W.I.D. Rae

Diffusion‐Weighted Magnetic Resonance Imaging of Spinal Infection and Malignancy

Background and Purpose. Pyogenic and tuberculous spondylitis can mimic malignancy. The purpose of this study was to deter mine the efficacy of diffusion‐weighted magnetic resonance imaging in differentiating spinal infection and malignancy. Methods. Fifty‐one consecutive patients with suspected spinal infection or malignancy were enrolled in the study. Apparent diffusion coefficients (ADCs) of paraspinal soft tissue mass and normal and abnormal vertebral bone marrow were determined on the diffusion‐weighted magnetic resonance images of the spine. The mean ADCs of normal and abnormal vertebral bodies in patients with confirmed infection or malignancy were compared using nonparametric tests. Results. ADCs of 69 tuberculous, 9 pyogenic, and 50 malignant vertebral marrow lesions were significantly higher than ADCs of normal marrow. ADCs of malignant bone marrow and 5 paraspinal soft tissue lesions were significantly lower than tuberculosis and pyogenic infection. There was no significant difference between the ADCs of 44 adult and 25 pediatric tuberculous bone lesions or between tuberculosis and pyogenic infection. Using the cutoff ADC of 1.02 × 10−3mm2/s for bone marrow, the sensitivity, specificity, and accuracy were 60.26%, 66.00%, and 62.50%, respectively, for distinguishing infection from malignancy. The sensitivity, specificity, and accuracy increased to 94.12%, 82.35%, and 90.20%, respectively, when the ADCs of associated soft tissue lesions were higher than 1.17 × 10−3mm2/s. Conclusions. Diffusion‐weighted magnetic resonance imaging has limited usefulness for differentiating spinal infection and malignancy.

Neurological, visual, and MRI brain scan findings in 87 South African patients with HIV-associated cryptococcal meningoencephalitis

BACKGROUND HIV-associated cryptococcal meningoencephalitis (CM) is a leading cause of adult meningitis in sub-Saharan Africa. Neuroradiological data is however limited to case reports and small case series from developed countries and/or immunocompetent patients. METHODS Eighty seven patients aged ≥18 hospitalized with a first episode of CM had magnetic resonance (MRI) imaging during the first two weeks of admission. A subset of eleven patients had follow-up scans approximately one month from their initial MRI scan. All had prospectively-recorded detailed neurological and visual examinations. RESULTS An abnormal finding on neurological examination was detected in 33 (39%) patients. 38 (48%) patients experienced some visual loss. Neuroradiological lesions presumed to be cryptococcosis-related, as defined by the presence of dilated Virchow Robin spaces, pseudocysts or cryptococcomas, enhancing nodules, hydrocephalus, meningitis, focal perilesional oedema and infarcts, were detected in 55 (63%) patients. MRI findings suggestive of a second diagnosis were found in 18 (21%) patients. Visual loss was associated with the presence of cryptococcal-related lesions (p = 0.02). Blindness was associated with raised intracranial pressure (ICP) (p = 0.02). Of eleven patients with paired scans, brain swelling was identified on the initial scan in only one patient. CONCLUSION The majority of patients had MRI brain scan abnormalities presumed secondary to CM. Dilated Virchow Robin spaces were the commonest neuroradiological lesion. Visual loss was associated with the degree of cerebral involvement as reflected by the presence of MRI abnormalities. Blindness was associated with the presence of raised ICP. Initial generalised brain swelling does not appear to be common, but further studies with paired scans are needed.

Identification of the breast edge using areas enclosed by iso-intensity contours

The segmentation of a mammogram into background and breast is a crucial first step in the computer aided diagnosis of mammograms that has the advantage of simplifying further processing of the image (by eliminating the background) and also provides a reference for the alignment of views when two views are being compared. A novel method of segmenting the breast from the background by analysing the area enclosed by iso-intensity contours is presented. Results are evaluated by comparison with manual borders drawn by three radiologists for a set of 25 mammograms. The effect of different pre-processing methods, on the accuracy of automated borders, is investigated. The best methods yielded average root-mean-square differences between the manual and automated iso-intensity borders of 3.0+/-0.3 mm for the image set containing clear breast edges and 4.8+/-0.5 mm for the image set containing indistinct breast edges compared to 5.1+/-0.8 and 7.9+/-0.9 mm for the two data sets with no pre-processing. Results are also compared to those obtained from global thresholding.

