Machine learning-based surrogate model assisting stochastic model predictive control of urban drainage systems.

Quantifying the uncertainty of stormwater inflow is critical for improving the resilience of urban drainage systems (UDSs). However, the high computational complexity and time consumption obstruct the implementation of uncertainty-addressing methods for real-time control of UDSs. To address this issue, this study developed a machine learning-based surrogate model (MLSM) that maintains high-fidelity descriptions of drainage dynamics and meanwhile diminishes the computational complexity. With stormwater inflow and controls as inputs and system overflow as the output, MLSM is able to fast evaluate system performance, and therefore stochastic optimization becomes feasible. Thus, a real-time control strategy was built by combining MLSM with the stochastic model predictive control. This strategy used stochastic stormwater inflow scenarios as input and aimed to minimize the expected overflow under all scenarios. An ensemble of stormwater inflow scenarios was generated by assuming the forecast errors follow normal distributions. To downsize the ensemble, representative scenarios with their probabilities were selected using the simultaneous backward reduction method. The proposed control strategy was applied to a combined UDS of China. Results are as follows. (1) MLSM fit well with the original high-fidelity urban drainage model, while the computational time was reduced by 99.1%. (2) The proposed strategy consistently outperformed the classical deterministic model predictive control in both magnitude and duration dimensions of system resilience, when the consumed time compatible is with the real-time operation. It is indicated that the proposed control strategy could be used to inform the real-time operation of complex UDSs and thus enhance system resilience to uncertainty.

Dam construction alters planktonic microbial predator‒prey communities in the urban reaches of the Yangtze River.

While dam construction supports social and economic development, changes in hydraulic conditions can also affect natural aquatic ecosystems, especially microbial ecosystems. The compositional and functional traits of multi-trophic microbiota can be altered by dam construction, which may result in changes in aquatic predator-prey interactions. To understand this process, we performed a large-scale sampling campaign in the urban reaches of the dam-impacted Yangtze River (1 995 km) and obtained 211 metagenomic datasets and water quality data. We first compared the compositional traits of planktonic microbial communities upstream, downstream, and in a dam reservoir. Results showed that Bacteroidetes (R-strategy) bacteria were more likely to survive upstream, whilst the reservoir and downstream regions were more conducive to the survival of K-strategy bacteria such as Actinobacteria. Eukaryotic predators tended to be enriched upstream, whilst phototrophs tended to be enriched in the reservoir and downstream regions. Based on bipartite networks, we inferred that the potential microbial predator-prey interactions gradually and significantly decreased from upstream to the downstream and dam regions, affecting 56% of keystone microbial species. Remarkably, functional analysis showed that the abundance of the photosynthetic gene psbO was higher in the reservoir and downstream regions, whilst the abundance of the KEGG carbohydrate metabolic pathway was higher upstream. These results indicate that dam construction in the Yangtze River induced planktonic microbial ecosystem transformation from detritus-based food webs to autotroph-based food webs.

Dam construction reshapes sedimentary pollutant distribution along the Yangtze river by regulating sediment composition.

Dam construction has far-reaching impacts on pollutant accumulation and the pollutant-induced quality of aquatic environments. Nonetheless, its large-scale effects on pollutant distribution in sediments, which greatly contribute to the environmental impacts of coexisting pollutants, remain poorly understood. We collected sediments from the Yangtze River during the dry and normal seasons (with 'normal' defined in terms of precipitation level), and examined how dam construction alters the spatial trajectories of both inorganic and organic pollutants in the sediments. Sediment composition exhibited linear variation from the upper to the lower reaches, with clay and silt particles dominating the sediment in the Three Gorges Reservoir and sand particles dominating in the middle-lower reaches. Accordingly, upstream of the Three Gorges Dam (TGD), sedimentary carbon, nitrogen, phosphorus, heavy metal, polycyclic aromatic hydrocarbons (PAHs), and oxygenated PAHs (OPAHs) contents increased toward the TGD owing to its regulation of the spatial variation in sediment particle size. The TGD caused upstream sedimentary accumulation of pollutants to be higher nearer to the TGD than in the upper reaches by 17%-129% for carbon, nitrogen, and phosphorus, 7%-51% for heavy metals, 30% for PAHs, and 140% for OPAHs. Pollutant content was sharply lower below the TGD, by 0.58-11.15 times for carbon, nitrogen, and phosphorus, 0.1-2.6 times for heavy metals, 1.7 times for PAHs, and 5.6 times for OPAHs. Upstream of the TGD, levels of NH4+-N, the main form of N in the interstitial water of the Yangtze River, increased lineary toward the TGD, whereas those of NO3--N and NO2--N decreased. Sedimentary organic matter source contributions were consistent along the Yangtze River, being on an average 46% for C3 plants and 28% for soil organic substances, further confirming the dam's regulatory effect on pollutants. These findings provide a foundation for future assessments of the environmental impact of dam-induced river fragmentation and hydrological alterations, and for developing advanced watershed pollutant management strategies.

