A novel image processing technique for 3D volumetric analysis of severely resorbed alveolar sockets with CBCT.

BACKGROUND: The aim of this study was to present and validate a novel procedure for the quantitative volumetric assessment of extraction sockets that combines cone-beam computed tomography (CBCT) and image processing techniques. METHODS: The CBCT dataset of 9 severely resorbed extraction sockets was analyzed by means of two image processing software, Image J and Mimics, using manual and automated segmentation techniques. They were also applied on 5-mm spherical aluminum markers of known volume and on a polyvinyl chloride model of one alveolar socket scanned with Micro-CT to test the accuracy. RESULTS: Statistical differences in alveolar socket volume were found between the different methods of volumetric analysis (P<0.0001). The automated segmentation using Mimics was the most reliable and accurate method with a relative error of 1.5%, considerably smaller than the error of 7% and of 10% introduced by the manual method using Mimics and by the automated method using ImageJ. CONCLUSIONS: The currently proposed automated segmentation protocol for the three-dimensional rendering of alveolar sockets showed more accurate results, excellent inter-observer similarity and increased user friendliness. The clinical application of this method enables a three-dimensional evaluation of extraction socket healing after the reconstructive procedures and during the follow-up visits.

Prospective monocentric study of non-tunnelled central venous catheter-related complications in hematological patients.

Indwelling central venous catheters (CVCs) are used in the management of hematologic patients. However, insertion and maintenance of CVCs are susceptible to complications. Study design and methods data concerning 388 consecutive catheterisations, performed in oncohematologic patients between April 2003 and December 2004, were prospectively collected. At insertion thrombocytopenia was present in 109 cases (28.1%) and neutropenia in 67 (17.3%). Hemorrhage after CVC insertion occurred in five thrombocytopenic patients (1.3%). The median duration of catheterisation was 18.8 days (range 1-89), longer in the 7-French CVCs utilised in leukemic patients (24.3 days) and shorter in 12-French CVCs (11 days), used for PBSC harvesting. Deep venous thrombosis was diagnosed in 13 cases (3.3%). Ninety-two catheterisations (12.6/1000 days-catheter) were complicated by infections: 19 local infections (4.8%) and 73 (18.8%) bacteraemias of which 45 (11.6%) were catheter-related, mainly due to Gram positive germs (32/45, 71.1%). The frequency of catheter-related bacteraemia was 7.2 events/1000 days-catheter. Thirteen CVCs were removed due to thrombosis, 15 due to infections, 20 due to malfunction, the remaining 333 at patients discharge. At univariate analysis high-dose chemotherapy (p = 0.013), 7-Fr lumen (p = 0.023), acute myeloid leukemia (AML) (p = 0.001), duration of neutropenia >10 days and length of catheterisation were significantly correlated to infection. Multivariate analysis confirmed the duration of catheterisation, AML and high-dose chemotherapy as risk factors. Even though hematological in-patients are at increased risk for bleeding and infections, non-tunnelled CVCs offer a safe venous access also in patients affected by severe thrombocytopenia and prolonged neutropenia.