Early Clinical and Subclinical Visual Evoked Potential and Humphrey's Visual Field Defects in Cryptococcal Meningitis

Cryptococcal induced visual loss is a devastating complication in survivors of cryptococcal meningitis (CM). Early detection is paramount in prevention and treatment. Subclinical optic nerve dysfunction in CM has not hitherto been investigated by electrophysiological means. We undertook a prospective study on 90 HIV sero-positive patients with culture confirmed CM. Seventy-four patients underwent visual evoked potential (VEP) testing and 47 patients underwent Humphreys visual field (HVF) testing. Decreased best corrected visual acuity (BCVA) was detected in 46.5% of patients. VEP was abnormal in 51/74 (68.9%) right eyes and 50/74 (67.6%) left eyes. VEP P100 latency was the main abnormality with mean latency values of 118.9 (±16.5) ms and 119.8 (±15.7) ms for the right and left eyes respectively, mildly prolonged when compared to our laboratory references of 104 (±10) ms (p<0.001). Subclinical VEP abnormality was detected in 56.5% of normal eyes and constituted mostly latency abnormality. VEP amplitude was also significantly reduced in this cohort but minimally so in the visually unimpaired. HVF was abnormal in 36/47 (76.6%) right eyes and 32/45 (71.1%) left eyes. The predominant field defect was peripheral constriction with an enlarged blind spot suggesting the greater impact by raised intracranial pressure over that of optic neuritis. Whether this was due to papilloedema or a compartment syndrome is open to further investigation. Subclinical HVF abnormalities were minimal and therefore a poor screening test for early optic nerve dysfunction. However, early optic nerve dysfunction can be detected by testing of VEP P100 latency, which may precede the onset of visual loss in CM.

New Insights into the Pathogenesis of Cryptococcal Induced Visual Loss Using Diffusion-Weighted Imaging of the Optic Nerve

Despite a successful antiretroviral drug rollout in South Africa, cryptococcal induced visual loss continues to be a major complication of cryptococcal meningitis. Preventive measures are lacking due to poor understanding of its pathogenesis. Optic nerve diffusion has shown some promise in investigating inflammatory and axonal changes of the optic nerve in vivo. The lack of post-contrast enhancement on T1 magnetic resonance imaging (MRI) and signal changes on T2 MRI with slightly increased apparent diffusion coefficient and reduced fractional anisotropy on diffusion imaging suggest that pressure-related effects are the dominant mechanism over inflammation and that there is early untested evidence of a compartment syndrome rather than papilloedema as the main contributor.

The Optic Nerve Compartment Syndrome in Cryptococcus-Induced Visual Loss

Abstract Visual loss in cryptococcal meningitis has been postulated to be due to papilloedema and/or optic neuritis. A 28-year-old human immunodeficiency virus (HIV)-positive female presented with visual loss, swollen optic discs, and elevated intracranial pressure due to cryptococcal meningitis. Computerised tomographic cisternography and T2-weighted magnetic resonance imaging showed occlusion of the peri-optic subarachnoid space and its reopening after serial lumbar punctures. Presumably lowering of the intracranial pressure resulted in equalisation of pressure across the pressure gradient created by the fungal block. This case supports a third mechanism of visual loss in cryptococcal meningitis, namely, an optic nerve compartment syndrome, that seems more plausible as the principal mechanism.

