Ulrich Raff

Parkinson's disease: a novel MRI method for determining structural changes in the substantia nigra

OBJECTIVES To use MRI in a novel way to image and quantify the changes occurring in the substantia nigra in Parkinsons disease. METHODS Six patients with idiopathic Parkinsons disease were compared with six age matched control subjects. The subjects were imaged using a combination of pulse sequences hypothesised to be sensitive to cell loss. RESULTS The images showed patterns of change in patients with Parkinsons disease. Highly significant differences between the patients and control population were found (p<0.001). CONCLUSIONS This methodology suggests the possibility of detecting presymptomatic disease in those judged to be at risk, and also in confirming the diagnosis in patients with early disease. Furthermore, the technique seems to hold promise as a means for staging the disease, and possibly differentiating other forms of parkinsonism.

Automated discrimination and quantification of idiopathic pulmonary fibrosis from normal lung parenchyma using generalized fractal dimensions in high-resolution computed tomography images

RATIONALE AND OBJECTIVES We computed generalized fractal dimensions for high-resolution computed tomography (HRCT) images to investigate their value in the discrimination and quantification of idiopathic pulmonary fibrosis (IPF) from normal lung parenchyma. METHODS A probability distribution that was based on the pixel value in each image was used to compute capacity, information, and higher fractal dimensions for a series of 52 HRCT slices obtained from four patients. Qualitative classification of normal, mild, moderate, and severe IPF cases was achieved by computing the following parameter: DD = D0 - 2D1 + D2, where D0, D1, and D2 represents the capacity, information, and pair correlation dimensions, respectively. A multiple linear regression analysis using morphometric quantification for the set of 52 slices was tested for all possible combinations of the parameters D0, D1, D2, and D3. The generalizability of the model was tested by predicting the extent of IPF for each patient from a regression model computed with the remaining slices in the database. RESULTS The best regression results were obtained using the independent parameters D1 and D2 to quantify the extent of diseased lung parenchyma. The technique was tested with 48 slices from 12 new patients. The results indicated that the extent of IPF could be predicted within the confidence limits given by the regression analysis. CONCLUSION The extent of IPF can be predicted well within the 90% confidence interval given by the model. The width of the confidence interval decreases as the number of slices used in the linear regression model increases. This operator-independent quantitative technique may be useful in the follow-up of patients with IPF.

Detection of Parkinson's disease by MRI: Spin-lattice distribution imaging†

We have developed an advanced MRI technique for detecting Parkinsons Disease (PD) which depends on an image constructed as a ratio of images from two inversion recovery sequences (one generating a white matter suppressed image, the other a gray matter suppressed image). This technique was designed to be exceptionally sensitive to the spin‐lattice relaxation time T1. It was refined with the introduction of segmentation analysis and given the acronym SIRRIM (Segmented Inversion Recovery Ratio Imaging). Our objectives are, first, to reinvestigate the sensitivity of MRI with new subjects and second, to investigate whether a new form of analysis, using the gray level distribution of signal in the image, may prove more sensitive than SIRRIM. For each subject, a ratio image was constructed (WMS/GMS) and the substantia nigra segmented out to be displayed as an isolated structure. From the segmented image a measure of disease severity, the Radiological Index (RI), was calculated for each subject. Since the pixel value in the ratio image is a strong function of the local T1 relaxation time, the distribution of pixel values gives the distribution of spin‐lattice relaxation times. A refinement in the analysis is introduced, the Spin‐Lattice Distribution Index (SI), which is an automated measure of MRI signal in the Substantia Nigra pars compacta (SNC). Both RI and SI were calculated for each of 24 subjects, 12 patients and 12 controls. The SI may further improve the separation of patient and control groups, and may therefore be more sensitive than the RI. Unlike the RI it is completely automatic and circumvents two of the limitations of the RI. The work is consistent with the proposition that MRI, when properly configured, is a highly sensitive marker for PD.

