Objective To propose a method for mitigate the impact of anomaly points(such as dust,bubbles,scratches on the chip surface,and minor indentations)in images on the results of digital droplet PCR(ddPCR)detection to achieve high-throughput,stable,and accurate detection.Methods We propose a Filter Faster R-CNN ddPCR detection model,which employs Faster R-CNN to generate droplet prediction boxes followed by removing the anomalies within the positive droplet prediction boxes using an outlier filtering module(Filter).Using a plasmid carrying a norovirus fragment as the template,we established a ddPCR dataset for model training(2462 instances,78.56%)and testing(672 instances,21.44%).Ablation experiments were performed to test the effectiveness of 3 filtering branches of the Filter for anomaly removal on the validation dataset.Comparative experiments with other ddPCR droplet detection models and absolute quantification experiments of ddPCR were conducted to assess the performance of the Filter Faster R-CNN model.Results In low-dust and dusty environments,the Filter Faster R-CNN model achieved detection accuracies of 98.23%and 88.35%for positive droplets,respectively,with composite F1 scores reaching 99.15%and 99.14%,obviously superior to the other models.The introduction of the filtering module significantly enhanced the positive accuracy of the model in dusty environments.In the absolute quantification experiments,a regression line was plotted using the results from commercial flow cytometry equipment as the standard concentration.The results show a regression line slope of 1.0005,an intercept of-0.025,and a determination coefficient of 0.9997,indicating high consistency between the two results.Conclusion The ddPCR detection technique using the Filter Faster R-CNN model provides a robust detection method for ddPCR under various environmental conditions.

Objective To propose a method for mitigate the impact of anomaly points(such as dust,bubbles,scratches on the chip surface,and minor indentations)in images on the results of digital droplet PCR(ddPCR)detection to achieve high-throughput,stable,and accurate detection.Methods We propose a Filter Faster R-CNN ddPCR detection model,which employs Faster R-CNN to generate droplet prediction boxes followed by removing the anomalies within the positive droplet prediction boxes using an outlier filtering module(Filter).Using a plasmid carrying a norovirus fragment as the template,we established a ddPCR dataset for model training(2462 instances,78.56%)and testing(672 instances,21.44%).Ablation experiments were performed to test the effectiveness of 3 filtering branches of the Filter for anomaly removal on the validation dataset.Comparative experiments with other ddPCR droplet detection models and absolute quantification experiments of ddPCR were conducted to assess the performance of the Filter Faster R-CNN model.Results In low-dust and dusty environments,the Filter Faster R-CNN model achieved detection accuracies of 98.23%and 88.35%for positive droplets,respectively,with composite F1 scores reaching 99.15%and 99.14%,obviously superior to the other models.The introduction of the filtering module significantly enhanced the positive accuracy of the model in dusty environments.In the absolute quantification experiments,a regression line was plotted using the results from commercial flow cytometry equipment as the standard concentration.The results show a regression line slope of 1.0005,an intercept of-0.025,and a determination coefficient of 0.9997,indicating high consistency between the two results.Conclusion The ddPCR detection technique using the Filter Faster R-CNN model provides a robust detection method for ddPCR under various environmental conditions.

