Faculty Bibliography

We have developed an interactive graphic editor to define subregions within a volume of medical images arranged in serial sections. The editing methods include tracing and automatic growing of connected components defined by the gray level range. The editor performs a statistical analysis of the signal contained in each subvolume and is used in studies of magnetic resonance (MR) signal in medical images. The graphic editor is also used for creating 3D views from MR images based on the new volume-rendering algorithm. The editor transforms the volume of images by remapping their gray levels and by multiplanar cuts. The ability to suppress regions having the same signal intensity as the region of interest is important in generating an unobstructed view of the anatomical structures

The authors developed new techniques for three-dimensional display of magnetic resonance (MR) images that preserve soft-tissue definition, are fully automatic, and work with routinely used section thicknesses. MR images are segmented, selectively enhanced, and displayed by means of a volumetric rendering algorithm. These techniques were used to illustrate normal anatomy of the brain, knee, and liver. Three-dimensional rendering of balanced spin-echo images shows the ventricles and extracerebral veins and of T1-weighted images, the sulci and gyri. The large hepatic and portal vessels can be seen with these enhancement techniques. Three-dimensional views of the knee reveal articular surfaces of the tibia and clearly depict menisci and posterior and anterior cruciate ligaments. These techniques make it possible to image multiple soft tissues simultaneously while preserving the detail contained in the original images. Three-dimensional presentation of complex, overlapping anatomic regions is helpful in surgical planning and should lead to improved diagnosis

We measured cerebrovascular reactivity to carbon dioxide in the cerebral cortex and the subcortical white matter of 12 healthy adult volunteers (four young subjects aged 21-24, four middle-aged subjects aged 34-40, and four elderly subjects aged 62-85 years). Blood flow was computed from the concentration history of xenon-133 in the volume of interest measured with an ultrapure germanium detector array. End-tidal PaCO2 ranged from 35.4 to 42.6 mm Hg. The mean +/- SD baseline blood flows in the cerebral cortex were 60 +/- 7, 51 +/- 9, and 33 +/- 4 ml/100 cm3/min in the young, the middle-aged, and the elderly subjects, respectively; the corresponding subcortical white matter baseline blood flows were 21 +/- 1, 22 +/- 3, and 16 +/- 5 ml/100 cm3/min. Mean +/- SD cerebrovascular reactivities to carbon dioxide in the cerebral cortex were 2.03 +/- 0.58, 1.36 +/- 0.41, and 0.72 +/- 0.19 ml/100 cm3/min/mm Hg PaCO2 for the young, the middle-aged, and the elderly subjects, respectively; the corresponding reactivities in the subcortical white matter were 0.69 +/- 0.11, 0.59 +/- 0.17, and 0.36 +/- 0.41 ml/100 cm3/min/mm Hg PaCO2. Blood flow and cerebrovascular reactivity in the cerebral cortex of the young subjects were significantly higher than those for white matter and significantly higher than those in the elderly subjects (p less than 0.001). Age vs. blood flow (for the cortex) and age vs. cerebrovascular reactivity (for both cortical gray and subcortical white matter) also showed significant linear correlation (p less than 0.05). However, the age-related changes in white matter blood flow and cerebrovascular reactivity were slow, and the differences among the age groups were not statistically significant

Volume reconstruction algorithms that provide a three-dimensional (3D) rendering of a tomographic study are now available but not directly applicable to magnetic resonance (MR) studies. The problem with images in such studies is that the gray ranges of tissues overlap. To overcome it, we have developed an interactive graphics editor (IGE) that allows its user to define subregions within the volume of MR images which comprises such a study. The editor performs a statistical analysis of the volume contained in each subregion. With this information the user can enhance individual subregions with such techniques as user-defined remapping of gray levels, the addition of color, the addition of transparency, and spatial filtering. The editor is implemented on a Pixar Image Computer which uses as a host a Sun 3/ISO computer. It has been used by us to create from MR studies three-dimensional images of the brain and knee

A new algorithm for computing cerebral blood flow is shown to overcome compartmental slippage and unstability problems associated with the conventional bicompartmental analysis. The tracer clearance curve is decomposed into a nonnegative linear combination of predetermined flow components. A weighted average of flows above (below) a fixed threshold yields the gray (white) matter flow. The accuracy and the stability of the new algorithm are analyzed by Monte-Carlo simulations, determining the effect of factors such as random error in tracer concentration, gray-white flow difference, tissue content of gray matter, and end-fit time. While the new algorithm requires 50-100% more CPU time and memory space than the bicompartmental method, its accuracy and stability is superior, especially as the conditions of the measurement deteriorate. The results suggest that more than twofold error reduction in measuring the blood flow in pathological brain tissue is possible

Distribution of cerebral blood flow was measured with an array of 200 ultra-pure germanium radiation detectors and 133-Xe by inhalation. The array 'sees' the head as a composite of different subvolumes and enables measurement of the concentration history of tracer every 1-10 sec in each subvolume simultaneously. Subvolume mean flows, (fm), and partition coefficients, lambda m, are derived by compartmental analysis of tissue concentration washout curves. Errors from 'cross talk,' scalp radiation, 'look through,' and assumed partition coefficients are eliminated. Average fm adjusted for 40 mm Hg PACO2 in 14 cortical subvolumes (7 right, 7 left) of four normal 21-24 year old controls ranged from 50 to 60 ml/100 cc tissue/min, and lambda m ranged from 0.97 to 1.14. Average fm and lambda m in white matter was 24 ml/100 cc/min and 1.42 - 1.14 respectively. During CO2 inhalation, right and left hemispheric fm increased 6.4% and 5.7%/mm Hg respectively, whereas white matter fm increased 2.2% and 3.4% mm Hg respectively. There was no systematic difference between front and back or dominant vs non-dominant sides. Three 73-84 year old controls had reduced fm and CO2 reactivity in all subvolumes, lambda m was in the same range as in younger controls. Two patients with intracranial cerebrovascular disease showed excellent localization of ischemic subvolumes. One patient with asymptomatic unilateral 98% stenosis of the internal carotid artery had a similar distribution of blood flow in both hemispheres