A prognostic computer model to individually predict post-procedural complications in interventional cardiology: the INTERVENT Project

Supporting diagnostic decisions using hybrid and complementary data mining applications: a pilot study in the pediatric emergency department

INTRODUCTION

This article demonstrates the capacity of a combination of different data mining (DM) methods to support diagnosis in pediatric emergency patients. By using a novel combination of these DM procedures, a computer-based diagnosis was created.

METHODS

A support vector machine (SVM), artificial neural networks (ANNs), fuzzy logics, and a voting algorithm were simultaneously used to allocate a patient to one of 18 diagnoses (e.g., pneumonia, appendicitis). Anonymized data sets of patients who presented in the emergency department (ED) of a pediatric care clinic were chosen. For each patient, 26 identical clinical and laboratory parameters were used (e.g., blood count, C-reactive protein) to finally develop the program.

RESULTS

The combination of four DM operations arrived at a correct diagnosis in 98% of the cases, retrospectively. A subgroup analysis showed that the highest diagnostic accuracy was for appendicitis (97% correct diagnoses) and idiopathic thrombocytopenic purpura or erythroblastopenia (100% correct diagnoses). During the prospective testing, 81% of the patients were correctly diagnosed by the system.

DISCUSSION

The combination of these DM methods was suitable for proposing a diagnosis using both laboratory and clinical parameters. We conclude that an optimized combination of different but complementary DM methods might serve to assist medical decisions in the ED.

Stent parameters predict major adverse clinical events and the response to platelet glycoprotein IIb/IIIa blockade: findings of the ESPRIT trial

BACKGROUND

Only limited data describe relationships between stent parameters (length and diameter), adverse events after percutaneous coronary intervention, and effects of platelet glycoprotein IIb/IIIa blockade by stent parameters.

METHODS AND RESULTS

In this post hoc analysis of the 1983 patients receiving a stent in the Enhanced Suppression of the Platelet Glycoprotein IIb/IIIa Receptor with Integrilin Therapy randomized percutaneous coronary intervention trial of eptifibatide versus placebo, rates of the major adverse cardiac event (MACE) end point (death, myocardial infarction, urgent target-vessel revascularization, or thrombotic bailout) at 48 hours and 1 year were correlated with stent parameters and then analyzed by randomization to eptifibatide versus placebo. In the placebo group, MACE increased with number of stents implanted, total stent length (by quartiles of <15, 15 to <18, 18 to <30, and >or=30 mm), and total stented vessel area (by quartiles of area <141, 141 to <188, 188 to <292, and >or=292 mm(2)). By stent parameters, MACE at 48 hours was reduced in the eptifibatide group at stent lengths of 18 to <30 mm (odds ratio [OR], 0.55; 95% CI, 0.32 to 0.94; P=0.030) and >or=30 mm (OR, 0.43; 95% CI, 0.25 to 0.75; P=0.003), stent diameters of >2.5 to <3.5 mm (OR, 0.56; 95% CI, 0.39 to 0.82; P=0.002), and with 2 stents implanted (OR, 0.39; 95% CI, 0.22 to 0.69; P=0.001). In the placebo group, near-linear relationships were observed between both increasing stent length and increasing stented vessel area and MACE at 48 hours and 1 year (all, P<0.001); these gradients were flattened in the eptifibatide group (P=0.005 for stent length).

CONCLUSIONS

Stent parameters predict MACE after percutaneous coronary intervention. Glycoprotein IIb/IIIa blockade mitigates much of the hazard of increasing procedural complexity.

Assessment of a novel angiographic image stabilization system for percutaneous coronary intervention

BACKGROUND

Optimization of coronary images for percutaneous coronary intervention (PCI) remains difficult due to cardiac motion throughout the respiratory and cardiac cycles. We tested a novel system to stabilize angiographic images at the region of interest in order to assist during PCI.

METHODS

Patients undergoing PCI to the right coronary artery (RCA) (group 1, n = 22) or complex PCI (group 2, n = 16) were prospectively enrolled and the angiographic image sequences of patients who died suddenly of confirmed or presumed stent thrombosis following PCI (group 3, n = 16) were retrospectively reviewed. All image sequences were analyzed off-line by three cardiologists before and after image stabilization for accuracy of stent placement, presence of residual edge dissection, and adequacy of procedural outcome.