Objective:To investigate the safety of minimally invasive liver resection for resectable hepatocellular carcinoma (HCC) complicated with portal hypertension.Methods:The propensity score matching and retrospective cohort study was conducted. The clinicopathological data of 807 patients with resectable HCC who underwent minimally invasive liver resection in 8 medical centers, including Sir Run Run Shaw Hospital, Affiliated with the Zhejiang University School of Medicine et al, from June 2011 to November 2022 were collected. There were 670 males and 137 females, aged 58(50,66)years. Of the 807 patients, 173 cases with portal hypertension were divided into the portal hypertension group, and 634 cases without portal hypertension were divided into the non-portal hypertension group. Observation indicators: (1) propensity score matching and comparison of general data of patients between the two groups after matching; (2) intraoperative and post-operative situations; (3) subgroup analysis. Propensity score matching was done by the 1:1 nearest neighbor matching method, with the caliper setting as 0.001. Measurement data with skewed distribution were represented as M( Q1, Q3), and comparison between groups was conducted using the rank sum test. Count data were described as absolute numbers, and comparison between groups was conducted using the chi-square test or Fisher exact probability. Comparison of ordinal data was constructed using the non-parameter rank sun test. Results:(1) Propensity score matching and comparison of general data of patients between the two groups after matching. Of the 807 patients, 268 cases were successfully matched, including 134 cases in the portal hypertension group and 134 cases in the non-portal hypertension group. The elimination of the tumor diameter and robot-assisted surgery confounding bias ensured comparability between the two groups after propensity score matching. (2) Intraoperative and postoperative situations. The occlusion time of porta hepatis, cases with intraoperative blood transfusion, cases with postoperative complication, cases with complication >Ⅱ grade of Clavien-Dindo classification, cases of Clavien-Dindo classification as Ⅰ grade, Ⅱ grade, Ⅲ grade, Ⅳ grade, cases with liver related complication were 27.0(15.0,43.0)minutes, 33, 55, 15, 13, 29, 14, 1, 37 in the portal hypertension group, versus 35.0(22.0,60.0)minutes, 17, 25, 5, 14, 9, 4, 1, 13 in the non-portal hypertension group, showing significant differences in the above indicators between the two groups ( Z=-2.15, χ2=6.30, 16.39, 4.38, 20.72, 14.16, P<0.05). (3) Subgroup analysis. Results of subgroups analysis showed that in cases with major live resection, the operation time, volume of intraoperative blood loss, duration of postoperative hospital stay were 243.5(174.6,296.3)minutes, 200.0(150.0,600.0)mL, 7.5(6.0,13.0)days in the portal hypertension group, versus 270.0(180.0,314.5)minutes, 200.0 (75.0,450.0)mL, 7.0(5.5,10.0)days in the non-portal hypertension group, showing no significant difference in the above indicators between the two groups ( Z=-0.54, -1.73, -0.92, P>0.05). In cases with non-major live resection, the operation time, volume of intraoperative blood loss, duration of postoperative hospital stay were 170.0(120.0,227.5)minutes, 100.0(50.0,200.0)mL, 8.0(5.0,10.0)days in the portal hypertension group, versus 170.0(120.0,227.5)minutes, 100.0(50.0,200.0)mL, 7.0(5.5,9.0)days in the non-portal hypertension group, showing no significant difference in the above indicators between the two groups ( Z=-1.39, -0.10, 1.05, P>0.05). In cases with anatomical liver resection, the operation time, volume of intraoperative blood loss, duration of postoperative hospital stay were 210.0(150.0,285.0)minutes, 150.0(50.0,200.0)mL, 8.0(6.0,9.3)days in the portal hypertension group, versus 225.5(146.3,306.8)minutes, 100.0(50.0,250.0)mL, 7.0(6.0,9.0)days in the non-portal hypertension group, showing no significant difference in the above indica-tors between the two groups ( Z=-0.75, -0.26, -0.91, P>0.05). In cases with non-anatomical liver resection, the operation time, volume of intraoperative blood loss, duration of postoperative hospital stay were 173.5(120.0,231.5)minutes, 175.0(50.0,300.0)mL, 7.0(5.0,11.0)days in the portal hyper-tension group, versus 186.0(123.0,262.5)minutes, 100.0(50.0,200.0)mL, 7.0(5.0,9.5)days in the non-portal hypertension group, showing no significant difference in the above indicators between the two groups ( Z=-0.97, -1.12, -0.98, P>0.05). Conclusion:Minimally invasive liver resection or even major liver resection is safe and feasible for screened HCC patients complicated with portal hyper-tension, but attention should be paid to the prevention and treatment of postoperative complications.