We introduce the notion of iceberg concept lattices and show their use in knowledge discovery in ... more We introduce the notion of iceberg concept lattices and show their use in knowledge discovery in databases. Iceberg lattices are a conceptual clustering method, which is well suited for analyzing very large databases. They also serve as a condensed representation of frequent itemsets, as starting point for computing bases of association rules, and as a visualization method for association rules. Iceberg concept lattices are based on the theory of Formal Concept Analysis, a mathematical theory with applications in data analysis, information retrieval, and knowledge discovery. We present a new algorithm called TITANIC for computing (iceberg) concept lattices. It is based on data mining techniques with a level-wise approach. In fact, TITANIC can be used for a more general problem: Computing arbitrary closure systems when the closure operator comes along with a so-called weight function. The use of weight functions for computing closure systems has not been discussed in the literature up to now. Applications providing such a weight function include association rule mining, functional dependencies in databases, conceptual clustering, and ontology engineering. The algorithm is experimentally evaluated and compared with Ganter's Next-Closure algorithm. The evaluation shows an important gain in efficiency, especially for weakly correlated data. Ó
In this paper, we propose the algorithm PASCAL which introduces a novel optimization of the well-... more In this paper, we propose the algorithm PASCAL which introduces a novel optimization of the well-known algorithm Apriori. This optimization is based on a new strategy called pattern counting inference that relies on the concept of key patterns. We show that the support of frequent non-key patterns can be inferred from frequent key patterns without accessing the database. Experiments comparing PAS-CAL to the three algorithms Apriori, Close and Max-Miner, show that PASCAL is among the most efficient algorithms for mining frequent patterns.
SymTA/S is a system-level performance and timing analysis approach based on formal scheduling ana... more SymTA/S is a system-level performance and timing analysis approach based on formal scheduling analysis techniques and symbolic simulation. The tool supports heterogeneous architectures, complex task dependencies and context aware analysis. It determines system-level performance data such as end-to-end latencies, bus and processor utilization, and worst-case scheduling scenarios. SymTA/S furthermore combines optimization algorithms with system sensitivity analysis for rapid design space exploration. This paper gives an overview of the current research interests in the SymTA/S project.
| Discovering association rules is one of the most important task in data mining. Many e cient al... more | Discovering association rules is one of the most important task in data mining. Many e cient algorithms have been proposed in the literature. The most noticeable are Apriori, Mannila's algorithm, Partition, Sampling and DIC, that are all based on the Apriori mining method: pruning the subset lattice (itemset lattice). In this paper we propose an e cient algorithm, called Close, based on a new mining method: pruning the closed set lattice (closed itemset lattice). This lattice, which is a sub-order of the subset lattice, is closely related to Wille's concept lattice in formal concept analysis. Experiments comparing Close to an optimized version of Apriori showed that Close is very e cient for mining dense and/or correlated data such as census style data, and performs reasonably well for market basket style data.
The objective of this study was to characterize the patient population with respect to patient se... more The objective of this study was to characterize the patient population with respect to patient selection, assess surgical morbidity and graft failures, and analyze the contribution of peri-operative clinical factors to recipient outcome in adult living donor liver transplantation (ALDLT). Summary Background Data: Previous reports have been centerspecific or from large databases lacking detailed variables. The Adult-to-Adult Living Donor Liver Transplantation Cohort Study (A2ALL) represents the first detailed North American multicenter report of recipient risk and outcome aiming to characterize variables predictive of graft failure. Methods: Three hundred eighty-five ALDLT recipients transplanted at 9 centers were studied with analysis of over 35 donor, recipient, intraoperative, and postoperative variables. Cox regression models were used to examine the relationship of variables to the risk of graft failure. Results: Ninety-day and 1-year graft survival were 87% and 81%, respectively. Fifty-one (13.2%) grafts failed in the first 90 days. The most common causes of graft failure were vascular thrombosis, primary nonfunction, and sepsis. Biliary complications were common (30% early, 11% late). Older recipient age and length of cold ischemia were significant predictors of graft failure. Center experience greater than 20 ALDLT was associated with a significantly lower risk of graft failure. Recipient Model for End-stage Liver Disease score and graft size were not significant predictors. Conclusions: This multicenter A2ALL experience provides evidence that ALDLT is a viable option for liver replacement. Older recipient age and prolonged cold ischemia time increase the risk of graft failure. Outcomes improve with increasing center experience. (Ann Surg 2005;242: 314 -325)
