Can preoperative spiral CT scans alone determine the feasibility of endovascular AAA repair? A comparison to angiographic measurements.

PURPOSE: To determine whether computed tomography (CT) alone can be used for excluding patients from endovascular repair for abdominal aortic aneurysms (AAA). METHODS: Among 71 patients evaluated for endovascular AAA repair using spiral CT imaging and angiography, 31 were selected who had both studies performed within 6 months of each other using a graduated measuring catheter or guidewire. Measurements of aneurysm neck diameter, neck length, and infrarenal aortic length were made from the CT and angiographic images using handheld calipers with calibration markers as guides. Infrarenal aortic length and neck length were determined from CT images by multiplying the width of the cuts by the number of slices between the lowest renal artery and the aortic bifurcation or the top of the aneurysm, respectively. RESULTS: CT neck diameter measurements differed significantly from the angiographic dimensions (6.3 +/- 5.1-mm mean difference, p < 0.001). In the majority of patients (25, 81%), CT neck diameters were larger (mean 7.3 +/- 3.8 mm). The mean difference in neck length measurements was 0.5 +/- 15.9 mm (p = NS). Twenty-two (71%) patients had aortic length measurements that were longer on the angiogram (mean 15.4 +/- 17.2 mm, p = NS). Five patients who would have been excluded as candidates based on overestimated CT neck diameter measurements subsequently underwent successful endovascular aneurysm repair. CONCLUSIONS: Considerable discrepancies exist between preoperative neck diameter and infrarenal aortic length measurements obtained from CT scans and angiograms used to evaluate candidates for endovascular aortic aneurysm repair. CT alone may not be adequate for predicting the feasibility of endovascular AAA repair.

A hybrid approach of neural network and memory-based learning to data mining.

We propose a hybrid prediction system of neural network and memory-based learning. Neural network (NN) and memory-based reasoning (MBR) are frequently applied to data mining with various objectives. They have common advantages over other learning strategies. NN and MBR can be directly applied to classification and regression without additional transformation mechanisms. They also have strength in learning the dynamic behavior of the system over a period of time. Unfortunately, they have shortcomings when applied to data mining tasks. Though the neural network is considered as one of the most powerful and universal predictors, the knowledge representation of NN is unreadable to humans, and this "black box" property restricts the application of NN to data mining problems, which require proper explanations for the prediction. On the other hand, MBR suffers from the feature-weighting problem. When MBR measures the distance between cases, some input features should be treated as more important than other features. Feature weighting should be executed prior to prediction in order to provide the information on the feature importance. In our hybrid system of NN and MBR, the feature weight set, which is calculated from the trained neural network, plays the core role in connecting both learning strategies, and the explanation for prediction can be given by obtaining and presenting the most similar examples from the case base. Moreover, the proposed system has advantages in the typical data mining problems such as scalability to large datasets, high dimensions, and adaptability to dynamic situations. Experimental results show that the hybrid system has a high potential in solving data mining problems.

Histology and electron microscopy of explanted bifurcated endovascular aortic grafts: evidence of early incorporation and healing.

PURPOSE: To report an examination of explanted bifurcated endovascular aortic grafts for histologic evidence of early healing and incorporation. METHOD: Two bifurcated endovascular aortic grafts composed of polycarbonate urethane and Elgiloy wire were explanted 42 and 21 days after successful endovascular exclusion of abdominal aortic aneurysms. Both patients expired from causes unrelated to endograft deployment. The explanted devices were examined using immunohistochemical analysis and electron microscopy. RESULTS: On explantation, both grafts appeared to have excluded the aneurysm with no evidence of endoleak, graft migration, or thrombosis. Histological examination showed numerous inflammatory cells and good ingrowth of tissue into the proximal 2 cm of the graft. Collagen and smooth muscle cells were evident in the proximal portion of the graft with only collagen in the distal segments. Neointimal formation was seen within the proximal 2 cm also, but not at the distal segments. Macrophages were present in the graft. Scanning electron microscopy showed an extensive matrix of fibers that most likely represented collagen. CONCLUSIONS: Bifurcated endovascular aortic grafts show inflammatory and mild foreign body reactions, collagen formation, and intimal ingrowth during healing. These findings are similar to some of the healing properties reported for sutured grafts, as well as other endovascular grafts.