Identifying key drivers of harmful algal blooms in a tributary of the Three Gorges Reservoir between different seasons: Causality based on data-driven methods.

Intense harmful algal blooms (HABs) can occur in the backwaters of tributaries supplying large-scale reservoirs. Due to the characteristics of process-based models and difficulties in modelling complex nonlinear processes, traditional models have difficulties disentangling the driving factors of HABs. In this study, we used data-driven methods (i.e., correlation analysis and machine-learning models) to identify the most important drivers of HABs in the Xiangxi River, a tributary of the Three Gorges Reservoir, China (2017-2018), for the dry season (from October to mid-April) and wet season (from April to September). We utilized the maximal information coefficient (MIC) combined with a time lag strategy and prior knowledge to quantitatively identify the driving variables of HABs. An extra trees regression (ETR) model was developed to assess the relative importance of causal variables driving algal blooms for the different periods. The results showed that water temperature was the most important driver for the duration of the study, followed by total nitrogen. Nitrogen had a stronger effect on algal blooms than phosphorus during both the wet and dry seasons. HABs were mainly affected by ammonia nitrogen in the wet season and by other forms of nitrogen in the dry season. In contrast, rather than the water temperature and nutrients, the operation of the Three Gorges Dam (difference between inflow and outflow discharge rate) was the most significant factor for algal blooms during the dry season, but its influence sharply declined during the wet season. This study showed that the key drivers of HABs can differ between seasons and suggests that HAB management should take seasonality into account.

Life-cycle assessment of two sewage sludge-to-energy systems based on different sewage sludge characteristics: Energy balance and greenhouse gas-emission footprint analysis.

Anaerobic digestion and incineration are widely used sewage sludge (SS) treatment and disposal approaches to recovering energy from SS, but it is difficult to select a suitable technical process from the various technologies. In this study, life-cycle assessments were adopted to compare the energy- and greenhouse gas- (GHG) emission footprints of two sludge-to-energy systems. One system uses a combination of AD with incineration (the AI system), whereas the other was simplified by direct incineration (the DI system). Comparison between three SS feedstocks (VS/TS: 57.61 -73.1 ds.%) revealed that the AI system consistently outperformed the DI system. The results of sensitivity analyses showed that the energy and GHG emission performances were mainly affected by VS content of the SS, AD conversion efficiency, and the energy consumption of sludge drying. Furthermore, the energy and GHG emission credit of the two systems increased remarkably with the increase in the VS content of the SS. For the high-organic-content sludge (VS/TS: 55%-80%), the energy and GHG emission credit of the AI system increase with the increase of AD conversion efficiency. However, for the low organic content sludge (VS/TS: 30%-55%), it has the opposite effect. In terms of energy efficiency and GHG performance, the AI system is a good choice for the treatment of high-organic-content sludge (VS/TS>55%), but DI shows superiority over AI when dealing with low organic content sludge (VS/TS<55%).

[Prognostic analysis of 280 patients with T4a gastric cancer who underwent curative resection].

OBJECTIVE: To evaluate the predicators of survival and recurrence after curative resection for gastric cancer patients with serosal invasion. METHODS: Clinical and follow-up data of 280 patients with T4a gastric cancer(according to the 7th UICC TNM Classification for gastric cancer) who underwent curative resection for gastric cancer between January 1997 and December 2003 were analyzed retrospectively. RESULTS: Five-year survival rate of these 280 patients was 50.0% and 158 patients (56.4%) presented with recurrence after surgery. The 3-year disease-free survival rate was 57.5%. The number of lymph node metastasis and metastatic lymph node ratio(MLR) were identified as the independent predicators for overall survival(P<0.05). The extent of lymph node metastasis was identified as the independent predicators for disease free survival (P<0.05). The extent of lymph node metastasis, the number of lymph node metastasis, and MLR were identified as the independent predicators for recurrence after surgery(P<0.05). CONCLUSIONS: The number of lymph node metastases, extent of lymph node metastasis and MLR are independent predictors of overall survival and recurrence in T4a gastric cancer patients after curative resection.