Objective: Few studies have evaluated long-term outcomes after orthotopic liver transplantation (... more Objective: Few studies have evaluated long-term outcomes after orthotopic liver transplantation (OLT). This work analyzes the experience of nearly 2 decades by the same team in a single center. Outcomes of OLT and factors affecting survival were analyzed. Methods: Retrospective analysis of 3200 consecutive OLTs that were performed at our institution, between February 1984 and December 31, 2001. Results: Of 2662 recipients, 578 (21.7%) and 659 (24.7%) were pediatric and urgent patients, respectively. Overall 1-, 5-, 10-, and 15-year patient and graft survival estimates were 81%, 72%, 68%, 64% and 73%, 64%, 59%, 55%, respectively. Patient survival significantly improved in the second (1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001) versus the era I (1984 -1991) of transplantation (P Ͻ 0.001). Similarly, graft survival was better in the era II of transplantation (P Ͻ 0.02). However, biliary and infectious complications increased in era II. When OLT indications were considered, best recipient survival was obtained in children with biliary atresia (82%, 79%, and 78% at 1, 5, and 10 years, respectively), while malignant disease in adult patients resulted in the worst outcomes of 68% and 43% at 1 and 5 years, post-OLT. Further, patients Ͻ18 years and nonurgent recipients exhibited superior survival when compared with recipients Ͼ18 years (P Ͻ 0.001) or urgent patients (P Ͻ 0.001). Of 13 donor and recipient variables, era of OLT, recipient age, urgent status, donor age, donor length of hospital stay, etiology of liver disease, retrans-plantation, warm and cold ischemia, but not graft type (whole, split, living-donor), significantly impacted patient survival. Conclusions: Long-term benefits of OLT are greatest in pediatric and nonurgent patients. Multiple factors involving the recipient, etiology of liver disease, donor characteristics, operative variables, and surgical experience influence long-term survival outcomes. By balancing and matching these factors with a given recipient, optimum results can be achieved. (Ann Surg 2005;241: 905-918) Survival curves were computed using Kaplan-Meier methods and compared using log-rank tests. Medians were Busuttil et al
To determine the factors affecting the outcome of orthotopic liver transplantation (OLT) for end-... more To determine the factors affecting the outcome of orthotopic liver transplantation (OLT) for end-stage liver disease caused by hepatitis C virus (HCV) and to identify models that predict patient and graft survival. The national epidemic of HCV infection has become the leading cause of hepatic failure that requires OLT. Rapidly increasing demands for OLT and depleted donor organ pools mandate appropriate selection of patients and donors. Such selection should be guided by a better understanding of the factors that influence the outcome of OLT. The authors conducted a retrospective review of 510 patients who underwent OLT for HCV during the past decade. Seven donor, 10 recipient, and 2 operative variables that may affect outcome were dichotomized at the median for univariate screening. Factors that achieved a probability value less than 0.2 or that were thought to be relevant were entered into a stepdown Cox proportional hazard regression model. Overall patient and graft survival rates at 1, 5, and 10 years were 84%, 68%, and 60% and 73%, 56%, and 49%, respectively. Overall median time to HCV recurrence was 34 months after transplantation. Neither HCV recurrence nor HCV-positive donor status significantly decreased patient and graft survival rates by Kaplan-Meier analysis. However, use of HCV-positive donors reduced the median time of recurrence to 22.9 months compared with 35.7 months after transplantation of HCV-negative livers. Stratification of patients into five subgroups, based on time of recurrence, revealed that early HCV recurrence was associated with significantly increased rates of patient death and graft loss. Donor, recipient, and operative variables that may affect OLT outcome were analyzed. On univariate analysis, recipient age, serum creatinine, donor length of hospital stay, donor female gender, United Network for Organ Sharing (UNOS) status of recipient, and presence of hepatocellular cancer affected the outcome of OLT. Elevation of pretransplant HCV RNA was associated with an increased risk of graft loss. Of 15 variables considered by multivariate Cox regression analysis, recipient age, UNOS status, donor gender, and log creatinine were simultaneous significant predictors for patient survival. Simultaneously significant factors for graft failure included log creatinine, log alanine transaminase, log aspartate transaminase, UNOS status, donor gender, and warm ischemia time. These variables were therefore entered into prognostic models for patient and graft survival. The earlier the recurrence of HCV, the greater the impact on patient and graft survival. The use of HCV-positive donors may accelerate HCV recurrence, and they should be used judiciously. Patient survival at the time of transplantation is predicted by donor gender, UNOS status, serum creatinine, and recipient age. Graft survival is affected by donor gender, warm ischemia time, and pretransplant patient condition. The authors' current survival prognostic models require further multicenter validation.
