Objective:To evaluate the efficacy of preoperative prediction of regional lymph node (RLN) metastasis in renal cell carcinoma (RCC) using a machine learning model based on habitat imaging radiomics from renal MRI.Methods:This cross-sectional study retrospectively analyzed 220 patients with RCC who underwent nephrectomy and RLN dissection at four medical centers of Chinese PLA General Hospital from January 2010 to August 2023. The cohort included 65 patients with RLN metastasis and 155 without. A stratified random sampling method was used to divide 175 patients from the first medical center into a training set ( n=140) and an internal test set ( n=35) in an 8∶2 ratio, while 45 patients from the third, fourth, and fifth medical centers constituted the external test set. The primary RCC lesions were categorized into 15 habitat subregions based on corticomedullary-phase enhancement and T 2WI signal intensity on MRI, and the volume fractions of different subregions were analyzed. In the training cohort, radiomics features derived from the habitat subregions were used to construct a radiomics model employing various machine learning algorithms, including extremely random trees (ET), gradient boosting decision trees (GBDT), random forest (RF), and support vector machine (SVM). The optimal model was selected and combined with RLN short-axis diameter to develop a combined model. The efficacy of each model in predicting RLN metastasis was evaluated using the receiver operating characteristic (ROC) curve. Results:The volume fraction of hyper-enhanced hyper-intense regions in the non-metastatic group was significantly higher than that in the metastatic group (0.05±0.09 vs. 0.02±0.03; t=3.00, P=0.003). Among the machine learning models constructed using 15 optimal habitat radiomics features, the SVM model demonstrated the best performance, with area under the ROC curve (AUC) values of 0.85 (95% CI 0.72-0.98) in the internal test set and 0.82 (95% CI 0.67-0.98) in the external test set, surpassing those of the ET, GBDT, and RF models. The combined model, integrating the SVM model with RLN short-axis diameter, achieved AUC values of 0.94 (95% CI 0.85-1.00) in the internal test set and 0.89 (95% CI 0.78-1.00) in the external test set, with RLN short-axis diameter contributing AUC values of 0.81 (95% CI 0.66-0.96) and 0.81 (95% CI 0.68-0.94), respectively. The diagnostic sensitivity of the combined model was 91.7% in the internal test set and 85.7% in the external test set, with specificities of 78.3% and 67.7%, respectively. Conclusion:The combined model based on MRI habitat imaging radiomics and RLN short-axis diameter demonstrates excellent preoperative assessment capability for RLN metastasis in RCC.

Objective:To evaluate the efficacy of preoperative prediction of regional lymph node (RLN) metastasis in renal cell carcinoma (RCC) using a machine learning model based on habitat imaging radiomics from renal MRI.Methods:This cross-sectional study retrospectively analyzed 220 patients with RCC who underwent nephrectomy and RLN dissection at four medical centers of Chinese PLA General Hospital from January 2010 to August 2023. The cohort included 65 patients with RLN metastasis and 155 without. A stratified random sampling method was used to divide 175 patients from the first medical center into a training set ( n=140) and an internal test set ( n=35) in an 8∶2 ratio, while 45 patients from the third, fourth, and fifth medical centers constituted the external test set. The primary RCC lesions were categorized into 15 habitat subregions based on corticomedullary-phase enhancement and T 2WI signal intensity on MRI, and the volume fractions of different subregions were analyzed. In the training cohort, radiomics features derived from the habitat subregions were used to construct a radiomics model employing various machine learning algorithms, including extremely random trees (ET), gradient boosting decision trees (GBDT), random forest (RF), and support vector machine (SVM). The optimal model was selected and combined with RLN short-axis diameter to develop a combined model. The efficacy of each model in predicting RLN metastasis was evaluated using the receiver operating characteristic (ROC) curve. Results:The volume fraction of hyper-enhanced hyper-intense regions in the non-metastatic group was significantly higher than that in the metastatic group (0.05±0.09 vs. 0.02±0.03; t=3.00, P=0.003). Among the machine learning models constructed using 15 optimal habitat radiomics features, the SVM model demonstrated the best performance, with area under the ROC curve (AUC) values of 0.85 (95% CI 0.72-0.98) in the internal test set and 0.82 (95% CI 0.67-0.98) in the external test set, surpassing those of the ET, GBDT, and RF models. The combined model, integrating the SVM model with RLN short-axis diameter, achieved AUC values of 0.94 (95% CI 0.85-1.00) in the internal test set and 0.89 (95% CI 0.78-1.00) in the external test set, with RLN short-axis diameter contributing AUC values of 0.81 (95% CI 0.66-0.96) and 0.81 (95% CI 0.68-0.94), respectively. The diagnostic sensitivity of the combined model was 91.7% in the internal test set and 85.7% in the external test set, with specificities of 78.3% and 67.7%, respectively. Conclusion:The combined model based on MRI habitat imaging radiomics and RLN short-axis diameter demonstrates excellent preoperative assessment capability for RLN metastasis in RCC.

