Automated lesion detectors in retinal fundus images.

Diabetic retinopathy (DR) is a sight-threatening condition occurring in persons with diabetes, which causes progressive damage to the retina. The early detection and diagnosis of DR is vital for saving the vision of diabetic persons. The early signs of DR which appear on the surface of the retina are the dark lesions such as microaneurysms (MAs) and hemorrhages (HEMs), and bright lesions (BLs) such as exudates. In this paper, we propose a novel automated system for the detection and diagnosis of these retinal lesions by processing retinal fundus images. We devise appropriate binary classifiers for these three different types of lesions. Some novel contextual/numerical features are derived, for each lesion type, depending on its inherent properties. This is performed by analysing several wavelet bands (resulting from the isotropic undecimated wavelet transform decomposition of the retinal image green channel) and by using an appropriate combination of Hessian multiscale analysis, variational segmentation and cartoon+texture decomposition. The proposed methodology has been validated on several medical datasets, with a total of 45,770 images, using standard performance measures such as sensitivity and specificity. The individual performance, per frame, of the MA detector is 93% sensitivity and 89% specificity, of the HEM detector is 86% sensitivity and 90% specificity, and of the BL detector is 90% sensitivity and 97% specificity. Regarding the collective performance of these binary detectors, as an automated screening system for DR (meaning that a patient is considered to have DR if it is a positive patient for at least one of the detectors) it achieves an average 95-100% of sensitivity and 70% of specificity at a per patient basis. Furthermore, evaluation conducted on publicly available datasets, for comparison with other existing techniques, shows the promising potential of the proposed detectors.

Astra Tech and Brånemark System implants: a prospective 5-year comparative study. Results after one year.

BACKGROUND: Endosseous dental implants are used frequently, and many implant systems are available. The scientific documentation of the implant system presents a great variation, and it is often difficult to compare studies of different systems. PURPOSE: The aim of this study was to compare two Swedish implant systems (Astra Tech and Brånemark System implants), in a prospective randomized study. MATERIALS AND METHODS: Sixty-six patients were equally distributed between the two implant systems; 184 Astra Tech and 187 Brånemark System implants were used. The patients have been followed annually with clinical and radiographic examinations. The results after 1 year are reported. RESULTS: The abutment procedure was found to be easier and less time-consuming with Astra Tech than with Brånemark implants. The operation times in minutes (mean +/- SEM) were for the respective implant 35 +/- 4.0 and 51 +/- 4.8 in the maxilla and 32 +/- 3.8 and 43 +/- 2.4 in the mandible. The differences in both cases were significant: p < .02 and p < .05, respectively. The failure rate for Astra Tech implants was 0.5% and for Brånemark implants 4.3%. The difference was significant (p < .05); however, taking into account that five of the eight implant losses in the Brånemark implant group occurred in one patient, an intraindividual correlation cannot be excluded. Therefore, this result should be interpreted with caution. The marginal bone level changes were examined already from the fixture installation. The major bone loss was found between fixture installation and baseline. This bone loss was several times greater than the bone loss between the baseline and the 1-year follow-up. The total bone loss during the observation period did not differ significantly between the systems, but they had different resorption patterns. The bone loss in the upper jaw between baseline and 1-year follow-up was 0.22 +/- 0.14 and 0.03 +/- 0.09 mm for the Astra Tech and Brånemark implants, respectively. In the lower jaw, the loss was -0.31 for both systems. The frequency of plaque accumulation and bleeding on probing did not differ between the implant systems. CONCLUSIONS: Abutment connection with Astra Tech implants was simpler than the corresponding surgery with Brånemark System implants and the survival rate of Astra Tech implants was higher than that of Brånemark system implants.

A 5-year multicenter study on implant-supported single crown restorations.

In this multicenter prospective study, the results achieved with the use of Brånemark implants for single tooth replacement were evaluated. The overall cumulative success rate was 95.9% for implants and 91.1% for crowns. Two of the 99 implants placed had to be removed before the prosthodontic stage of treatment; thus, 97 were restored with CeraOne crowns. Seventy-seven implants were evaluated radiographically at the 1-year follow-up, 57 at 3 years, and 47 at 5 years. Mean marginal bone resorption was well within the limits set by Albrektsson et al in 1986. The status of the soft tissue around crowns and adjacent teeth remained stable over the evaluation period. The gold abutment screw in the CeraOne system seems to have eliminated the problem of loosening abutment screws in single tooth replacements. The results suggest that the Brånemark system can be safely used for tissue-integrated replacement of single teeth.