Endoluminal repair of aneurysms containing ostia of essential branch arteries: an experimental model.

PURPOSE: To examine experimentally the feasibility of transfemoral endoluminal repair of aneurysms containing the ostia of essential branch arteries. METHODS: In a canine model (n = 4), suprarenal aortic aneurysms were created by suturing an artificial patch onto an anterior arteriotomy. Following a 2-week recovery period, the dogs underwent endovascular exclusion of their aneurysms using an aortic stent-graft with separate renal artery branch grafts. Outcome was evaluated using angiography, intravascular ultrasound (IVUS), Doppler flow, invasive pressure monitoring, and autopsy, respectively. RESULTS: Successful creation and subsequent endovascular exclusion of the aneurysm using aortic stent-grafts and separate bilateral renal artery stent-grafts was achieved in all trials. Angiographically, all aneurysms were excluded from aortic flow and all renal arteries were patent at completion of the procedure. With IVUS, good graft apposition and absence of perigraft flow were demonstrated in all animals. Mean pressure in the aneurysmal sac at completion of the procedure was 40 +/- 7 mmHg, compared to a mean systemic blood pressure of 105 +/- 8 mmHg (p < 0.05). At autopsy, no gross intimal damage was seen in the aorta or the renal arteries, and intact aortic grafts and branch grafts without twisting, coiling, or kinking were found in all trials. CONCLUSIONS: In an acute animal model, suprarenal aortic aneurysms can be excluded from the circulation with preservation of renal flow using an endoluminally placed aortic stent-graft with separate branch grafts.

Electrophoretic profiles and biological activities: intraspecific variation in the venom of the Malayan pit viper (Calloselasma rhodostoma)

The Malayan pit viper (Calloselasma rhodostoma) is of major clinical significance both as a leading cause of snakebite and as the source of ancrod (Arvin). Although its venom has been extensively studied, the degree to which venom composition varies between individuals is poorly known. We individually analysed the venoms of over 100 C. rhodostoma using isoelectric focusing. In all populations, females produced an intense band that was absent from all males, and significant ontogenetic variation was detected. Principal components analysis of the banding profiles also revealed strong geographic variation, which was significantly congruent with variation in the biological activities of the venom (phosphodiesterase, alkalinephosphoesterase, L-amino acid oxidase, arginine ester hydrolase, 5'-nucleotidase, thrombin-like enzyme, haemorrhagic activity). Studies of captive-bred snakes indicate that the intraspecific variation in venom is genetically inherited rather than environmentally induced. The intraspecific variation in venom composition and biological activity could be of applied importance to snakebite therapy, both in correct diagnosis of the source of envenomation and in the development of a more effective antivenom. Greater attention should be given to the source of C. rhodostoma venom used in research to ensure reproducibility of results.

Fibrinogen sialic acid residues are low affinity calcium-binding sites that influence fibrin assembly.

Calcium ions occupy low (n congruent to 10; Kd congruent to 1 mM) and high (n = 3; Kd congruent to 1 microM) affinity sites on fibrinogen and facilitate fibrin monomer polymerization. We have previously localized two of the three high affinity Ca2+ sites to gamma 311-gamma 336. However, optimal enhancement of fibrin monomer polymerization occurs only at physiological millimolar Ca2+ concentrations which are two orders of magnitude higher than the concentration required for occupancy of the high affinity Ca2+-binding sites. In this study, we show that removal of fibrinogen sialic acid residues results in loss of low affinity Ca2+-binding sites. Clotting of asialofibrinogen appears to be Ca2+-independent and results in fiber bundles thicker in diameter than normal fibrin bundles as determined by turbidometry and scanning and transmission electron microscopy. By using a Ca2+-sensitive electrode, free sialic acid is shown to bind Ca2+ (Kd congruent to 1 mM). These observations suggest that the high affinity fibrinogen D-domain Ca2+-binding sites may play a role in the tertiary structure of the D-domain, whereas, sialic acid residues are low affinity sites whose occupancy by Ca2+ at physiological calcium concentration facilitates fibrin polymerization.