Quantitation of grey matter, white matter, and cerebrospinal fluid from spin-echo magnetic resonance images using an artificial neural network technique

An operator independent technique has been developed to quantitate the volume of white matter (WM), grey matter (GM), and cerebrospinal fluid (CSF) using spin-echo magnetic resonance images. Using skull stripped spin-echo images, CSF was removed using an automated thresholding technique. The bimodal histogram of the remaining images was used to train a perceptron and a single hidden layer neural network to output the percentage of GM and WM in the image. The output values were compared with those of a semiautomated technique employing a least square fitting technique [graduated nonconvexity algorithm (GNC)] applied to the bimodal histogram. This semiautomated technique allowed for intervention by the radiologist. Fourteen normal volunteers with eight contiguous slices each were analyzed. The individual percentages of WM, GM, and CSF of 40 slices from five subjects not used in the training set as well as the total percentages of GM, WM, and CSF in each individual were predicted using the two artificial network architectures. GM, WM, and CSF percentages were predicted within 7% for individual slices while total percentages of WM, GM, and CSF were computed accurately with an absolute error of less than 5% when compared to the semiautomated technique involving a trained neuroradiologist.

Computer assessment of neurodegeneration in Parkinson disease using data fusion techniques with MR images

RATIONALE AND OBJECTIVES Recently developed MR imaging techniques using inversion recovery are a sensitive tool to identify and quantify morphologic changes in the substantia nigra due to neurodegeneration. Using a semi-automated computer segmentation technique to isolate the substantia nigra pars compacta (SN(c)), we propose a colored image fusion technique to visually assess the sites of damage in the SN(c) and integrate the information obtained from two implemented inversion-recovery sequences. PATIENTS AND METHODS Six patients and six age-matched control subjects were scanned using a combination of two MR imaging inversion-recovery (IR) pulse sequences. A subgroup of them was used to develop our technique. Images were blended together into a final (RGBA) image, where A stands for the alpha channel describing transparency. RESULTS Abnormalities in the SN(c) can be accurately assessed in location, shape, and variations of signal intensities within the segmented SN(c) by varying the transparency (alpha) channel of the color fusion image. Several previous findings such as the lateral-medial gradient of signal change and a ventral-dorsal broadening of the pars compacta are accompanied by an overall mild-to-severe heterogeneity of neurodegeneration patterns. CONCLUSION Color fusion techniques revealed subtle changes in the neurodegeneration of the substantia nigra in Parkinson disease, which can be helpful for an objective and hence effective visual assessment of disease progression.

Quantitation of T2 lesion load in patients with multiple sclerosis: A novel semiautomated segmentation technique*

RATIONALE AND OBJECTIVES The authors designed a segmentation technique that requires only minimal operator input at the initial and final supervision stages of segmentation and has computer-driven segmentation as the primary determinant of lesion boundaries. The technique was applied to compute total T2-hyperintense lesion volumes in patients with multiple sclerosis (MS). A semi-automated segmentation technique is presented and shown to have a test-retest reliability of <5%. MATERIALS AND METHODS The method used a single segmented section with MS lesions. A probabilistic neural net performed segmentation into four tissue classes after supervised training. This reference section was deconstructed into the entire set of possible 4 x 4-pixel subregions, which was used to segment all-brain sections in steps of 4 x 4-pixel, adjacent image blocks. Intra- and interimage variabilities were tested by using 3-mm-thick, T2-weighted, dual-echo, spin-echo MR images from five patients, each of whom was imaged twice on the same day. Five different reference sections and three temporally separated. training sessions involving the same reference section were used to test the segmentation technique. RESULTS The coefficient of variation ranged from 0.013 to 0.068 (mean +/- standard deviation, 0.037 +/- 0.039) for results from five different reference sections for each brain and from 0.007 to 0.037 (mean, 0.027 +/- 0.021) for brains segmented with the same reference section on three temporally separated occasions. Test-retest (intra-imaging) reliability did not exceed 5% (except for a small lesion load of 1 cm3 in one patient). Interimaging differences were approximately 10%. CONCLUSION The segmentation technique yielded intra-imaging variabilities (2%-3%, except for very small MS lesion loads) that compare favorably with previously published results. New repositioning techniques that minimize imaging-repeat imaging variability could make this approach attractive for resolving MS lesion detection problems.

