Mining data from intensive care patients

In this paper we describe the application of data mining methods for predicting the evolution of patients in an intensive care unit. We discuss the importance of such methods for health care and other application domains of engineering. We argue that this problem is an important but challenging one for the current state of the art data mining methods and explain what improvements on current methods would be useful. We present a promising study on a preliminary data set that demonstrates some of the possibilities in this area.     

Slowly moving hull forms in short waves

The purpose of this paper is to provide a mathematical tool to improve the optimal design of ship forms. It is common practice that hull forms are designed such that they have minimum wave resistance in calm water. In this paper a theory is described by which the effect of short waves may be incorporated.The basic tool we use is the ray theory. First, the appropriate free-surface condition is shown. Then, the standard ray method, well-known in geometric optics, is formulated in the fluid region and at the free surface. After an elimination process the eiconal equation and the transport equation are obtained. The characteristic equation for the nonlinear eiconal equation is derived, keeping in mind that the characteristics are not perpendicular to the wave fronts, due to the effect of the double-body potential due to the forward speed of the ship, which is assumed to be a good approximation for the steady potential.Numerical computations are carried out by means of the RK4 method to obtain the ray pattern. After some manipulations the amplitude may be computed just as well. Finally, the nonlinear added-resistance force is calculated. Pictures of ray patterns for several angles of incidence are shown. Also the forces are shown.     

Long-term results after allogeneic bone marrow transplantation for chronic myelogenous leukemia in chronic phase: a report from the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation

The purpose of this study was to determine the long-term results of allogeneic bone marrow transplantation for chronic myeloid leukemia. A retrospective analysis was carried out of the outcome of 373 consecutive transplants performed at 38 European institutions between 1980 and 1988 and reported to the registry of the European Group for Blood and Marrow Transplantation. All transplants were carried out for first chronic phase of chronic myelogenous leukemia using unmanipulated marow cells from HLA-identical sibling donors. The probability of survival and leukemia-free survival at 8 years were 54% (95% CI: 49-59) and 47% (95% CI: 41-52) respectively. The probabilities of developing acute GVHD (II-IV) at 100 days and chronic GVHD at 4 years after transplant were 47% (95% CI: 41-53) and 52% (95% CI: 46-58) respectively. The probabilities of transplant-related mortality and leukemic relapse 8 years after BMT were 41% (95% CI: 36-48) and 19% (95% CI: 14-25), respectively. Transplant within 12 months of diagnosis was associated with reduced transplant-related mortality (34 vs 45%, P = 0.013) and resulted in improved leukemia-free survival (52 vs 44%, P = 0.03). The probability of relapse was significantly reduced in patients who developed chronic GVHD (RR = 0.33, P = 0.004). The probability of relapse occurring more than 2 years after transplant was increased more than five-fold in patients transplanted from a male donor (RR = 5.5, P = 0.006). Sixty-seven patients in hematologic remission were studied for residual disease by two-step RT/PCR for BCR-ABL mRNA and 61 (91%) tested negative. We conclude that bone marrow transplantation can induce long-term survival in approximately one-half of CML patients; the majority of survivors have no evidence of residual leukemia cells when studied by molecular techniques. The probability of late relapse is increased with use of a male donor.     

Autoxidation of Hydrocarbons: From Chemistry to Catalysis

This contribution summarizes our recent efforts to unravel the radical chemistry of autoxidations. Abstraction of the weakly bonded αH-atom of the primary hydroperoxide product by chain carrying peroxyl radicals is identified as the source of major end products such as alcohol and ketone/aldehyde. In the case of cyclohexane autoxidation, this reaction is even responsible for the majority of ring-opened by-products. In a second part, different approaches to autoxidation catalysis are discussed, ranging from transition metal ion catalysis to organocatalysis with immobilized N-hydroxyphthalimide.     

Risk assessment for patients with chronic myeloid leukaemia before allogeneic blood or marrow transplantation. Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation

BACKGROUND: Transplantation of blood or bone-marrow stem cells is the treatment of choice for selected patients with chronic myeloid leukaemia (CML). Transplantation is used with increasing frequency and success, but remains associated with substantial risks of morbidity and mortality. Other treatments with satisfactory short-term outcome are available. For appropriate counselling of patients, a rapid and simple way to assess risk is needed. METHODS: Data from 3142 patients (1873 [60%] male, 1269 [40%] female; mean age 34 years, range <1-60 years) treated with allogeneic blood or marrow transplants for CML between 1989 and 1997, reported to the European Group for Blood and Marrow Transplantation (EBMT), were used to develop and test a simple risk score based on previously reported major pretransplant risk factors: histocompatibility, stage of disease at time of transplantation, age and sex of donor and recipient, and time from diagnosis to transplantation. We analysed probabilities of survival, leukaemia-free survival, transplant-related mortality, and relapse incidence with respect to these risk factors. FINDINGS: At the time of analysis, 1922 (61%) of the 3142 patients were alive-1567 (65%) of those with HLA-identical sibling donors and 417 (57%) of those with unrelated donors. 1682 (54%) were alive without relapse. 1220 (39%) patients had died, 1013 (83%) of transplant-related causes, 207 (17%) of relapse. 447 (14%) patients had relapsed. The final scoring system was highly predictive for leukaemia-free survival, survival and transplant-related mortality. Survival at 5 years was 72%, 70%, 62%, 48%, 40%, 18%, and 22% for patients with scores 0, 1, 2, 3, 4, 5, and 6, respectively. Risk of transplant-related mortality was 20%, 23%, 31%, 46%, 51%, 71%, and 73%. Data showed the same trends for HLA-identical sibling transplants and unrelated transplants for transplants done in 1989-93 and 1994-96. INTERPRETATION: Pretransplant risk factors are cumulative for individual patients with CML having blood or marrow transplantation. A simple system based on five main factors gives adequate risk assessment for counselling of patients and taking decisions.     

