2017 to March 2018, including 44 males and 16 females, aged 15-48 years (mean, 26.3 years). Computer navigation system was used to locate the internal orifice of intra-articular femoral and tibial tunnel in 30 patients (navigation group), and artificial location of intra-articular femoral and tibial tunnel in 30 patients (control group). The operation time and complications of two groups were recorded after operation. The tunnel position was evaluated by CT three-dimensional reconstruction, and the ratio of the tunnel position on the lateral surface of the femoral condyle to the tibial platform was evaluated by the grid method. The proportion sites from the deep posterior edge of lateral intercondylar surface of the femur to the shallow anterior edge (DS) and from the upper edge to the lower edge (HL) were used to represent the position of the femoral tunnel, and the proportion sites of the line connecting the anterior and posterior edge of the tibial plateau were used to represent the tibial tunnel position. The postoperative tunnel position was compared between the two groups. The knee joint stability (Lachman test, pivot shift test) and knee joint function score (Lysholm score, Tegner score) were compared between the two groups before operation and 12 months after operation.Results:All patients were followed up for 12-24 months (mean, 15 months). The operation time was (56.1±8.1)minutes in control group versus (76.0±6.7)minutes in navigation group ( P<0.05). In navigation group, 2 patients had skin edge necrosis at the place where the navigation needle was installed. The proportion of DS in femoral tunnel was (27.2±3.7)% in navigation group versus (33.9±4.4)% in control group ( P<0.05). HL proportion site was (36.6±4.8)% in navigation group versus (38.9±4.9)% in control group ( P<0.05). Proportion of the tibial tunnel at the anterior and posterior edge of the tibial plateau was (44.9±1.8)% in navigation group and (44.7±3.0)% in control group ( P>0.05). Both groups showed significant improvements in Lysholm score, Tegner score, Lachman test and pivot shift test 12 months after operation compared to the preoperative levels ( P<0.05), but the knee joint function score and knee joint stability score had no significant difference between the two groups 12 months after operation ( P>0.05). Conclusions:No significant differences are found between computer navigated and manually navigated ACL reconstruction with regards to knee joint stability, knee joint function, and tibial tunnel localization. However, computer assisted navigation enables the location of the lateral femoral tunnel closer to the anatomical position.