Computerized left ventricular pressure–volume relationships (pV-loops) using disposable angiographic tip transducer pigtail catheters

Left ventricular pressure–volume relationships expressed as pV loops could yield important hemodynamic information in the cardiac catheterization laboratory. Many clinical situations might benefit from a quantitative assessment of left ventricular function. Potential applications of pV loops include the assessment of vasoactive and inotropic drugs, balloon valvuloplasty, coronary angioplasty, and surgical treatment of valvular heart disease. For many years the clinical use of pV loops has been hindered by logistical difficulties. The ability to merge on-line concurrent digital imaging data for computation of left ventricular volume and digital left ventricular pressure wave forms obtained from high fidelity tip-transducer angiocatheters has allowed us to develop a technique which can generate pV loops during cardiac catheterization procedures. The method offers an automated measurement of left ventricular volume independent of edge detection or an interactive technique for tracing endocardial borders by a trained operator. Illustrative case studies are included to demonstrate the potential of the method during ventricular angiographic procedures. Implementation and computational time requirements of the method are discussed. The concept and the value of pV loop generation to study left ventricular performance has been known for many years. Combining digital imaging and digital physiologic data obtained with disposable tip-transducer angiocatheters with modern networking technology, the technique can more easily be applied to catheterization procedures and could enhance invasive hemodynamic assessment of left ventricular function.

Spin-lattice distribution MRI maps nigral pathology in progressive supranuclear palsy (PSP) during life: a pilot study.

An MRI biomarker for Parkinsonism has long been sought, but almost all attempts at conventional field strengths have proved unsatisfactory, since patients and controls are not separated. The exception is Spin-Lattice Distribution MRI (SLD-MRI), a technique which detects changes in the substantia nigra (SN) due to changes in the spin-lattice relaxation time, T1. This easily separates patients with Parkinsons disease (PD) from control subjects at 1.5 Tesla, suggesting that it may be sensitive to presymptomatic disease. SLD-MRI demonstrates a topography of signal change within the SN which is the same as the known topography of pathological change, where the lateral portions of the nucleus are more affected than the medial. In a further step towards its validation, we apply SLD-MRI to a disease control, Progressive Supranuclear Palsy (PSP), the most common of the atypical forms of Parkinsonism. In PSP the topography of pathological change in the SN is reversed. We therefore hypothesized that PSP would show a topography of SLD-MRI signal change in the SN that is the reverse of PD (i.e. the medial portion is more affected than the lateral). All 7 patients showed such a topography of MR signal, and all patients were separated from control subjects. Although this is a step toward validation of SLD-MRI with respect to sensitivity and disease specificity, nevertheless we stress that this is a pilot project only. Validation will only be possible when comparing larger cohorts of PSP, PD and control subjects.

Automated determination of left ventricular volume curves from bi-plane digital angiography without explicit use of edge detection algorithms

Automated computation of left ventricular (LV) global and regional function using contrast angiography has not yet become a routine procedure with the advent of digital cardiac imaging systems. We describe a new technique to compute LV volume curves which does not require the use of manual or semi-automated detection of endocardial borders and provides on-line implementation of volumetric curves and computation of pressure volume loops during catheterization. The approach uses the concepts of variableentropy (orinformation) of left ventricular images throughout the cardiac cycle. LV volume curves are computed with an interpolation scheme using those LV volume curves of a patient data base which are associated with the closest variation in entropy in the RAO projection to the analyzed patient data according to a simple metric. Computed LV volume curves were correlated with those obtained with manual tracing. Left ventricular ejection fraction (LVEF), time to end systole (TES) and angiographic cardiac output (CO) were compared to those obtained with the manual method. Results using a data base of 365 patients revealed excellent correlation (r=0.97) between manually derived volume curves and volume curves computed with the automated technique within a large range of LVEFs. In 87% of all cases the computed LVEF values were found within ± 10% of the value obtained with thegold standard method. The systolic phase of the volume curves showed that 81% of all cases had the same accuracy. The TES showed much more variation due to undersampling of the cardiac cycle in time (r=0.71).

Computation of left ventricular volume curves from gated blood pool studies without explicit use of edge detection algorithms: concise communication

A new technique has been developed to compute left ventricular (LV) time activity curves from gated blood pool (GBP) studies without the use of manual, semiautomated or fully automated edge detection algorithms. The method utilizes the correlation of entropy calculated from the counts of a fixed region of interest covering the left ventricle during a cardiac cycle to compute the LV volume curve for a new patient. The new LV volume curve is obtained through interpolation of those volume curves of a data base which are associated with the closest variations in normalized entropy to the new one. The computed LV time activity curves agree with those obtained from manual or fully automated outlines of the left ventricle within 9 percent for the selected set of 67 patients demonstrating the potential of the method. The accuracy of calculated LV volume curves can be improved theoretically to any degree by increasing the number of cases in the data base of known statistical feature vectors associated with the LV images and LV volume curves. The new method for computation of LV curves is very efficient and robust when compared to traditional techniques.