The authors determined the long-term outcome of patients undergoing hepatic retransplantation at ... more The authors determined the long-term outcome of patients undergoing hepatic retransplantation at their institution. Donor, operative, and recipient factors impacting on outcome as well as parameters of patient resource utilization were examined.
To assess the efficacy of orthotopic liver transplantation (OLT) for hepatocellular carcinoma (HC... more To assess the efficacy of orthotopic liver transplantation (OLT) for hepatocellular carcinoma (HCC) and the impact of current staging criteria on long term survival. Summary Background Data: HCC is becoming an increasingly common indication for OLT. Medicare approves OLT only for HCCs meeting the Milan criteria, thus limiting OLT for an expanding pool of potential liver recipients. We analyzed our experience with OLT for HCC to determine if expansion of criteria for OLT for HCC is warranted. Methods: All patients undergoing OLT for HCC from 1984 to 2006 were evaluated. Outcomes were compared for patients who met Milan criteria (single tumor Յ5 cm, maximum of 3 total tumors with none Ͼ3 cm), University of California, San Francisco (UCSF) criteria (single tumor Ͻ6.5 cm, maximum of 3 total tumors with none Ͼ4.5 cm, and cumulative tumor size Ͻ8 cm), or exceeded UCSF criteria. Results: A total of 467 transplants were performed for HCC. At mean follow up of 6.6 Ϯ 0.9 years, recurrence rate was 21.2%, and overall 1, 3, and 5-year survival was 82%, 65%, and 52%, respectively. Patients meeting Milan criteria had similar 5-year posttransplant survival to patients meeting UCSF criteria by preoperative imaging (79% vs. 64%; P ϭ 0.061) and explant pathology (86% vs. 71%; P ϭ 0.057). Survival for patients with tumors beyond UCSF criteria was significantly lower and was below 50% at 5 years. Multivariate analysis showed that tumor number (P Ͻ 0.001), lymphovascular invasion (P Ͻ 0.001), and poor differentiation (P ϭ 0.002) independently predicted poor survival. Conclusions: This largest single institution experience with OLT for HCC demonstrates prolonged survival after liver transplantation for tumors beyond Milan criteria but within UCSF criteria, both when classified by preoperative imaging and by explant pathology. Measured expansion of OLT criteria is justified for tumors not exceeding the UCSF criteria.
We introduce the notion of iceberg concept lattices and show their use in knowledge discovery in ... more We introduce the notion of iceberg concept lattices and show their use in knowledge discovery in databases. Iceberg lattices are a conceptual clustering method, which is well suited for analyzing very large databases. They also serve as a condensed representation of frequent itemsets, as starting point for computing bases of association rules, and as a visualization method for association rules. Iceberg concept lattices are based on the theory of Formal Concept Analysis, a mathematical theory with applications in data analysis, information retrieval, and knowledge discovery. We present a new algorithm called TITANIC for computing (iceberg) concept lattices. It is based on data mining techniques with a level-wise approach. In fact, TITANIC can be used for a more general problem: Computing arbitrary closure systems when the closure operator comes along with a so-called weight function. The use of weight functions for computing closure systems has not been discussed in the literature up to now. Applications providing such a weight function include association rule mining, functional dependencies in databases, conceptual clustering, and ontology engineering. The algorithm is experimentally evaluated and compared with Ganter's Next-Closure algorithm. The evaluation shows an important gain in efficiency, especially for weakly correlated data. Ó
In this paper, we propose the algorithm PASCAL which introduces a novel optimization of the well-... more In this paper, we propose the algorithm PASCAL which introduces a novel optimization of the well-known algorithm Apriori. This optimization is based on a new strategy called pattern counting inference that relies on the concept of key patterns. We show that the support of frequent non-key patterns can be inferred from frequent key patterns without accessing the database. Experiments comparing PAS-CAL to the three algorithms Apriori, Close and Max-Miner, show that PASCAL is among the most efficient algorithms for mining frequent patterns.
SymTA/S is a system-level performance and timing analysis approach based on formal scheduling ana... more SymTA/S is a system-level performance and timing analysis approach based on formal scheduling analysis techniques and symbolic simulation. The tool supports heterogeneous architectures, complex task dependencies and context aware analysis. It determines system-level performance data such as end-to-end latencies, bus and processor utilization, and worst-case scheduling scenarios. SymTA/S furthermore combines optimization algorithms with system sensitivity analysis for rapid design space exploration. This paper gives an overview of the current research interests in the SymTA/S project.
| Discovering association rules is one of the most important task in data mining. Many e cient al... more | Discovering association rules is one of the most important task in data mining. Many e cient algorithms have been proposed in the literature. The most noticeable are Apriori, Mannila's algorithm, Partition, Sampling and DIC, that are all based on the Apriori mining method: pruning the subset lattice (itemset lattice). In this paper we propose an e cient algorithm, called Close, based on a new mining method: pruning the closed set lattice (closed itemset lattice). This lattice, which is a sub-order of the subset lattice, is closely related to Wille's concept lattice in formal concept analysis. Experiments comparing Close to an optimized version of Apriori showed that Close is very e cient for mining dense and/or correlated data such as census style data, and performs reasonably well for market basket style data.
