Using machine-learning techniques, clinical diagnostic model research extracts diagnostic models from patient data. Traditionally, patient data are often collected using electronic Case Report Form ...(eCRF) systems, while mathematical software is used for analyzing these data using machine-learning techniques. Due to the lack of integration between eCRF systems and mathematical software, extracting diagnostic models is a complex, error-prone process. Moreover, due to the complexity of this process, it is usually only performed once, after a predetermined number of data points have been collected, without insight into the predictive performance of the resulting models.
The objective of the study of Clinical Data Miner (CDM) software framework is to offer an eCRF system with integrated data preprocessing and machine-learning libraries, improving efficiency of the clinical diagnostic model research workflow, and to enable optimization of patient inclusion numbers through study performance monitoring.
The CDM software framework was developed using a test-driven development (TDD) approach, to ensure high software quality. Architecturally, CDM's design is split over a number of modules, to ensure future extendability.
The TDD approach has enabled us to deliver high software quality. CDM's eCRF Web interface is in active use by the studies of the International Endometrial Tumor Analysis consortium, with over 4000 enrolled patients, and more studies planned. Additionally, a derived user interface has been used in six separate interrater agreement studies. CDM's integrated data preprocessing and machine-learning libraries simplify some otherwise manual and error-prone steps in the clinical diagnostic model research workflow. Furthermore, CDM's libraries provide study coordinators with a method to monitor a study's predictive performance as patient inclusions increase.
To our knowledge, CDM is the only eCRF system integrating data preprocessing and machine-learning libraries. This integration improves the efficiency of the clinical diagnostic model research workflow. Moreover, by simplifying the generation of learning curves, CDM enables study coordinators to assess more accurately when data collection can be terminated, resulting in better models or lower patient recruitment costs.
The aim of this study was to find the best 3D reconstruction technique to visualize the endometrial-myometrial junction (EMJ).
Retrospective observational study on 240 stored 3D volumes of 80 ...patients. The first author reconstructed the 2D midcoronal image without volume contrast imaging (VCI), with VCI set at 4 mm and with VCI set at 2 mm. Three images per patient (240 images) were saved and integrated in the web-based electronic data capture software Clinical Data Miner (CDM) (http://cdm.esat.kuleuven.be). Five experienced gynaecologists analysed the images shown in random order. They scored the image quality (good, moderate, poor, insufficient) and described the EMJ of these images using IETA terminology (regular, irregular, interrupted, not defined). One of the examiners (CVP) also re-evaluated the same set of images after 12 days to assess intra-observer variability.
The use of VCI significantly improved the recorded subjective image quality. The Fleiss' kappa coefficient for evaluating the inter-observer variability of the EMJ description using coronal view without VCI, with VCI at 4 mm and VCI at 2 mm were 0.36 ± 0.05, 0.34 ± 0.05 and 0.42 ± 0.05, respectively. The corresponding figures for the intra-observer variability were 0.58 ± 0.08, 0.36 ± 0.08 and 0.68 ± 0.07, respectively.
In this study on 3D reconstructed coronal images of the uterine cavity, the 2 mm VCI slices gave the best quality images of the EMJ.
To document the action of dopamine on gastrointestinal motility in mechanically ventilated patients.
Crossover, randomized, placebo-controlled study.
General intensive care unit (ICU) in a university ...hospital.
Twelve mechanically ventilated patients in a stable hemodynamic condition, with no contraindication to enteral feeding.
Dopamine (4 microg/kg per minute) and placebo were infused over 8 h (4 h fasting, followed immediately by 4 h nasogastric feeding at 100 kcal per hour) on two consecutive days, in a random order. Pressure changes in the gastric antrum (four sites) and in the duodenum (two sites) were recorded by perfused catheter manometry. Each session started with the institution of dopamine or placebo infusion.
The migrating motor complex and its three successive phases were identified (phase I, period of quiescence; phase II, period of irregular contractile activity; phase III or activity front, period of high-frequency, regular contractions). Contractions and activity fronts at each site were quantified during fasting and feeding. The mean duration of the fasting migrating motor complex was determined in the duodenum, as well as the contribution of each phase (phases I, II, III) to the length of the complete cycle. The propagation characteristics of each activity front were assessed visually. The number of contractions was lower in the antrum (p = 0.024) and phase III motor activity higher in the duodenum incidence of activity fronts (p = 0.008); number of phase III contractions (p = 0.009) during dopamine infusion than with placebo. These modifications observed under dopamine were related to decreased antral contractions during fasting (p = 0.050), increased incidence of activity fronts during feeding (p = 0.031), and increased number of phase III contractions during fasting (p = 0.037). In both groups (placebo and dopamine) activity fronts rarely started in the antrum, and abnormally propagated activity fronts were found in the duodenum in some patients.
Low-dose dopamine adversely affects gastroduodenal motility in mechanically ventilated critically ill patients.