Functional image-guided surgery of intracranial tumors located in or near the sensorimotor cortex.

OBJECT: The purpose of this study was to evaluate the efficacy of noninvasive preoperative functional imaging data used in an interactive fashion in the operating room. The authors describe a method of registering preoperative functional magnetic resonance (fMR) imaging localization of sensorimotor cortex with a frameless stereotactic surgical navigation device. METHODS: The day before surgery, patients underwent blood oxygen level-dependent fMR imaging while performing a finger-tapping motor paradigm. Immediately afterward an anatomical stereotactic MR image was acquired. Raw fMR imaging data were analyzed offline at a separate workstation, and the resulting functional maps were registered to a high-resolution anatomical scan. The fused functional-anatomical images were then downloaded onto a surgical navigation computer via an ethernet connection. At surgery, the brain was exposed in the standard fashion, and the sensorimotor cortex was identified by direct cortical stimulation, the use of somatosensory evoked potentials, or both. This localization was then compared with that predicted by the registered fMR study. Thirteen procedures were performed in 12 patients. The mean registration error was 2.2 mm. The predicted location of motor and/or sensory cortex matched that found on intraoperative mapping in all 12 patients tested. Maximal tumor resection was accomplished in each case and no new permanent neurological deficits resulted. CONCLUSIONS: Compared with conventional brain mapping techniques, fMR image-guided surgery may allow for smaller brain exposures, localization of the language cortex with the patient under general anesthesia, and the mapping of multiple functional sites. The scanning equipment used in this method may be more readily available than for other functional imaging techniques such as positron emission tomography or magnetoencephalography.

Lung volumes before and after lung volume reduction surgery: quantitative CT analysis.

The volume and severity of pulmonary emphysema in individual lungs were measured by means of quantitative computed tomography (CT) studies in 28 patients (14 women, 14 men, median age 65 yr) who underwent either bilateral (n = 15) or unilateral (n = 13) lung volume reduction surgery (LVRS). Spirometric, total body plethysmographic, and CT data (at TLC and RV) were correlated before and after LVRS. Lung volumes determined by CT correlated well with volumes obtained by total body plethysmography (p < 0.0001). For individual lungs after LVRS, CT-derived mean lung capacity decreased 13% and residual volume 20% (p < 0.00001 for each), while mean total functional lung volume (TFLV, defined as the volume of lung with CT attenuation greater than -910 Hounsfield units) increased 9% (p < 0.01), and the mean ratio of the air space to tissue space volume (V(AS)/V(TS)) decreased more at RV (23%) than at TLC (14%) (p < 0.0005 for each). In contrast, unilateral LVRS did not affect exhalation from the unoperated lung (2% reduction in RV, p = NS). The magnitude of the postoperative response (CT-derived TLC, RV, TFLV, V(AS)/V(TS)) of each operated lung was comparable for unilateral and bilateral LVRS. Thus, a lung's response to LVRS was independent from that of the contralateral lung. Moreover, postoperative alterations in TFLV and FEV1 correlated significantly (r = 0.80, p < 0.0001), which suggests that the expansion of functioning tissue may contribute to the mechanism by which LVRS palliates airway obstruction.

Intraoperative functional MRI using a real-time neurosurgical navigation system.

A 42-year-old-man had focal left hand motor seizures. MR studies demonstrated a right posterior frontal brain tumor. Functional MRI was performed, localizing the motor cortex posterior to the lesion. The functional images were integrated with a neurosurgical navigation computer. A real-time intraoperative display of the anatomic and functional images was produced, registered to a neurosurgical probe. Excellent correlation was demonstrated between the functional maps and invasive electrophysiologic mapping performed at the time of craniotomy.

Functional MRI-guided surgery of intracranial tumors.

We describe a technique of registering functional magnetic resonance image (fMRI) data sets with the anatomical images used for intraoperative navigation. Three patients with supratentorial tumors were studied, 2 with astrocytomas and 1 with a meningioma. The day before surgery, a functional blood-oxygen-level-dependent MR scan was done. During the same imaging session, an anatomic stereotactic MR scan was performed. The raw data were reconstructed off line on a workstation, and the fMR maps were then registered to the anatomic studies using image header information. Fused images were then transferred to the navigational workstation over an Ethernet connection. At surgery, brain exposure was performed in a standard manner. In each case, the registered scans accurately localized the motor and/or sensory cortex. No new permanent neurological deficits resulted from surgery. The potential advantages of fMRI-guided surgery include: (1) smaller brain exposures; (2) localization of language cortex under general anesthesia; (3) mapping of multiple functional sites, with decreased time and increased case of surgery.

Superconducting receiver coils for sodium magnetic resonance imaging.

We present the results from sodium magnetic resonance imaging (MRI) experiments using high-temperature superconducting (HTS) receiver coils. Sodium imaging has been shown to have great potential for the assessment of cell integrity but suffers from a substantially lower signal-to-noise ratio (SNR) than that of a hydrogen imaging. The use of an HTS receiver coil was found to significantly increase the SNR relative to an equivalent copper receiver coil at room temperature. The SNR gains afforded by HTS coils can also be used to decrease the imaging time.

A prospective controlled study of magnetic resonance imaging of the brain in gay men and parenteral drug users with human immunodeficiency virus infection.