The objective of this study was to characterize the patient population with respect to patient se... more The objective of this study was to characterize the patient population with respect to patient selection, assess surgical morbidity and graft failures, and analyze the contribution of peri-operative clinical factors to recipient outcome in adult living donor liver transplantation (ALDLT). Summary Background Data: Previous reports have been centerspecific or from large databases lacking detailed variables. The Adult-to-Adult Living Donor Liver Transplantation Cohort Study (A2ALL) represents the first detailed North American multicenter report of recipient risk and outcome aiming to characterize variables predictive of graft failure. Methods: Three hundred eighty-five ALDLT recipients transplanted at 9 centers were studied with analysis of over 35 donor, recipient, intraoperative, and postoperative variables. Cox regression models were used to examine the relationship of variables to the risk of graft failure. Results: Ninety-day and 1-year graft survival were 87% and 81%, respectively. Fifty-one (13.2%) grafts failed in the first 90 days. The most common causes of graft failure were vascular thrombosis, primary nonfunction, and sepsis. Biliary complications were common (30% early, 11% late). Older recipient age and length of cold ischemia were significant predictors of graft failure. Center experience greater than 20 ALDLT was associated with a significantly lower risk of graft failure. Recipient Model for End-stage Liver Disease score and graft size were not significant predictors. Conclusions: This multicenter A2ALL experience provides evidence that ALDLT is a viable option for liver replacement. Older recipient age and prolonged cold ischemia time increase the risk of graft failure. Outcomes improve with increasing center experience. (Ann Surg 2005;242: 314 -325)
Objective: Few studies have evaluated long-term outcomes after orthotopic liver transplantation (... more Objective: Few studies have evaluated long-term outcomes after orthotopic liver transplantation (OLT). This work analyzes the experience of nearly 2 decades by the same team in a single center. Outcomes of OLT and factors affecting survival were analyzed. Methods: Retrospective analysis of 3200 consecutive OLTs that were performed at our institution, between February 1984 and December 31, 2001. Results: Of 2662 recipients, 578 (21.7%) and 659 (24.7%) were pediatric and urgent patients, respectively. Overall 1-, 5-, 10-, and 15-year patient and graft survival estimates were 81%, 72%, 68%, 64% and 73%, 64%, 59%, 55%, respectively. Patient survival significantly improved in the second (1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001) versus the era I (1984 -1991) of transplantation (P Ͻ 0.001). Similarly, graft survival was better in the era II of transplantation (P Ͻ 0.02). However, biliary and infectious complications increased in era II. When OLT indications were considered, best recipient survival was obtained in children with biliary atresia (82%, 79%, and 78% at 1, 5, and 10 years, respectively), while malignant disease in adult patients resulted in the worst outcomes of 68% and 43% at 1 and 5 years, post-OLT. Further, patients Ͻ18 years and nonurgent recipients exhibited superior survival when compared with recipients Ͼ18 years (P Ͻ 0.001) or urgent patients (P Ͻ 0.001). Of 13 donor and recipient variables, era of OLT, recipient age, urgent status, donor age, donor length of hospital stay, etiology of liver disease, retrans-plantation, warm and cold ischemia, but not graft type (whole, split, living-donor), significantly impacted patient survival. Conclusions: Long-term benefits of OLT are greatest in pediatric and nonurgent patients. Multiple factors involving the recipient, etiology of liver disease, donor characteristics, operative variables, and surgical experience influence long-term survival outcomes. By balancing and matching these factors with a given recipient, optimum results can be achieved. (Ann Surg 2005;241: 905-918) Survival curves were computed using Kaplan-Meier methods and compared using log-rank tests. Medians were Busuttil et al
