Objective:To develop an artificial intelligence (AI)-based platelet review strategy to identify abnormal platelet histograms with no significant difference between initial impedance platelet count (PLT-I) and PLT-F results.Methods:This study included 5 119 routine blood analysis in Nanfang Hospital of Southern Medical University and its Ganzhou branch from July 2023 and March 2024. Specimens exhibiting abnormal platelet histograms and an initial platelet count >40×10?/L underwent review using the fluorescent platelet count (PLT-F) channel. Consistency of the results was defined as a difference between impedance platelet count (PLT-I) and PLT-F less than ±20% of the PLT-F results. A deep learning model was developed using platelet and red blood cell histogram data from a training set of 3 807 specimens. The model′s diagnostic performance was evaluated on an independent external validation set ( n=805) using receiver operating characteristic (ROC) curve analysis. Changes in the number of reviewed samples and sample turnaround time were analyzed to assess its clinical utility. Results:The deep learning model based on platelet and red blood cell histograms achieved an area under the ROC curve (AUC) of 0.854 in the training set. At a cutoff value of 0.1, the sensitivity was 0.954 and specificity was 0.358. The model could reduce review by 16.80% (190/1 131). In the validation set, the AUC was 0.805, with a sensitivity of 0.955 and specificity of 0.307, corresponding to a reduction of 17.41% (47/270) in reviewed specimens.Conclusion:The platelet review prediction model developed based on deep learning technology can efficiently identify samples with consistent results before and after review, reducing unnecessary reviews and shortening specimen testing time, thereby improving the efficiency of platelet test.

Objective:To develop an artificial intelligence (AI)-based platelet review strategy to identify abnormal platelet histograms with no significant difference between initial impedance platelet count (PLT-I) and PLT-F results.Methods:This study included 5 119 routine blood analysis in Nanfang Hospital of Southern Medical University and its Ganzhou branch from July 2023 and March 2024. Specimens exhibiting abnormal platelet histograms and an initial platelet count >40×10?/L underwent review using the fluorescent platelet count (PLT-F) channel. Consistency of the results was defined as a difference between impedance platelet count (PLT-I) and PLT-F less than ±20% of the PLT-F results. A deep learning model was developed using platelet and red blood cell histogram data from a training set of 3 807 specimens. The model′s diagnostic performance was evaluated on an independent external validation set ( n=805) using receiver operating characteristic (ROC) curve analysis. Changes in the number of reviewed samples and sample turnaround time were analyzed to assess its clinical utility. Results:The deep learning model based on platelet and red blood cell histograms achieved an area under the ROC curve (AUC) of 0.854 in the training set. At a cutoff value of 0.1, the sensitivity was 0.954 and specificity was 0.358. The model could reduce review by 16.80% (190/1 131). In the validation set, the AUC was 0.805, with a sensitivity of 0.955 and specificity of 0.307, corresponding to a reduction of 17.41% (47/270) in reviewed specimens.Conclusion:The platelet review prediction model developed based on deep learning technology can efficiently identify samples with consistent results before and after review, reducing unnecessary reviews and shortening specimen testing time, thereby improving the efficiency of platelet test.

Objective:To compare the efficacy and safety of robot-assisted laparoscopic and laparoscopic partial nephrectomy (RAPN and LPN) for patients with completely endophytic renal tumor.Methods:A total of 73 patients with completely endophytic renal tumor receiving RAPN (n=29) or LPN (n=44) in our center between January 2015 and June 2021 were retrospectively collected. There were 21 males and 8 females in RAPN group. The average age was 48.6±13.7 years old, average tumor size was 2.9±0.9 cm with 13 left tumors and 16 right tumors, average R. E.N.A.L. score was 9.2±1.0, and average preoperative eGFR was 82.6±10.7 ml/(min·1.73 m 2). There were 27 males and 17 females in LPN group. The average age was 50.1±12.3 years old, average tumor size was 2.9±0.9 cm with 24 left tumors and 20 right tumors, average R. E.N.A.L. score was 9.1±1.3, and average preoperative eGFR was 81.7±9.6 ml/(min·1.73 m 2). There was no significant difference in above variables between two groups. The operative time, warm ischemia time, blood loss, postoperative complication, postoperative hospital stay and postoperative 3 months renal function of two groups were compared. Results:All 73 patients successfully underwent RAPN or LPN and no patient converted to radical nephrectomy or open surgery. There was no significant difference in operation time [140(80, 160) min vs. 150 (90, 180) min, P=0.264], intraoperative estimated blood loss[150 (100, 200)vs. 180 (120, 200) ml, P=0.576]and postoperative hospital stay (7.0±2.7 vs. 7.4±2.1 days, P=0.480) between two groups. Compared with LPN group, RAPN group had obvious less warm ischemia time (23.1±3.3 vs. 27.6±4.7 min, P<0.001). No obvious complication occurred in RAPN group and one case with postoperative hemorrhage occurred in LPN group. No positive margin occurred in either group. There was no difference in renal function 3 months after operation between the two groups [73.2±6.3 vs.70.5±7.6ml/(min·1.73 m 2), P=0.117]. The median follow-up period was 22.6 months with no tumor recurrence or metastasis. Conclusions:For experienced surgeons, both RAPN and LPN are safe and feasible for patients with completely endophytic renal tumor. Compared with LPN, RAPN has advantages of perioperative curative effect, which could reduce the operating difficulty and shorten the warm ischemia time.