Ratio between negative and positive lymph nodes is suitable for evaluation the prognosis of gastric cancer patients with positive node metastasis.

OBJECTIVE: To date, there is no consensus to evaluate the most appropriate category of the nodal metastasis for precise predication the prognosis of gastric cancer patients with positive node metastasis after curative surgery. METHODS: We retrospectively analyzed the clinicopathologic characteristics and overall survival (OS) of 299 gastric cancer patients with positive node metastasis after curative surgery for evaluation the optimal category of the nodal metastasis. RESULTS: With the univariate and multivariate survival analyses, the depth of primary tumor invasion was identified as the independent predicators with the OS of 299 gastric cancer patients with nodal metastasis postoperatively, as were the number of positive lymph nodes (PLNs), the number of negative lymph nodes (NLNs), and the ratio between negative and positive lymph nodes (RNPL). The RNPL was identified to be more suitable for predication the OS of gastric cancer patients with positive node metastasis than the ratio between positive and dissected lymph nodes (RPDL) by using the stratum procedure of survival analysis. Besides, we found both PLNs and NLNs were independently correlated with OS of gastric cancer patients with nodal metastasis when RNPL, instead of RPDL, was controlled in the partial correlation model. CONCLUSIONS: RNPL, a new category of the nodal metastasis, was suitable for predication the OS of gastric cancer patients with nodal metastasis after curative resection, as were the PLNs, and NLNs.

Extended lymphadenectomy improvement of overall survival of gastric cancer patients with perigastric node metastasis.

BACKGROUND: The superiority of extended lymphadenectomy for the prognosis of gastric cancer (GC) is still controversial. The authors hypothesized that extended lymphadenectomy could improve the overall survival (OS) of GC patients with a specific extent of lymph node metastasis. METHODS: Data from 456 GC patients who underwent curative gastrectomy with lymphadenectomy were used to illuminate the difference of OS between patients who underwent limited lymphadenectomy and patients who underwent extended lymphadenectomy. RESULTS: As a whole, there was no significant difference of OS between patients who underwent extended lymphadenectomy and patients who underwent limited lymphadenectomy in all 456 GC patients. However, we demonstrated that extended lymphadenectomy significantly improved the OS of GC patients with perigastric lymph node metastasis (n1 stage based on the Japanese Gastric Cancer Association classification) compared to limited lymphadenectomy (P = 0.023). Furthermore, the more the negative lymph nodes were, the longer the OS of GC patients with perigastric node metastasis following extended lymphadenectomy was (P < 0.001). CONCLUSIONS: Extensive lymph node dissection and good harvest of negative lymph nodes should be deemed as the most important factors to improve the OS of GC patients with perigastric node metastasis.

Investigation of the recurrence patterns of gastric cancer following a curative resection.

PURPOSE: The goal of this study was to investigate the recurrence patterns of gastric cancer and determine the predictive information of recurrence patterns of gastric cancer following a curative resection. METHODS: This study retrospectively analyzed the data of 308 gastric cancer patients who underwent a curative resection, to identify the factors associated with the recurrence patterns. RESULTS: One hundred and sixty-nine gastric cancer patients had recurrence following curative resection. One hundred and twenty-six patients were observed for 3 years after the operation. Locoregional recurrence formed part of the recurrence pattern in 107 patients, peritoneal dissemination was observed in 98 patients, and distant metastasis occurred in 22 patients. A multivariate analysis revealed that locoregional recurrence was only associated with the Lauren classification (P = 0.003); peritoneal dissemination was only associated with N stage (P < 0.001); and distant metastasis was only associated with the Lauren classification (P = 0.016). CONCLUSIONS: Locoregional recurrence, peritoneal dissemination, and distant metastasis were the most common recurrence patterns of gastric cancer following a curative resection. Each recurrence pattern is associated with specific clinicopathological factors.

Suitability of 7th UICC N stage for predicting the overall survival of gastric cancer patients after curative resection in China.