To determine the factors affecting the outcome of orthotopic liver transplantation (OLT) for end-... more To determine the factors affecting the outcome of orthotopic liver transplantation (OLT) for end-stage liver disease caused by hepatitis C virus (HCV) and to identify models that predict patient and graft survival. The national epidemic of HCV infection has become the leading cause of hepatic failure that requires OLT. Rapidly increasing demands for OLT and depleted donor organ pools mandate appropriate selection of patients and donors. Such selection should be guided by a better understanding of the factors that influence the outcome of OLT. The authors conducted a retrospective review of 510 patients who underwent OLT for HCV during the past decade. Seven donor, 10 recipient, and 2 operative variables that may affect outcome were dichotomized at the median for univariate screening. Factors that achieved a probability value less than 0.2 or that were thought to be relevant were entered into a stepdown Cox proportional hazard regression model. Overall patient and graft survival rates at 1, 5, and 10 years were 84%, 68%, and 60% and 73%, 56%, and 49%, respectively. Overall median time to HCV recurrence was 34 months after transplantation. Neither HCV recurrence nor HCV-positive donor status significantly decreased patient and graft survival rates by Kaplan-Meier analysis. However, use of HCV-positive donors reduced the median time of recurrence to 22.9 months compared with 35.7 months after transplantation of HCV-negative livers. Stratification of patients into five subgroups, based on time of recurrence, revealed that early HCV recurrence was associated with significantly increased rates of patient death and graft loss. Donor, recipient, and operative variables that may affect OLT outcome were analyzed. On univariate analysis, recipient age, serum creatinine, donor length of hospital stay, donor female gender, United Network for Organ Sharing (UNOS) status of recipient, and presence of hepatocellular cancer affected the outcome of OLT. Elevation of pretransplant HCV RNA was associated with an increased risk of graft loss. Of 15 variables considered by multivariate Cox regression analysis, recipient age, UNOS status, donor gender, and log creatinine were simultaneous significant predictors for patient survival. Simultaneously significant factors for graft failure included log creatinine, log alanine transaminase, log aspartate transaminase, UNOS status, donor gender, and warm ischemia time. These variables were therefore entered into prognostic models for patient and graft survival. The earlier the recurrence of HCV, the greater the impact on patient and graft survival. The use of HCV-positive donors may accelerate HCV recurrence, and they should be used judiciously. Patient survival at the time of transplantation is predicted by donor gender, UNOS status, serum creatinine, and recipient age. Graft survival is affected by donor gender, warm ischemia time, and pretransplant patient condition. The authors' current survival prognostic models require further multicenter validation.
The authors determined the long-term outcome of patients undergoing hepatic retransplantation at ... more The authors determined the long-term outcome of patients undergoing hepatic retransplantation at their institution. Donor, operative, and recipient factors impacting on outcome as well as parameters of patient resource utilization were examined.
To assess the efficacy of orthotopic liver transplantation (OLT) for hepatocellular carcinoma (HC... more To assess the efficacy of orthotopic liver transplantation (OLT) for hepatocellular carcinoma (HCC) and the impact of current staging criteria on long term survival. Summary Background Data: HCC is becoming an increasingly common indication for OLT. Medicare approves OLT only for HCCs meeting the Milan criteria, thus limiting OLT for an expanding pool of potential liver recipients. We analyzed our experience with OLT for HCC to determine if expansion of criteria for OLT for HCC is warranted. Methods: All patients undergoing OLT for HCC from 1984 to 2006 were evaluated. Outcomes were compared for patients who met Milan criteria (single tumor Յ5 cm, maximum of 3 total tumors with none Ͼ3 cm), University of California, San Francisco (UCSF) criteria (single tumor Ͻ6.5 cm, maximum of 3 total tumors with none Ͼ4.5 cm, and cumulative tumor size Ͻ8 cm), or exceeded UCSF criteria. Results: A total of 467 transplants were performed for HCC. At mean follow up of 6.6 Ϯ 0.9 years, recurrence rate was 21.2%, and overall 1, 3, and 5-year survival was 82%, 65%, and 52%, respectively. Patients meeting Milan criteria had similar 5-year posttransplant survival to patients meeting UCSF criteria by preoperative imaging (79% vs. 64%; P ϭ 0.061) and explant pathology (86% vs. 71%; P ϭ 0.057). Survival for patients with tumors beyond UCSF criteria was significantly lower and was below 50% at 5 years. Multivariate analysis showed that tumor number (P Ͻ 0.001), lymphovascular invasion (P Ͻ 0.001), and poor differentiation (P ϭ 0.002) independently predicted poor survival. Conclusions: This largest single institution experience with OLT for HCC demonstrates prolonged survival after liver transplantation for tumors beyond Milan criteria but within UCSF criteria, both when classified by preoperative imaging and by explant pathology. Measured expansion of OLT criteria is justified for tumors not exceeding the UCSF criteria.
Uploads
Papers by rafik rafik