菅 幹生

One of the major unsolved issues of PET-MRI is the PET attenuation correction using MR images. Conventionally, in PET or PET-CT, attenuation maps (μ-maps) have been obtained by a PET transmission scan, or a CT scan. For PET-MRI, in order to obtain μ-maps from MR images, many studies have been reported, and they can be classified into two methods; the atlas-based method (ABM) and the segmentation-based method (SBM). In the ABM, individual differences such as lesions are not supported. In the SBM, it is difficult to discriminate bone and air, which have large differences in their attenuation coefficients, because these tissues have similar MR signal values in the T1 weighted (T1w) MR images. In this work, therefore, we proposed a hybrid segmentation-atlas method (HSAM) to utilize the advantages and compensate for the disadvantages of both the ABM and the SBM. At first, the proposed method follows the SBM approach. In the bone and air regions where T1w MRI signals are similar and low, the HSAM uses information from a standard μ-map obtained through the ABM. For evaluation, the head data from 6 healthy volunteers were obtained by PET (ECAT Exact HR+) and MRI (Philips Intera 1.5T). We estimated μ-maps by the ABM, the SBM and the HSAM, and PET images were reconstructed though attenuation correction with those μ-maps. In comparison of the μ-maps, the HSAM and ABM outperformed the SBM. In comparison of the final PET images, a similar tendency was seen. For patient data, which would contain different distribution from the database, the HSAM is expected to outperform the ABM. © 2012 IEEE.

We are developing the OpenPET which can provide an open space observable and accessible to the patient during PET measurements. In addition, we have proposed the real-time imaging system for the OpenPET which is expected to be used in PET-guided tumor tracking radiation therapy and demonstrated its tracking ability using a point source and a small OpenPET prototype. However, tumor tracking in the human body still remains as a challenging task when we use F-18-FDG which is the best available tracer for tumors because of its background activity, scatter and attenuation in the body. In this study, we assess conditions under which tumor tracking is feasible in the human body by using the 4D XCAT phantom which is a realistic 4D human whole body phantom. To simulate realistic F-18-FDG distributions, we assigned standardized uptake values (SUVs) to normal organs based on the literature. We conducted Monte Carlo simulation of a human-sized OpenPET geometry by using Geant4 Application for Tomographic Emission (GATE) ver. 6.1 assuming a measurement at 100 minutes after the F-18-FDG injection of 370 MBq and a spherical tumor with the diameter of 10 to 30 mm and SUV of 3 to 10. List-mode data were generated for each 0.5 s time frame of a respiratory cycle of 5 s. Image reconstruction was done in a frame-by-frame manner and tumor position was automatically extracted for each time frame by a pattern matching technique. Tumor movement in the 4D XCAT phantom was about 17 mm at the maximum. The mean error of the tumor positions extracted from the reconstructed images was similar to the PET image resolution when the tumor size was 20 mm or more and SUV was 5 or more. We showed that tumor tracking by the OpenPET is feasible even in the human body scale and for realistic conditions.

One of the major unsolved issues of PET-MRI is the PET attenuation correction using MR images. Conventionally, in PET or PET-CT, attenuation maps (mu-maps) have been obtained by a PET transmission scan, or a CT scan. For PET-MRI, in order to obtain mu-maps from MR images, many studies have been reported, and they can be classified into two methods; the atlas-based method (ABM) and the segmentation-based method (SBM). In the ABM, individual differences such as lesions are not supported. In the SBM, it is difficult to discriminate bone and air, which have large differences in their attenuation coefficients, because these tissues have similar MR signal values in the T1 weighted (T1w) MR images. In this work, therefore, we proposed a hybrid segmentation-atlas method (HSAM) to utilize the advantages and compensate for the disadvantages of both the ABM and the SBM. At first, the proposed method follows the SBM approach. In the bone and air regions where T1w MRI signals are similar and low, the HSAM uses information from a standard mu-map obtained through the ABM. For evaluation, the head data from 6 healthy volunteers were obtained by PET (ECAT Exact HR+) and MRI (Philips Intera 1.5T). We estimated mu-maps by the ABM, the SBM and the HSAM, and PET images were reconstructed though attenuation correction with those mu-maps. In comparison of the mu-maps, the HSAM and ABM outperformed the SBM. In comparison of the final PET images, a similar tendency was seen. For patient data, which would contain different distribution from the database, the HSAM is expected to outperform the ABM.