Population balances for particulate processes—a volume approach

Population balance models have been used in chemical engineering since the 1960s and have evolved to become the most important tools for design and control of particulate processes. In this paper we show that the intrinsic particle parameter that determines changes in the process and should thus be included in the population balance is the particle volume. The basic population that is modeled should be the mass distribution, or the volume distribution if the density is constant. The population balance thus describes the change of the volume distribution of volume with time. Furthermore, we suggest that the “birth” and “death” terms that are often used to describe discrete events in particulate processes can almost always be replaced by a rate of change term.To design and control existing and future processes, a multi-dimensional population balance model is required. We propose a volume-based model in which the particle properties that are modeled are the volumes of solid, liquid, and air, respectively. In the most general case the model will consist of a properties vector and a distribution tensor. Depending on the complexity of the process, one or more of the properties may be omitted from the model. This is shown in three examples of increasing complexity: comminution, sintering, and granulation.     

Augmented Panoramic Video

Many video sequences consist of a locally dynamic background containing moving foreground subjects. In this paper we propose a novel way of re‐displaying these sequences, by giving the user control over a virtual camera frame. Based on video mosaicing, we first compute a static high quality background panorama. After segmenting and removing the foreground subjects from the original video, the remaining elements are merged into a dynamic background panorama, which seamlessly extends the original video footage. We then re‐display this augmented video by warping and cropping the panorama. The virtual camera can have an enlarged field‐of‐view and a controlled camera motion. Our technique is able to process videos with complex camera motions, reconstructing high quality panoramas without parallax artefacts, visible seams or blurring, while retaining repetitive dynamic elements.     

A generalized multiple layer data envelopment analysis model for hierarchical structure assessment: A case study in road safety performance evaluation

Data envelopment analysis (DEA) is a powerful analytical research tool for measuring the relative efficiency of a homogeneous set of decision making units (DMUs) by obtaining empirical estimates of relations between multiple inputs and multiple outputs related to the DMUs. To further embody multilayer hierarchical structures of these inputs and outputs in the DEA framework, which are prevalent in today’s performance evaluation activities, we propose a generalized multiple layer DEA (MLDEA) model. Starting from the input-oriented CCR model, we elaborate the mathematical deduction process of the MLDEA model, formulate the weights in each layer of the hierarchy, and indicate different types of possible weight restrictions. Meanwhile, its linear transformation is realized and further extended to the BCC form. To demonstrate the proposed MLDEA model, a case study in evaluating the road safety performance of a set of 19 European countries is carried out. By using 13 hierarchical safety performance indicators in terms of road user behavior (e.g., inappropriate or excessive speed) as the model’s input and 4 layered road safety final outcomes (e.g., road fatalities) as the output, we compute the most optimal road safety efficiency score for the set of European countries, and further analyze the weights assigned to each layer of the hierarchy. A comparison of the results with the ones from the one layer DEA model clearly indicates the usefulness and effectiveness of this improvement in dealing with a great number of performance evaluation activities with hierarchical structures.     

增强战斗力的有效途径:虚拟战争

在军事领域中,随着军事斗争的复杂性的不断增加,人们一直在寻求一种能在实验条件下“预先实践战争”的方法,来深入地研究未来战争的规律。现代科学技术的飞速发展,使得人们这一愿望得以实现。在虚拟现实技术基础上发展起来的“虚拟战争”方法,正是研究增强军队战斗力的有效途径。本文对这一课题作了概要的阐述。     

用MBR工艺处理印染前处理废水的试验研究

棉纺织印染前处理废水成分复杂、水质较差、处理难度较大。采用膜生物反应器(MBR)处理经过简单预处理的印染前处理废水,可使废水水质大为改善、后续处理压力大大减轻。在近100 d的现场试验中,得出了MBR的最佳操作参数为:温度25℃,HRT 18 h,MLSS 8 g/L,DO 2.8 mg/L,pH 7.5～8.5。在最佳工艺条件下,SS、色度、NH_3—N和COD_(Cr)去除率分别约为90%、30%、72%、72%。该工艺运行成本约为1.643元/m~3。