To detect the earliest structural changes in the brain in human immunodeficiency virus (HIV) infection, 118 gay men and 115 parenteral drug users enrolled in a study of the natural history of HIV infection underwent magnetic resonance imaging evaluations. Routine T2-weighted and heavily T2-weighted scans for quantification of brain water were obtained, blinded to HIV serostatus. Atrophy and foci of increased signal did not correlate with any medical, immunologic, neurologic, or neuropsychologic parameters in the group as a whole, or in the gay men or parenteral drug user subgroups. Three subjects had progressive multifocal leukoencephalopathy and one had central nervous system lymphoma. In a subgroup in whom intracranial water percent was calculated, correlations were found with CD4 counts and CD4/CD8 ratios. We conclude that standard magnetic resonance imaging of the brain does not differentiate asymptomatic and mildly symptomatic HIV-positive individuals from HIV-negative individuals, regardless of risk group. However, intracranial water percent may distinguish HIV-positive from HIV-negative individuals because it correlates with raw CD4 counts and CD4/CD8 ratios.

New spatial localization method using pulsed high-order field gradients (SHOT: Selection with High-Order gradienT).

A new spatial localization method using an additional set of high-order magnetic field gradients is described. The method uses a nonlinear part of high-order magnetic field gradient patterns which allows us to select a volume in conjunction with the selective radiofrequency (RF) pulse. Unlike the other existing volume selection methods such as ISIS or SPARS, the proposed selection method requires only one RF-gradient pulse pair to select a volume in two directions. The center of the selected volume can be moved to any arbitrary location within the body by the addition of precalculated lower order gradients which are simultaneously pulsed with the high-order gradient. The method also has the potential for localized spectroscopy from the FID signal which can be realized by using oscillating second- and first-order gradients for 3D selection with a single RF pulse. By using the proposed localization method, it is possible to design more flexible pulse sequences, e.g., the shorter echo-time spectroscopic pulse sequence. We have designed and constructed a six-loop r2 (or x2 + y2) gradient coil for initial application. By simultaneously applying this second-order gradient and proper x, y, and/or z gradients, 2D selections were achieved in arbitrarily selected positions in conjunction with a single selective RF pulse. Phantom and animal experiments have been performed and the results appear promising, especially in areas of NMR spectroscopic imaging applications where spatial localization is essential.

An optimized multislice acquisition sequence for the inversion-recovery MR imaging.

An optimized multislice data acquisition scheme for inversion-recovery MR imaging is proposed and experimental results are presented. In this new scheme, instead of forming a set of multislice inversion-recovery sequences in series for a given phase encoding step, 180 degrees inversion pulses corresponding to different slices are interwoven with the spin echo data acquisition sequence in an optimal way depending on the desired inversion-recovery time. For example, between the 180 degrees inversion RF pulse and the spin-echo imaging sequence, a number of imaging and inversion sequences are inserted with different slice combinations, i.e., long inversion-recovery time is effectively utilized for the other slice pre-inversion and data acquisition. With the optimized sequence, imaging time has been reduced by as much as a factor of four compared with the existing methods.

Application of region of interest definition to quadtree-based compression of CT images.

A quadtree-based data compression algorithm can provide different levels of compression within and outside of regions of interest (ROIs). The current study shows whether ROI compression can provide greater compression or diagnostic accuracy than uniform quadtree compression. In 75 single CT images from 75 consecutive abdominal examinations, 43 abnormalities were identified and surrounded by ROIs. Three radiologists interpreted the images following (1) 50:1 compression of the entire image; (2) ROI compression at five decreasing compression ratios (with 50:1 compression outside the ROI); and (3) reversible (lossless) compression of the entire image. Reversible compression (compression ratio 3:1) yielded a sensitivity of 96%. ROI compression of 15:1 was achieved with no loss of sensitivity; ROI compression of 28:1 yielded a sensitivity of 91% (not significantly different). At any given compression ratio, diagnostic sensitivity was greater with ROI compression than with uniform quadtree compression. For purposes of image archiving, quadtree-based ROI compression is superior to uniform compression of CT images.

Radio frequency field intensity mapping using a composite spin-echo sequence.

A novel radio frequency (RF) field intensity mapping or imaging method using a composite NMR spin-echo sequence is proposed. A composite spin-echo RF pulse with 90 degrees y-180 degrees x-90 degrees y sequence makes phase change in the final image depending on the RF field intensity on the object. The resultant phase change or phase map can be used to obtain the actual RF flip-angle map for a given condition which includes the status of tuning and RF inhomogeneity, etc. Bloch equation has been solved numerically to obtain the effects of the RF field intensity as well as the main magnetic field inhomogeneity and the results are used for the mapping (imaging) of the RF field intensity. Phantom studies have been performed using a 1.5 Tesla whole body MRI system and the results are presented.

Quadtree-based data compression of abdominal CT images.

An ROC study was performed to evaluate the effect of quadtree-based data compression on the diagnostic yield of CT images. Seventy images were selected from a series of consecutive abdominal/pelvic CT scans. Following the application of quadtree-based compression, all images were reviewed independently by three radiologists. The images were analyzed at six decreasing irreversible compression ratios (30.6:1 to 7.4:1), as well as after reversible compression (2.9:1). ROC curves reveal a gradual decrease in clinical accuracy with increasing compression ratios. At a compression ratio of 7.4:1, sensitivity for all major abnormalities was 99% with a specificity of 93%. As the compression ratio was increased to 30.6:1, sensitivity and specificity dropped to 75% and 83% respectively. Execution times for compression and decompression of the CT images with a PC-AT based digital radiography system varied from 24.7 to 18.5 seconds and from 16.2 to 5.1 seconds respectively, decreasing with higher levels of compression. We conclude that quadtree-based compression of abdominal CT images may have practical applications for a PC based digital radiography system.

In vivo magnetic resonance imaging of sodium in the human body.

Sodium magnetic resonance imaging of the human body in vivo has been tried and its clinical application is considered. A short T2 imaging algorithm with total volume excitation and a specialized RF coil focused to the region of interest have been adopted to improve the signal-to-noise ratio. Using a 1.5-T human body imaging system, several important organs including heart, liver, gallbladder, kidney, and spine have been examined to demonstrate their sodium concentration in vivo.