Computed diffusion-weighted MRI for prostate cancer detection: the influence of the combinations of b-values

OBJECTIVE

To evaluate the influence of the combinations of b-values on computed diffusion-weighted images (cDWIs) for prostate cancer (PCa) detection at b = 2000 s mm(-2).

METHODS

Diffusion-weighted imaging (DWIs) for 31 patients with PCa (65.2 ± 7.1 years) were obtained pre-operatively at different b-values (0, 100, 500, 1000 and 2000 s mm(-2)) on a 3-T MRI. cDWIs at b = 2000 were generated by using six b-value combinations: 0-100 s mm(-2) (cDWI0-100); 0-500 s mm(-2) (cDWI0-500); 100-500 s mm(-2) (cDWI100-500); 0-1000 s mm(-2) (cDWI0-1000); 100-1000 s mm(-2) (cDWI100-1000); and 500-1000 s mm(-2) (cDWI500-1000). These cDWIs and measured DWIs with b = 2000 s mm(-2) (mDWI2000) were evaluated in this setting. To assess image quality for each DWI, contrast ratios (CRs) of cancerous and non-cancerous lesions were evaluated. To compare the detectability of PCa for each DWI, receiver operating characteristic analysis was used.

RESULTS

CRs of all cDWIs were significantly higher than those of mDWI2000 (p < 0.05). Areas under the curve of cDWI0-100 (0.62) and cDWI0-500 (0.65) were significantly smaller (p < 0.05) than those of others (cDWI100-500, 0.72; cDWI0-1000, 0.73; cDWI100-1000, 0.71; cDWI500-1000, 0.74; mDWI2000, 0.72).

CONCLUSION

The combinations of b-values influenced image quality and diagnostic ability of cDWIs for PCa detection. The combinations of b ≥ 100 and b ≥ 500 s mm(-2), as well as b = 0 and b = 1000 s mm(-2), were optimal in this study.

ADVANCES IN KNOWLEDGE

For generating the useful cDWI for PCa detection, radiologists should take care of the combination of b-values when including low b-values.

Identification of a novel fibroblast growth factor, FGF-22, preferentially expressed in the inner root sheath of the hair follicle

We isolated cDNA encoding a novel fibroblast growth factor (FGF-22) (170 amino acids) from human placenta. Of the FGF family members, FGF-22, which appears to be a secreted protein, is most similar to FGF-10 and FGF-7 (approximately 46% and approximately 40% amino acid identities, respectively). The human FGF-22 gene was localized on chromosome 19p13.3. We also isolated mouse cDNA encoding FGF-22 (162 amino acids) from the skin. Mouse FGF-22 shows high homology (87% amino acid identity) to human FGF-22. Mouse FGF-22 mRNA was found to be preferentially expressed in the skin among the mouse adult tissues examined by Northern blotting analysis. By in situ hybridization, FGF-22 mRNA in the skin was found to be preferentially expressed in the inner root sheath of the hair follicle. Therefore, FGF-22 is expected to be a unique FGF that plays a role in hair development.

Non-contrast-enhanced MR angiography using 3D ECG-synchronized half-Fourier fast spin echo

A non-contrast-enhanced three-dimensional (3D) magnetic resonance angiography (MRA) technique, which acquires images in a reasonably short scanning time and requires no contrast agent, is developed. An electrocardiographically (ECG) synchronized 3D half-Fourier fast spin-echo (FSE) technique with an appropriate ECG delay time for every slice encoding in 3D terms was used to examine the thoracic and iliac regions in 16 healthy volunteers at both 0.5 and 1.5 T. Prior to each 3D fresh blood imaging (FBI) experiment, an ECG preparation (ECG-prep) scan was acquired, and an appropriate ECG triggering time was selected for 3D FBI acquisition to optimize visualization of the vessel of interest. In the thoracic and abdominal regions, good-quality 3D MRA images were obtained. Furthermore, the weighted subtraction of two images in different phases provides contrast enhancement between arteries and veins.

High-resolution MR images of inner ear internal anatomy using a local gradient coil at 1.5 Tesla: correlation with histological specimen

PURPOSE

To obtain high-resolution MR images of the inner ear at 1.5 Tesla with a local gradient coil and to correlate these images with the histological specimen.

MATERIALS AND METHODS

All studies were performed on a 1.5 Tesla MR unit with a local gradient coil (23 mT/m, slew rate of 107 mT/m/ms). The cranio-facial region of a cadaver was examined using 3D-fast spin echo (FSE) imaging with the voxel size of 0.27 mm x 0.27 mm x 0.5 mm in 9 h 37 min. Two normal volunteers were examined with the same system using 3D-FSE imaging with the voxel size of 0.20 mm x 0.26 mm x 1.0 mm in 57 min. These images were correlated with the cadaver images and histological specimens.