We are developing a novel, general purpose isotropic-3D PET detector X'tal cube which has high spatial resolution in all three dimensions. The research challenge for this detector is implementing effective detection of scintillation photons by covering six faces of a segmented crystal block with silicon photomultipliers (SiPMs). In this paper, we developed the second prototype of the X'tal cube for a proof-of-concept. We aimed at realizing an ultimate detector with 1.0 mm(3) cubic crystals, in contrast to our previous development using 3.0 mm(3) cubic crystals. The crystal block was composed of a 16 x 16 x 16 array of lutetium gadolinium oxyorthosilicate (LGSO) crystals 0.993 x 0.993 x 0.993 mm(3) in size. The crystals were optically glued together without inserting any reflector inside and 96 multi-pixel photon counters (MPPCs, S10931-50P, i.e. six faces each with a 4 x 4 array of MPPCs), each having a sensitive area of 3.0 x 3.0 mm(2), were optically coupled to the surfaces of the crystal block. Almost all 4096 crystals were identified through Anger-type calculation due to the finely adjusted reflector sheets inserted between the crystal block and light guides. The reflector sheets, which formed a belt of 0.5 mm width, were placed to cover half of the crystals of the second rows from the edges in order to improve identification performance of the crystals near the edges. Energy resolution of 12.7% was obtained at 511 keV with almost uniform light output for all crystal segments thanks to the effective detection of the scintillation photons.

The OpenPET geometry is our new idea to visualize a physically opened space between two detector rings. In this paper, we developed the first small prototype to show a proof-of-concept of OpenPET imaging. Two detector rings of 110 mm diameter and 42 mm axial length were placed with a gap of 42 mm. The basic imaging performance was confirmed through phantom studies; the open imaging was realized at the cost of slight loss of axial resolution and 24% loss of sensitivity. For a proof-of-concept of PET image-guided radiation therapy, we carried out the in-beam tests with C-11 radioactive beam irradiation in the heavy ion medical accelerator in Chiba to visualize in situ distribution of primary particles stopped in a phantom. We showed that PET images corresponding to dose distribution were obtained. For an initial proof-of-concept of real-time multimodal imaging, we measured a tumor-inoculated mouse with F-18-FDG, and an optical image of the mouse body surface was taken during the PET measurement by inserting a digital camera in the ring gap. We confirmed that the tumor in the gap was clearly visualized. The result also showed the extension effect of an axial field-of-view (FOV); a large axial FOV of 126 mm was obtained with the detectors that originally covered only an 84 mm axial FOV. In conclusion, our initial imaging studies showed promising performance of the OpenPET.

Various PET-MRI systems are being developed by a number of research groups. Since almost all the PET detectors used in these PET-MRI systems have no depth-of-interaction (DOI) capability. the parallax error restricts their performances in studies for rodents and the human brain. The DOI detector can reduce the parallax error and lead to improved performance, especially for the spatial resolution of the PET system portion. We have proposed a four-layer DOI encoding method and are developing a new PET system integrated with an RF-coil of the MRI using the four-layer DOI detectors. In the proposed system, PET detectors which consist of a scintillator block, photo sensors and front-end circuits are placed close to the objective. The photo sensors and front-end circuits should be shielded to minimize noises from the MRI and influence on MRI imaging. Elements of the RF coil are inserted between gaps of the scintillation crystal block. The RF-coil element is positioned inside the shielding material. We constructed a prototype four-layer DOI-PET detector and carried out experiments for performance evaluation in an actual MRI system. Additionally, we investigated influences on the PET detector and the MRI system from each other. As a result, sufficient detector performance regarding crystal identification and energy resolution was achieved in simultaneous measurements. The MRI images were also obtained as usual even though the DOI-PET detector was positioned close to the RF coil elements.