Visual loss in HIV-associated cryptococcal meningitis: A case series and review of the mechanisms involved

Permanent visual loss is a devastating yet preventable complication of cryptococcal meningitis. Early and aggressive management of cerebrospinal fluid pressure in conjunction with antifungal therapy is required. Historically, the mechanisms of visual loss in cryptococcal meningitis have included optic neuritis and papilloedema. Hence, the basis of visual loss therapy has been steroid therapy and intracranial pressure lowering without clear guidelines. With the use of high-resolution magnetic resonance imaging of the optic nerve, an additional mechanism has emerged, namely an optic nerve sheath compartment syndrome (ONSCS) caused by severely elevated intracranial pressure and fungal loading in the peri-optic space. An improved understanding of these mechanisms and recognition of the important role played by raised intracranial pressure allows for more targeted treatment measures and better outcomes. In the present case series of 90 HIV co-infected patients with cryptococcal meningitis, we present the clinical and electrophysiological manifestations of Cryptococcus-induced visual loss and review the mechanisms involved.

Interactive breast mass segmentation using a convex active contour model with optimal threshold values

INTRODUCTION A convex active contour model requires a predefined threshold value to determine the global solution for the best contour to use when doing mass segmentation. Fixed thresholds or manual tuning of threshold values for optimum mass boundary delineation are impracticable. A proposed method is presented to determine an optimized mass-specific threshold value for the convex active contour derived from the probability matrix of the mass with the particle swarm optimization method. We compared our results with the Chan-Vese segmentation and a published global segmentation model on masses detected on direct digital mammograms. METHODS AND MATERIALS The regional term of the convex active contour model maximizes the posterior partitioning probability for binary segmentation. Suppose the probability matrix is binary thresholded using the particle swarm optimization to obtain a value T1, we define the optimal threshold value for the global minimizer of the convex active contour as the mean intensity of all pixels whose probabilities are greater than T1. RESULTS The mean Jaccard similarity indices were 0.89±0.07 for the proposed/Chan-Vese method and 0.88±0.06 for the proposed/published segmentation model. The mean Euclidean distance between Fourier descriptors of the segmented areas was 0.05±0.03 for the proposed/Chan-Vese method and 0.06±0.04 for the proposed/published segmentation model. CONCLUSIONS This efficient method avoids problems of initial level set contour placement and contour re-initialization. Moreover, optimum segmentation results are realized for all masses improving on the fixed threshold value of 0.5 proposed elsewhere.

Dependence of Shape-Based Descriptors and Mass Segmentation Areas on Initial Contour Placement Using the Chan-Vese Method on Digital Mammograms

Variation in signal intensity within mass lesions and missing boundary information are intensity inhomogeneities inherent in digital mammograms. These inhomogeneities render the performance of a deformable contour susceptible to the location of its initial position and may lead to poor segmentation results for these images. We investigate the dependence of shape-based descriptors and mass segmentation areas on initial contour placement with the Chan-Vese segmentation method and compare these results to the active contours with selective local or global segmentation model. For each mass lesion, final contours were obtained by propagation of a proposed initial level set contour and by propagation of a manually drawn contour enclosing the region of interest. Differences in shape-based descriptors were quantified using absolute percentage differences, Euclidean distances, and Bland-Altman analysis. Segmented areas were evaluated with the area overlap measure. Differences were dependent upon the characteristics of the mass margins. Boundary moments presented large percentage differences. Pearson correlation analysis showed statistically significant correlations between shape-based descriptors from both initial locations. In conclusion, boundary moments of digital mass lesions are sensitive to the placement of initial level set contours while shape-based descriptors such as Fourier descriptors, shape convexity, and shape rectangularity exhibit a certain degree of robustness to changes in the location of the initial level set contours for both segmentation algorithms.

Status of Education and Training in Africa: Focus on South Africa

Medical Physics training and education in Africa continues to be limited in many ways. Information gathering is difficult in the region due to limited communication channels and relatively few formal contact people within the field. It appears only five countries in Africa have formal courses offered toward recognized Medical Physics qualifications. Much of the training is achieved via International Atomic Energy Agency (IAEA) sponsored training courses. These focus on specific training needs and are presented around the continent in member countries. Many qualified individuals currently in Africa obtained their qualifications in other countries.