RESULTS

On cadaver images, internal structures such as the macula utriculi, macula sacculi, crista ampullaris, lamina spiralis ossea, ligamentum spirale cochleae, modiolus, scala tympani, scala vestibuli, and cochlear aqueduct were visualized. On the images of both volunteers, the same structures were visualized as on the cadaver images.

CONCLUSIONS

This study confirmed that high-resolution MR images obtained at 1.5 Tesla can visualize inner ear internal anatomy. Knowledge obtained in this study may be of significant value for the diagnosis of pathology in the area of the inner ear.

[2D-FASE MRCP for pediatrics with congenital biliary dilatation: usefulness of non-breath-hold one-shot MRCP]

The usefulness of magnetic resonance cholangiopancreatography (MRCP) using the non-breath-hold one-shot technique was evaluated. Ten children suffering from congenital biliary dilatation (CBD) were included. Four of them were preoperative cases, and the remaining six postoperative. All MR images taken were compared with endoscopic retrograde cholangiopancreatography or intraoperative cholangiography. MR images using the non-breath-hold one-shot technique clearly showed the confluence of the common bile duct and the main pancreatic duct in seven of the cases. The confluence of the common bile duct and main pancreatic duct was obscure in the other three cases, mainly due to motion artifact. These results show that this non-breath-hold one-shot technique is useful for diagnosis and postoperative follow-up of congenital biliary dilatation in children.

Combination of surface anatomy MRI and MR venography to demonstrate cerebral cortex and cortical veins on one image

We assessed in 12 patients the value of MR images in which a surface anatomy scanning (SAS) image and an MR venogram are added. We obtained SAS images using the half-Fourier technique and MR venograms with the 2D phase-contrast technique. The added images provided sufficient information concerning brain surface as well as cortical veins. Their findings correlated with those at surgery in six patients who underwent scalp incision marking. This technique provides valuable anatomical information preoperatively.

MR imaging of the inner ear: comparison of a three-dimensional fast spin-echo sequence with use of a dedicated quadrature-surface coil with a gadolinium-enhanced spoiled gradient-recalled sequence

PURPOSE

To prospectively evaluate the sensitivity and specificity of magnetic resonance (MR) imaging in the inner ear with a long echo train, three-dimensional (3D), asymmetric Fourier-transform, fast spin-echo (SE) sequence with use of a dedicated quadrature-surface phased-array coil to detect vestibular schwannoma in the cerebellopontine angle and the internal auditory canal.

MATERIALS AND METHODS

In 205 patients (410 ears) with ear symptoms, 1.5-T MR imaging was performed with unenhanced 3D asymmetric fast SE and gadolinium-enhanced 3D gradient-recalled (SPGR) sequences with use of a quadrature surface phased-array coil. The 3D asymmetric fast SE images were reviewed by two radiologists, with the gadolinium-enhanced 3D SPGR images used as the standard of reference.

RESULTS

Nineteen lesions were detected in the 410 ears (diameter range, 2-30 mm; mean, 10.5 mm +/- 6.4 [standard deviation]; five lesions were smaller than 5 mm). With 3D asymmetric fast SE, sensitivity, specificity, and accuracy, respectively, were 100%, 99.5%, and 99.5% for observer 1 and 100%, 99.7%, and 99.8% for observer 2.

CONCLUSION

The unenhanced 3D asymmetric fast SE sequence with a quadrature-surface phased-array coli allows the reliable detection of vestibular schwannoma in the cerebellopontine angle and internal auditory canal.

MR cisternography using a three-dimensional half-fourier single-shot fast spin-echo sequence

Application of a three-dimensional half-fourier single-shot fast spin-echo sequence to MR cisternography is presented. This technique is capable of demonstrating normal cranial nerves. It is also useful in screening for acoustic neuroma as well as in the diagnosis of neurovascular compression.

Three-dimensional fast spin-echo MR of the inner ear: ultra-long echo train length and half-Fourier technique

We compared the image quality of the newly developed ultra-long echo train length (ETL) 3-D fast spin-echo (FSE) and half-Fourier technique, which is performed in less than 3 minutes, with the conventional 3D-FSE imaging technique, which takes 15 minutes, in assessing MR examinations of the inner ear. The new method's images were almost comparable to the conventional 3D-FSE images in depicting anatomic details and pathologic findings. Implementation of the ultra-long ETL and half-Fourier 3D-FSE imaging technique enables acquisition of inner ear MR studies in a vastly reduced time and with high spatial resolution without significant penalty, opening the possibility for low-cost screening of acoustic tumors without contrast enhancement in less than 3 minutes.

A novel MR angiography technique: SPEED acquisition using half-Fourier RARE

A novel MR angiography (MRA) method, swap phase encode extended data (SPEED), was developed. Two one-shot images with the phase-encode directions swapped were collected within a single breath-hold period and processed with a maximum intensity projection (MIP) to obtain an image. In this study, a long echo train two-dimensional rapid acquisition with relaxation enhancement (RARE) sequence with half-Fourier (half-RARE) was used to obtain the pulmonary MRA images. The MIP image obtained using the SPEED technique presented promising results for pulmonary vessels.