BACKGROUND: Metastatic lymph node count (MLNC) is the most intensively prognostic indicator of gastric cancer. How to efficiently evaluate the best classification of MLNC is unclear. The purpose of this study was to evaluate the overall survival (OS) prediction of new UICC N stage in gastric cancer after curative surgery. MATERIALS AND METHODS: Data from 456 patients who underwent curative resection were used to choose the most efficient classification of MLNC for evaluation the OS of gastric cancer. RESULTS: Using univariate analysis, both the N stage of 7th edition UICC TNM classification (7th UICC N stage) and N stage of 5th/6th edition UICC TNM classification (5th/6th UICC N stage) were associated with the OS of gastric cancer after curative surgery. However, Cox regression multivariate analysis showed the 7th UICC N stage was an independent factor for predicting the OS of gastric cancer instead of the 5th/6th UICC N stage. Besides, we used the case-control matched fashion for further validation of the superiority of the 7th UICC N stage in prognostic prediction of gastric cancer. Last, we adopted the cut-point survival analysis to determine the most appropriate cutoffs for MLNC of all gastric cancer patients after curative surgery. We demonstrated the cutoff of 7th UICC N stage was similar to that produced from the cut-point survival analysis. CONCLUSION: The 7th UICC N stage appears to provide a reliable prognostic category of MLNC of gastric cancer than the 5th/6th UICC N stage, and it is the efficiently prognostic indicator of gastric cancer after curative surgery.

Enhancement the prediction of postoperative survival in gastric cancer by combining the negative lymph node count with ratio between positive and examined lymph nodes.

BACKGROUND: The purpose of this study was to evaluate the impact of the negative lymph node (NLN) count on the prognostic prediction of the ratio between positive and examined lymph nodes (RML) in gastric cancer after curative resection. METHODS: The positive and negative node counts were determined for 456 patients who underwent curative resection for gastric cancer. Overall survival was examined according to clinicopathologic variables. The correlation between the NLN count and the aforementioned best variable for prediction the disease-specific overall survival was examined. RESULTS: The NLN count cutoffs were designed as 0-9, 10-14, and > or =15, with the 5-year survival rate 4.1, 30.7, and 74.8%, respectively. RML of 98 patients who had an NLN count of nine or fewer was > or =40%. The median survival of these patients was 12 months. Of 88 patients who had 10 to 14 NLN count, 7 had 74-month median survival with 0.1-10% RML, 52 had 47-month median survival with 10.1-40% RML, and 29 had 22-month median survival with >40% RML. Of 270 patients who had > or =15 NLN count, 157 had 114-month median survival without positive nodes, 62 had 98-month median survival with 0.1-10% RML, 45 had 40-month median survival with 10.1-40% RML, and 6 had 14-month median survival with >40% RML. CONCLUSIONS: The NLN count is a key factor for improvement of survival prediction of RML in gastric cancer.

Prognostic factors of young patients with colon cancer after surgery.

AIM: To investigate the prognostic factors of 96 young patients with colon cancer within a cancer center by univariate and multivariate analysis. METHODS: A total of 723 patients with colon cancer were treated surgically during a period of 10 years. Ninety six of them were 40 years old or younger.R0, R1 and R2 operations were performed in 69 (71.9%), 4(4.1%) and 23 patients(24%), respectively. Left hemicolectomy was performed in 43 patients, right hemicolectomy in 37 patients, transverse colon resection in 9 patients and low anterior resection in 7 patients. Cox multivariate regression analysis was performed to identify predictors of survival. RESULTS: The operation mortality was 0%, 54 patients died within 111 mo after operation due to occurrence or metastases of the tumor. Liver,lung and bone metastases occurred in 3, 1 and 5 patients, respectively. The mean survival time for all patients was 77.9 +/- 5.01 mo and the overall 3-,5- and 10- year survival rates were 66.68%, 58.14% and 46.54%, respectively. In the univariate survival analysis,patient age,type of operation, radical resection, blood transfusion, histological type, diameter of tumor,depth of tumor invasion, lymphatic invasion, distant metastases, liver metastases and TNM stage were found to be predictors of survival in young patients with colon cancer. In the Cox-regression analysis,blood transfusion and lymphatic invasion were determined as independent prognostic factors of survival. CONCLUSIONS: Age, type of operation, radical resection, blood transfusion, histological type, diameter of tumor,depth of tumor invasion, lymphatic invasion, distant metastasis and TNM stage are the predictors of survival in young patients with colon cancer after surgery.

[Determinants of long-term survival in 38 patients with carcinoma of ampulla of Vater treated by local resection].