Imported malaria has become a major risk factor for malaria prevention and control in China. How to screen malaria quickly for people entering China is an urgent problem to be solved. Protein microarrays are widely used in high-throughput screening and diagnosis. In this study, surface plasmon resonance (SPR) technique for malaria detection was established by using the specific adsorption surface treated by polyethylene glycol polymer, and the malaria specific antigen HRP2 was used as capture probe. The optimal concentration of antigen, sensitivity and specificity of detection, as well as anti-interference ability of the chip were analyzed. The SPR protein chip was applied to detect specific antibodies of malignant malaria in serum with the advantage of label-free, instant and fast. Compared with fluorescence quantitative PCR, there were no significant difference in sensitivity and specificity between the two methods. This study lays a foundation for further development of protein microarray for malaria typing identification, and it is conducive to the rapid screening of malaria for people entering.
Antibodies ; China ; Humans ; Malaria/diagnosis* ; Protein Array Analysis ; Surface Plasmon Resonance

Objective:To understand the diagnostic value of tuberculosis (TB) pathogenic detection methods (TPDM) in pathological tissue for TB.Methods:A retrospective study was conducted with 190 pathological specimens from different tissues suspected with TB from Third People′s Hospital of Shenzhen during May 2016 and May 2019. Specimens were divided into four groups according to histomorphology: group one, necrotizing granulomatous inflammation (109 cases); group two, non-necrotic granulomatous inflammation (20 cases); group three, non-granulomatous inflammation (45 cases); group four, non-tuberculous lesions (16 cases). The positive rates of each TPDM among specimens from four groups were compared. The positive rates of all TPDM for specimens from group one were compared. Meanwhile, the influence of antituberculosis treatment course on the TPDM was analyzed. Chi-square test or Fisher′s exact test was used for statistical analysis.Results:The positive rates of Ziehl-Neelsen acid-fast staining among the four groups were 17.4%(19/109), 5.0%(1/20), 4.4%(2/45) and 0(0/16), respectively. The positive rates of Mycobacterium tuberculosis (MTB) complex culture were 32.0%(32/100), 4/19, 4.8%(2/42) and 0(0/16), respectively. The positive rates of Mycobacterium tuberculosis/rifampin resistance real-time quantitative nucleic acid amplification detection system (Xpert MTB/RIF) were 74.3%(81/109), 15.0%(3/20), 13.3%(6/45) and 0(0/16), respectively. The positive rates of fluorescent quantitative polymerase chain reaction (FQ-PCR) were 63.0%(58/92), 0(0/15), 2.6%(1/38) and 0(0/10), respectively. The positive rates of simultaneous amplification and testing (SAT) were 32.4%(24/74), 0(0/10), 0(0/15) and 0(0/10), respectively. The differences of each TPDM among four groups were all statistically significant (all P<0.05). The positive rate of Xpert MTB/RIF in group one specimens was significantly higher than those of acid-fast staining, MTB culture and SAT ( χ2=71.016, 37.162 and 35.679, respectively, all P<0.01), while the difference was not statistically significant when compared with FQ-PCR ( χ2=2.517, P=0.112). The positive rate of combined TPDM (85.3%(93/109)) was significantly higher than Xpert MTB/RIF(74.3%(81/109)) ( χ2=4.100, P=0.043). The positive rates of acid-fast staining group 1A (anti-tuberculosis treatment course was less than one month) and group 1B (anti-tuberculosis treatment course was longer than one month) were 14.3%(7/49) and 20.0% (12/60), respectively ( χ2=0.612, P=0.434); those of MTB culture were 48.9% (22/45) and 18.2% (10/55), respectively ( χ2=10.721, P=0.001); those of Xpert MTB/RIF were 69.4%(34/49) and 78.3%(47/60), respectively ( χ2=1.131, P=0.287); those of FQ-PCR were 55.0%(22/40) and 69.2%(36/52), respectively ( χ2=1.965, P=0.161); those of SAT were 43.3%(13/30) and 25.0%(11/44), respectively ( χ2=2.736, P=0.098). Conclusions:The results of TPDM correlate closely with the typical histomorphological features of tuberculosis. Xpert MTB/RIF possesses significantly higher sensitivity than any other single TPDM, and is not attenuated by early anti-tuberculosis treatment. Combined TPDM could significantly improve the sensitivity of TB pathogenic detection, which is suggested to be applied when the tissue specimen is sufficient.