Implant-supported overdentures: a longitudinal prospective study.

The aim of this study was to evaluate the clinical function and long-term prognosis of overdentures retained by a small number of implants in the maxilla and mandible using one of two different attachment systems. Included in the study were all patients referred to specialty clinics in Jönköping and Linköping, Sweden, during the treatment period who needed an overdenture and could be provided with a minimum number of two bilaterally-placed implants. Excluded were patients with bone-grafted jaws, irradiated cancer patients, heavy bruxers, and patients who had lost a fixed prosthesis because of implant losses. The patients were randomly assigned to receive one retentive system, either a round 2-mm-diameter bar with clips or ball attachments (Nobel Biocare). Eighteen overdentures were placed in maxillae and 32 in mandibles, supported by a total of 115 Brånemark implants. Of the implants placed, 86.1% were continuously osseointegrated. The cumulative implant survival rates after 7 years of loading were 75.4% in the maxillae and 100% in the mandibles. There was no difference in implant survival rate between the attachment systems. Patients with implant losses were characterized by severely resorbed maxillary ridges and inferior bone quality, together with unfavorable loading circumstances such as short implants combined with long leverages. Complications and prosthetic adjustments were mostly resolved early and easily.

Single-tooth replacement by osseointegrated Brånemark implants. A retrospective study of 82 implants.

The aim of the study was to evaluate the outcome of single-tooth restorations on Brånemark implants performed during the period 1984-1989. The material comprised 58 consecutive patients with 82 implants. The observation period varied from 1 to 5 years. Two implants were lost, both of them before loading. The overall survival rate was 97.6%. Radiographic follow-up was performed for 72 implants after 1 year and for 51 implants after 2 years. The bone loss during the first year was on average 9.6 +/- 0.1 mm and during the second year 0.1 +/- 0.1 mm. Although the observation time for most of the patients was only 2 years, the results indicate that the outcome of single-tooth restorations may be as favorable as for implants in edentulous jaws.

Osseointegrated oral implants. A Swedish multicenter study of 8139 consecutively inserted Nobelpharma implants.

Fourteen Swedish teams outside the University of Gothenburg, each with minimally three years' experience in the Nobelpharma osseointegrated implant participated in a retrospective multiclinic study. The total number of consecutively inserted implants at the 14 clinics was 8139. The outcome of every implant was reported and all implant failures, irrespective of when they occurred, were published. The success criteria included absence of implant mobility, absence of radiolucent zones on x-rays, and an annual bone loss after the first year of less than 0.2 mm. In the mandible 334 implants were followed for five to eight years, with only three failures, for a success rate of 99.1%. In the maxilla 106 implants were followed for five to seven years, with a success rate of 84.9%. In irradiated and grafted mandibles, 56 implants were inserted and none was lost during a follow-up of up to five years. In the irradiated maxilla there were 16 implants inserted with three reported failures and in the grafted upper jaw 71 implants were inserted with 12 failures. The proportions of mandibular and maxillary sleeping implants were 0.8 and 0.3%, of patient drop-out implants 0.3 and 0.6%, and of patient death implants 0.9 and 1.2%, respectively. It was concluded that the osseointegrated implant, if inserted according to the guidelines of Brånemark, results in a very high degree of clinical success, thereby meeting any published oral implant success criteria.

One-sided difference schemes and transonic flow.

Two one-sided conservation form difference approximations to a scalar one-dimensional convex conservation law are introduced. These are respectively of first- and second-order accuracy and each has the minimum possible band-width. They are nonlinearly stable, they converge only to solutions satisfying the entropy condition, and they have sharp monotone profiles. No such stable approximation of order higher than two is possible. Dimensional splitting algorithms are constructed and used to approximate the small-disturbance equation of transonic flow. These approximations are also nonlinearly stable and without nonphysical limit solutions.

Absorbing boundary conditions for numerical simulation of waves.

In practical calculations, it is often essential to introduce artificial boundaries to limit the area of computation. Here we develop a systematic method for obtaining a hierarchy of local boundary conditions at these artifical boundaries. These boundary conditions not only guarantee stable difference approximations, but also minimize the (unphysical) artificial reflections that occur at the boundaries.