OBJECTIVE: To investigate determinants of long-term survival for carcinoma of ampulla of Vater treated by local resection. METHODS: The clinical and pathological data of 38 such patients treated by local resection from 1983 to 2003 were retrospectively analyzed. According to UICC staging system, there were T1 30, T2 7 and T3 1. Lymph nodes were involved in 4 during operation which was present in primary lesions larger than 2 cm across. All patients were treated by local resection. At first, external palpation was carried out to ascertain accessibility. Then with the duodenum opened, direct exploration was carried out. On deciding for resection, the common bile duct was probe explored which guided the circumferential ring resection 1 cm, away from the tumor, including all layers of duodenum, ampula and partial bile and terminal pancreatic ducts and the posterial wall of duodenum was completed in steps. Meticulous care was taken not to suture the pancreatic duct and endotheliation was ensured at the mouth of common bile duct and duodenum. The basal tissue was frozen sectioned to ensure negative stumps. The gall bladder of 6 patients was also resected. SPSS 10.0 software was used in data processing, log-rank test used in univariate analysis and Cox equation for multivariate analysis and Kaplan-Meirer method for the survival rates. RESULTS: Thirty-eight patients received local resection giving an operative mortality of 0% and morbidity of 13.2%. The 1-, 5- and 10-year survival rate was 83.5%, 51.4%, and 38.9%, respectively, with a median survival of 3.35 years. Up to now, 13 patients have survived for more than five years and 2 patients beyond ten years. The tumour size, tumour grading, lymph node status and UICC stage were significant prognostic factors in univariate analysis. However, only lymph node status was a statistically independent predictor of prognosis in multivariate analysis. CONCLUSION: Local excision is safe giving low morbidity and good survival in carefully selected cases. Preferably it is indicated only in high risk patients with a pT1 and well differentiated ampullary cancer smaller than 1 cm in diameter.

[Analysis of prognostic factors of rectal cancer in the elderly].

BACKGROUND & OBJECTIVE: The relation between patient age and outcomes from colorectal cancer surgery is complex. It is generally believed that age should not be a determinant in consideration of prognosis for colorectal cancer patient. But there were few studies on the prognostic factors of rectal cancer in the elderly. We therefore performed a retrospective analysis of clinicopathologic characteristics of rectal cancer in 343 old patients with rectal cancer by univariate and multivariate analysis. METHODS: A total of 343 patients (older than 60 years) with rectal cancer were treated surgically during a period of 10 years. R0, R1, and R2 operations were carried out in 261 patients (76.09%), 29 patients (8.45%), and 53 patients (15.45%), respectively. Low anterior resection was performed in 116 patients; 169 patients underwent Miles operation; 58 patients underwent the other operations. RESULTS: The operation mortality was 0.87%; 149 patients died of occurrence or metastases of the tumor within 108 months postoperatively. Liver, lung, and bone metastases were occurred in 17, 18, and 1 patients, respectively. The mean survival time for all patients was 72.12+/-2.60 months and the overall 3-, 5-, and 10-year survival rates were 69.62%, 55.73%, and 34.23%, respectively. Univariate analysis showed that the predictors of survival were type of operations, radical resection, histological type, diameter of the tumors, depth of tumor invasion, lymphatic invasion, distance metastases, liver and lung metastases. Multivariate analysis showed that only radical resection, lymphatic invasion, liver and lung metastases were independent factors. CONCLUSION: The follow prognostic factors can influence the survival of rectal cancer in the elderly: type of operations, nature of operation, histological type, diameter of the tumors, depth of tumor invasion, lymphatic invasion, distance metastases, liver and lung metastases. The independent factors were nature of operation, lymphatic invasion, liver and lung metastases.

[Multivariate Cox analysis on prognostic factors after surgery for rectal carcinoma].

