Hematopoietic SCT (HSCT) is a well-recognized therapeutic procedure to prolong life and cure patients with lifethreatening hematological malignancies; however, the risk of developing secondary carcinoma may increase in longterm survivors. The objective of this study was to determine the incidence and risk factors for secondary squamous carcinoma after HSCT. Between 1984 and 2004, 170 allogeneic HSCT recipients aged 415 years, who had survived for 45 years were enrolled. Demographic data and the characteristics of secondary carcinoma were collected and analyzed for the determination of the incidence and risk of developing secondary carcinoma. Eight patients developed secondary carcinoma, including five oral squamous cell carcinomas, one esophageal, one gastric and one ovarian carcinoma, but no cutaneous carcinomas were detected at a median follow-up of 14.1 years (range, 5.1-23.3 years) after HSCT. The accrual 10-year cumulative incidence of secondary carcinoma was 2.89%. In univariate and multivariate analyses, chronic GVHD and age 440 years at the time of HSCT were both significant risk factors independently associated with the development of secondary carcinoma. Thus, the occurrence of secondary carcinoma is one of the late complications in patients undergoing HSCT. Oral squamous cell carcinoma was more common in our patients after HSCT, indicating the need for lifelong surveillance of the oral cavity. Moreover, because of the relatively long latency in developing secondary carcinoma, extended follow-up is required for a thorough understanding of the incidence and characteristics of secondary carcinoma after HSCT.

Funding Acknowledgements Type of funding sources: None. Background Timely recognition and intervention of severe tricuspid regurgitation (TR) improve clinical outcomes, especially given advancing transcatheter devices nowadays. We developed deep neural network model to automatically recognize significant tricuspid regurgitation through echocardiography in both assessing color flow and leaflet morphology. Method With careful annotation of corresponded images, we developed three stage models. First model to perform view classification, second joint models to do image processing including 2D image segmentation, leaflet pose estimation, and color image regurgitation flow segmentation, and the third model conduct detection of significant TR. Cross validation was performed stage by stage to confirm performance and model stability. Result The first view classification EfficientNetB5 model reached averaged classification accuracy 94.8%. The second joint models designed in Hourglass structure to achieve averaged 97.1% and 83.3% intersection over union (IOU) for regurgitation flow and chamber segmentation, respectively, and 0.9%-1.6% mean square error for 4 nodes each of three cuspid of tricuspid valve. With the joint of abovementioned post-processing information and original raw image, our model possessed averaged 88% accuracy to detect significant tricuspid regurgitation. Conclusion Our study confirmed that with the utilization of color flow and morphological information, deep neural network model can achieve reliable automatic recognition presence of significant TR.