Automatic measurement of mandibular cortical bone width on cone-beam computed tomography images.

OBJECTIVE: The computed tomography cortical index (CTCI), computed tomography mandibular index (CTMI), and computed tomography index (inferior) [CTI(I)] are indexes obtained from cone-beam computed tomography images for the assessment of the mandibular cortex quality for implant planning or osteoporosis. However, cross-sectional image reconstruction for the measurements is labor-intensive. This study aimed to develop and evaluate a method to automatically reconstruct cross-sectional images and measure the cortex width in all areas inferior to the mental foramen (MF). METHODS: Seventy-one women (mean age: 52.4 years; range: 20-78 years) were enrolled. They were divided into four age and CTCI groups, including females younger (FY) and females older (FO) than 50 years (C1: normal, C2: mild/moderate erosion, and C3: severe porosity). Automatic and manual measurements of CTMI and CTI(I) were compared, and the inter- and intraobserver agreements were assessed using the intraclass correlation coefficient (ICC). The relationships between CTMI or CTI(I) and CTCI were also assessed. RESULTS: The mean processing times for reconstruction and measurements were 31.9 s and 1.22 s, respectively. ICCs for the comparison of automatic and manual measurements were 0.932 and 0.993 in the C1 and C2/C3 groups, respectively. Significant differences in CTMI and CTI(I) were observed between the FY or the FO-C1 and FO-C3 groups (p < 0.05). CONCLUSION: The automatic and manual measurements showed a strong agreement. The new method could drastically reduce routine clinical workload. Additionally, our method enables the measurement of the cortex width in all the mandibular bones inferior to the MF.

Automatic Remasking of Digital Subtraction Angiography Images in Pulmonary Angiography.

In pulmonary angiography, the heartbeat creates artifacts that hinder extraction of blood vessel images in digital subtraction angiography. Remasking according to the cardiac phase of the angiogram may be effective but has yet to be automated. Here, automatic remasking was developed and assessed according to the cardiac phase from electrocardiographic information collected simultaneously with imaging. Manual remasking, fixed remasking, and our proposed automatic remasking were applied to 14 pulmonary angiography series from five participants with either chronic thromboembolic pulmonary hypertension or pulmonary arteriovenous malformation. The processing time and extent of artifacts from the heartbeat were compared. In addition, the peak signal-to-noise ratio (PSNR) was measured from differential images between mask image groups before the injection of the contrast medium to investigate optimal mask images. The mean time required for automatic remasking was 4.7 s/series, a significant reduction in processing time compared with the mean of 266 s/series for conventional manual processing. A visual comparison of the different approaches showed virtually no misregistration artifacts from the heartbeat in manual or automatic remasking according to cardiac phase. The results from measuring the PSNR for differential images between mask image groups also showed that smaller cardiac phase difference and time difference between two images ensure higher PSNR (p < 0.01). Automatic remasking according to the cardiac phase was fast and easy to implement and reduced misregistration artifacts from heartbeat.

Usability of unbiased nonlocal means for de-noising intraoperative magnetic resonance images in neurosurgery.

PURPOSE: Intraoperative magnetic resonance imaging (iMRI) is a powerful tool that allows real-time image-guided excision of brain tumors. However, low magnetic field iMRI devices may produce low-quality images due to nonideal imaging conditions in the operating room and additional noise of unknown origin. The purpose of this study was to evaluate a three-dimensional unbiased nonlocal means filter for iMRI (UNLM-i) that we developed in order to enhance image quality and increase the diagnostic value of iMRI. METHODS: We first evaluated the effect of UNLM by assessing the modulation transfer function (MTF) and Weiner spectrum (WS) of UNLM in simulated imaging. We then tested the diagnostic value of UNLM-i de-noising by applying it to a series of randomly chosen iMR images that were assessed by 4 neurosurgeons and 4 radiological technologists using a 5-point rating scale to compare 13 parameters, including tumor visibility, edema, and sulci, before and after de-noising. RESULTS: Unbiased nonlocal means provided better MTF in comparison with other filters, and the WS for UNLM de-noising was reduced for all spatial frequencies. Postprocessing UNLM-i allowed de-noising with preserved edges and >twofold improvement in the signal-to-noise ratio without extending the MRI scanning time (p< 0.001) . The diagnostic value of UNLM-i de-noising was rated as "superior" or "better" in >80 % of cases in terms of contrast between white and gray matter and visibility of sulci, tumor, and edema (p< 0.001). CONCLUSIONS: Unbiased nonlocal means filter for iMRI de-noising proved very useful for image quality enhancement and assistance in the interpretation of iMR images.

Adapting non-local means of de-noising in intraoperative magnetic resonance imaging for brain tumor surgery.

In image-guided brain tumor surgery, intraoperative magnetic resonance imaging (iMRI) is a powerful tool for updating navigational information after brain shift, controlling the resection of brain tumors, and evaluating intraoperative complications. Low-field iMRI scans occasionally generate a lot of noise, the reason for which is yet to be determined. This noise adversely affects the neurosurgeons' interpretations. In this study, in order to improve the image quality of iMR images, we optimized and adapted an unbiased non-local means (UNLM) filter to iMR images. This noise appears to occur at a specific frequency-encoding band. In order to adapt the UNLM filter to the noise, we improved the UNLM, so that de-noising can be performed at different noise levels that occur at different frequency-encoding bands. As a result, clinical iMR images can be de-noised adequately while preserving crucial information, such as edges. The UNLM filter preserved the edges more clearly than did other classical filters attached to an anisotropic diffusion filter. In addition, UNLM de-noising can improve the signal-to-noise ratio of clinical iMR images by more than 2 times (p < 0.01). Although the computational time of the UNLM processing is very long, post-processing of UNLM filter images, for which the parameters were optimized, can be performed during other MRI scans. Therefore, The UNLM filter was more effective than increasing the number of signal averages. The iMR image quality was improved without extension of the MR scanning time. UNLM de-noising in post-processing is expected to improve the diagnosability of low-field iMR images.