OBJECTIVE: To analyze a large cohort of patients with rectal cancer within a cancer center to determine the prognostic factors by univariate and multivariate analyses. METHODS: A total of 952 patients with rectal cancer were treated surgically during a period of 10 years. R0, R1 and R2 operations were carried out in 741 patients (77.8%), 75 patients (7.9%) and 136 patients (14.3%), respectively. There were more Miles operation (53.5%) than lower abdominal resection (LAR, 33.7%). RESULTS: The operation mortality was 0.3%, 418 patients were dead within 108 months postoperatively due to recurrence or metastases to liver, lung and bone in 53, 39 and 12 patients. The overall mean survival time for all patients was 73.52 +/- 1.70 months and the overall 3-, 5-and 10-year survival rates were 67.6%, 55.4% and 38.2%. The overall 3-, 5- and 10-year survival rates for patients treated by radical operation were 81.4%, 70.3%, 48.8%, respectively. Kaplan-Meier estimate showed that patient gender, age, radicality of resection, histological type, liver and pulmonary metastasis and TNM stage were the predictors of survival. Multivariate analysis showed statistically significant correlation with radicality of operation, histological type, depth of tumor invasion, lymphatic invasion, TNM stage, liver and pulmonary metastasis. CONCLUSION: For survival, statistically significant differences among prognostic factors in relation to radicality of resection, lymphatic invasion, TNM stage, depth of tumor invasion, histological type, liver and pulmonary metastasis are found.

[Prospective randomized trial of prophylaxis of postoperative peritoneal carcinomatosis of advanced gastric cancer: intraperitoneal chemotherapy with mitomycin C bound to activated carbon particles].

OBJECTIVE: To evaluate the beneficial effect of intraperitoneally applied mitomycin bound to activated carbon particles (MMC-CH) in the prevention and treatment of intraabdominal recurrence after curative surgery for gastric cancer. METHODS: One hundred and twenty-four patients with radically resected gastric cancer infiltrating the serosal surface were randomly divided into group receiving 50 mg mitomycin bound to a solution of 375 mg carbon adsorbent intraperitoneally before closure of the abdominal wound (n = 62) and a control group (n = 62). The patients with MMC-CH and the control group were received systemic chemotherapy 3 months or 3 weeks after operation respectively. The postoperative recurrence-free survival was evaluated to analyze the benefits of this treatment. RESULTS: After observation for 8 months (range, 2 - 65). The 3-, 5-year postoperative recurrence-free survival rates were significantly higher in the MMC-CH group (70.16%, 44.51%) than in the control group (27.09%, 14.45%), P < 0.01. CONCLUSION: Adjuvant intraperitoneal chemotherapy of gastric cancer by mitomycin bound to activated carbon particles is effected by an increased postoperative recurrence-free survival rate.

Diagnosis and treatment of primary gastric non-Hodgkin's lymphoma: analysis of 157 patients.

OBJECTIVE: To seek the optimum treatment for patients with primary gastric non-Hodgkin's lymphoma and factors associated with prognosis. METHODS: A retrospective study was conducted on 157 primary gastric non-Hodgkin's lymphoma patients who had received operation for 45 years. RESULTS: The X-ray diagnosis rate was 39.4% before operation. The diagnosis rate by gastroscopy was 52.7%. Among the 157 patients, 32 belonged to stage I(E), 40 stage II(E), 29 stage III(E), and 56 stage IV(E). All of the patients were received chemotherapy or radiation. The 3-, 5-, 10-, 15-year survival rates were 51.1% (69/135), 42.3% (55/130), 20.7% (23/111), and 13.5% (14/104). CONCLUSIONS: The 3-, 5-year survival rates in stage I(E) and stage II(E) were 2 to 5 times higher than those in stage III(E) and IV(E) (P < 0.01). The 3-, 5-year survival rates of primary gastric non-Hodgkin's lymphoma were 60.2% (65/108) and 50.0% (52/104) respectively. The prognosis was better than the 5-year survival rate of gastric cancer patients with D(2) lymphodenectomy (33.3%). Early diagnosis and treatment are effective to prevent complications, enhance quality of patient's life, and prolong the survival.

[Primary lymphoma of the spleen: clinical analysis of 23 cases].

OBJECTIVE: To investigate the best diagnostic and therapeutic method for primary lymphoma of the spleen. METHODS: Clinicopathologic features of 23 patients treated from January 1956 to August 1999 were analyzed retrospectively. RESULTS: All patients but one for exploration only (96%) underwent resection of the tumor. They accepted chemotherapy after operation. 23 patients were confirmed pathologically. B-cell type non-Hodgkin's lymphoma was noted in 21 patients and T-cell letion in 2. According to Ahman's staging, 9 patients belonged to stage I, 8 stage II, and 6 stage III. The 5-year survival rates were 50%, 40% and 16% respectively. CONCLUSIONS: The diagnosis of splenic lymphoma is dependent mainly on B-ultrasound examination and CT scanning. Splenectomy combined with chemotherapy may provide optimum therapy for patients with splenic lymphoma.