Improvement of MR cholangiopancreatography at .5 T: three-dimensional half-averaged single-shot fast spin echo with multi-breath-hold technique

The purpose of this study was to determine the advantage of a three-dimensional (3D) single-shot fast-spin-echo (SSFSE) sequence to obtain MR cholangiopancreatography (MRCP) with a .5-T MR unit by comparison with a two-dimensional (2D) SSFSE sequence. MRCP with 2D-SSFSE and with 3D-SSFSE with 128 echo train lengths was performed on 15 volunteers and 38 patients with pancreatobiliary disease using a .5-T MR unit. For maximum intensity projection (MIP) reconstruction, the section thickness of source images was 4 mm in the 2D-SSFSE and 3 mm in the 3D-SSFSE. 3D volume data in 3D-SSFSE were obtained using repeated short breath-hold of 2 seconds for every repetition time throughout the examination. The image quality, duct conspicuity, signal-intensity ratio (SIR), and contrast-to-noise ratio (CNR) were evaluated. In 23 of the patients who underwent both MRCP and direct cholangiopancreatography (endoscopic retrograde cholangiopancreatography [ERCP]/percutaneous transhepatic cholangiography [PTC]), a comparison between these two modalities was also conducted. The image quality of the MIP image with 3D-SSFSE (49 of 53, 92.5% graded excellent or good) was superior to that with 2D-SSFSE (31 of 53, 58.4%). Duct conspicuity, SIR, and CNR were significantly higher with 3D-SSFSE than with 2D-SSFSE. 3D-SSFSE also showed a stronger relationship with the ERCP/PTC findings compared to 2D-SSFSE. 3D-SSFSE provided satisfactory quality, SIR, and CNR of MRCP images, even when a .5-T MR unit was used, because the breath-hold technique used during 3D data sampling minimized all types of motion effects.

Fast MR imaging of abdomen: application of the QD body coil and low refocusing flip angle

The fast spin-echo (FSE) method as applied to abdominal imaging has undergone considerable technical and clinical study over the past several years. However, except in a limited number of institutions, this time-saving sequence has not replaced the conventional spin-echo (CSE) method. In particular, FSE is less frequently employed for examination of the liver because FSE and CSE images of hepatic tumors and normal hepatic tissues differ significantly in terms of contrast. The signal reduction effect of magnetization transfer contrast (MTC) is affected by the number of slice sections and by the specific absorption rate (SAR). We have employed a newly developed quadrature detection (QD) body coil to enhance the signal-to-noise ratio (S/N) and a low refocusing flip angle to suppress SAR and MTC effects. We have also evaluated the effectiveness of this new method by carrying out visual assessment of clinical images and quantitative measurements of signal intensity in these images. Several problems must be overcome before the FSE method is able to replace the CSE method in the clinical examination of the abdomen. The selection of sequences with a low refocusing flip angle combined with the use of the QD body coil promises to be one possible solution to these problems.

The function of gingival lymphocytes on the establishment of human periodontitis

Human periodontitis has been confirmed to be an IgG plasma cell-rich lesion. However, we also detected many T cells, both CD4-positive and CD8-positive cells, in periodontal lesions. Some of these T cells expressed HLA-DR (la-like) antigen on their surfaces, and the proportion of HLA-DR+ cells was approximately equal in both CD4+ and CD8+ cell populations (Okada et al., 1983, 1984). Consequently, both helper and suppressor T cells were believed to participate in the establishment of periodontal lesions.

On the other hand, B cells were thought to be activated polyclonally in periodontal lesions, because a variety of periodontal florae possessed polyclonal B-cell-activating activity. We demonstrated that Actinomyces viscosus T14V stimulated mouse spleen B cells polyclonally and induced many IgM-producing cells but few IgG-producing cells. Moreover, IgG-producing cells were differentiated from only surface IgG-positive B cells but not from surface IgG-negative B cells-namely, surface IgM- or IgA-positive B cells (Harada et al., 1988). These results suggested that memory B cells, which had already been primed with appropriate antigens, might migrate into periodontal lesions, and then be activated polyclonally and develop into IgG-producing cells.

The periodontal lesion could, therefore, be induced by the interactions of immunoregulatory mechanisms of T cells and polyclonal B cell activity of periodontal florae. In fact, L3T4-positive T cells (helper-inducer T cells) enhanced IgG synthesis of mouse spleen B cells which had been activated with T-independent B cell activators such as LPS and A. viscosus preparations (Okada et al., 1987; Ito et al., 1988). We hypothesized from the above results that autoreactive T cells recognized the increasing self-MHC class II(Ia) antigen on B cells which had been activated with polyclonal B cell activators, and then produced soluble factors, which could enhance IgG synthesis of these B cells. Autoreactive T cells as well as PBAs, thus, may play an important role in the establishment of the IgG plasma cell-rich periodontal lesion.