RESULTS

Image stabilization was successful in 100% of cases in a mean time of 95 +/- 71 seconds and was considered to be helpful in 13.6% of group 1, in 18.3% of group 2, and in 10% of group 3 cases. There was good correlation between observers with a kappa statistic of 0.85 to 1.0 for all observations. However, there was no difference in the reviewers' opinions of stent placement, presence of edge dissection, or adequacy of procedural result when comparing the standard angiographic views and the stabilized images. In particular, no previously unrecognized edge dissections were apparent in group 3 with stabilized display.

CONCLUSION

Image stabilization centered on the region of interest was considered helpful in a small subset of patients, particularly the complex PCI patients. However, no differences in objective parameters could be demonstrated.

Automatic analysis of auditory nerve electrically evoked compound action potential with an artificial neural network

The auditory nerve's electrically evoked compound action potential is recorded in deaf patients equipped with the Nucleus 24 cochlear implant using a reverse telemetry system (NRT). Since the threshold of the NRT response (NRT-T) is thought to reflect the psychophysics needed for programming cochlear implants, efforts have been made by specialized management teams to develop its use. This study aimed at developing a valid tool, based on artificial neural networks (ANN) technology, for automatic estimation of NRT-T. The ANN used was a single layer perceptron, trained with 120 NRT traces. Learning traces differed from data used for the validation. A total of 550 NRT traces from 11 cochlear implant subjects were analyzed separately by the system and by a group of physicians with expertise in NRT analysis. Both worked to determine 37 NRT-T values, using the response amplitude growth function (AGF) (linear regression of response amplitudes obtained at decreasing stimulus intensity levels). The validity of the system was assessed by comparing the NRT-T values automatically determined by the system with those determined by the physicians. A strong correlation was found between automatic and physician-obtained NRT-T values (Pearson r correlation coefficient >0.9). ANOVA statistics confirmed that automatic NRT-Ts did not differ from physician-obtained values (F = 0.08999, P = 0.03). Moreover, the average error between NRT-Ts predicted by the system and NRT-Ts measured by the physicians (3.6 stimulation units) did not differ significantly from the average error between NRT-Ts measured by each of the three physicians (4.2 stimulation units). In conclusion, the automatic system developed in this study was found to be as efficient as human experts for fitting the amplitude growth function and estimating NRT-T, with the advantage of considerable time-saving.

Simplified scoring system for predicting mortality after percutaneous coronary intervention

OBJECTIVES

We sought to develop a simplified scoring system based on pre-intervention clinical characteristics to predict in-hospital mortality after percutaneous coronary intervention (PCI).

BACKGROUND

Percutaneous coronary intervention is associated with variety of complications, including the risk of death. Factors leading to poor outcomes need to be identified. Currently available indexes are cumbersome and therefore seldom used.

METHODS

Crude mortality and univariate odds ratios (ORs) for mortality associated with multiple clinical characteristics were calculated for 9,954 patients undergoing PCI at the William Beaumont Hospital during 1996 to 1998. Based on the OR, each factor was assigned a weighted score. Using these scores, a classification was constructed to determine the probability of death after PCI, with classes I through IV representing an increasing probability of procedural mortality. This classification was validated in a separate group of patients.

RESULTS

The factors with the highest univariate odds of dying and their scores were: myocardial infarction <14 days = 7; elevated creatinine = 4; multivessel disease = 4; and age >65 years = 3. Classes were created based on the presence of these factors in a given patient. The odds of dying and mortality increased significantly with each class. These results were reproduced in the validation subset.

CONCLUSIONS

Preprocedural clinical risk factors have a differential influence on the probability of death after PCI. Risk classification based on these factors can be used to accurately predict the procedural outcome. This simple classification can be used by interventionalists to assist in management decisions, to provide an estimate of procedural risk to the patients and relatives, and for quality assurance.

Ischemic complications after percutaneous transluminal coronary angioplasty

The ischemic complications ofpercutaneous transluminal coronary angioplasty (PTCA) include abrupt closure, which occurs in 2% to 10% of patients and is associated with increased morbidity and mortality. Periprocedural myocardial infarction due to side branch occlusion or embolization of platelet aggregates or thrombus occurs in 5% to 20% of patients. Patients with acute coronary syndromes, older age, and complex lesions are at greater risk of periprocedural complications. Technical advances, primarily stenting, are useful in the prevention and management of acute closure, but are also accompanied by thrombotic complications. It remains to be seen whether the new antithrombin agents reduce the rate of periprocedural complications if used in combination with aspirin and new antiplatelet therapies. These new antiplatelet agents (ticlopidine, clopidogrel, abciximab, eptifibatide, and tirofiban) reduce the rate of ischemic complications and have become standard adjunctive therapy for patients who undergo PTCA.