Various PET-MRI systems are being developed by a number of research groups. Since almost all the PET detectors used in these PET-MRI systems have no depth-of-interaction (DOI) capability. the parallax error restricts their performances in studies for rodents and the human brain. The DOI detector can reduce the parallax error and lead to improved performance, especially for the spatial resolution of the PET system portion. We have proposed a four-layer DOI encoding method and are developing a new PET system integrated with an RF-coil of the MRI using the four-layer DOI detectors. In the proposed system, PET detectors which consist of a scintillator block, photo sensors and front-end circuits are placed close to the objective. The photo sensors and front-end circuits should be shielded to minimize noises from the MRI and influence on MRI imaging. Elements of the RF coil are inserted between gaps of the scintillation crystal block. The RF-coil element is positioned inside the shielding material. We constructed a prototype four-layer DOI-PET detector and carried out experiments for performance evaluation in an actual MRI system. Additionally, we investigated influences on the PET detector and the MRI system from each other. As a result, sufficient detector performance regarding crystal identification and energy resolution was achieved in simultaneous measurements. The MRI images were also obtained as usual even though the DOI-PET detector was positioned close to the RF coil elements. © 2011 IEEE.

One of the challenging applications of PET is for in-beam PET, which is an in situ monitoring method for charged particle therapy. For this purpose, we have previously proposed an open-type PET scanner, Open PET. The original OpenPET has a physically opened field-of-view (FOV) between two detector rings which irradiation beams pass through. This dual-ring OpenPET has a wide axial FOV including the gap. Therefore this geometry is not necessarily the most efficient when it is applied to in-beam PET in which only a limited FOV around the irradiation field is required. In this paper, we proposed new single-ring OpenPET geometry as more efficient geometry dedicated to in-beam PET. The detector ring of the proposed geometry is a cylinder both ends of which are cut by parallel aslant planes. The proposed geometry can be made compact so that the beam port can be placed close to the patient.One of the problems for the proposed geometry is arranging the rectangular block detectors. We proposed two arrangement types, a slanted ellipse type where oval detector rings are slanted and stacked and an axial shift type where block detectors originally forming a conventional PET scanner are axially shifted little by little. We compared the two types through numerical simulations. The simulated systems were a small prototype that we designed for a proof-of-concept and a human sized system. We evaluated the effect of utilizing depth-of-interaction (DOl) measurement in both sized system in both types.In the small sized simulations, both types resulted in acceptable image quality with DOl capability.In the human sized simulation, on the other hand, the effect of DOl utilizing was not clearly shown because the size of spots was too large compared with the size of the crystals. Therefore, our future work is further investigation the quality of reconstructed images for both types in the human sized simulation.

We developed a new PET depth-of-interaction (DOI) detector in which the scintillation crystal array is extended in one direction. It can be utilized as a PET detector having, for instance, long axial direction. The detector has the "crystal cube (X'tal cube)" structure and the extension was enabled by the feature of the X'tal cube that some MPPCs are always placed close to an originating position of the scintillation photons.X'tal cube is a DOI PET detector we have developed and it can provide isotropic spatial resolution. X'tal cube consists of a scintillation crystal block segmented into cubes and multi-pixel photon counters (MPPCs). A number of MPPCs are coupled on all six sides of the crystal block to detect scintillation photons near their originating position. In a preliminary experiment previously carried out, we confirmed sufficient detector performance of X'tal cube using a prototype composed of a 6 x 6 x 6 array of Lu2xGd2(1-x) SiO5:Ce (LGSO, x=0.9) crystals with dimensions of 3.0 x 3.0 x 3.0 mm(3) and MPPCs of 3.0 x 3.0 mm(2) active area. In this study, the crystal block was extended in one direction from 6 crystals to 14 crystals. The results of performance evaluation proved that the 6 x 6 x 14 crystals could be identified with the structure. Although MPPCs were coupled on all sides of the crystal block in the extended X'tal cube, at the central part, the detected radiation position would be calculated with mainly the signals of close MPPCs on the four surfaces. That means the detector can be long as considered appropriate for a system while keeping the isotropic spatial resolution.

The OpenPET, which has a physically opened space between two detector rings, is our new geometry to enable PET imaging during radiation therapy. Especially, tracking a moving target such as a tumor in the lung will become possible if the real-time imaging system is realized. In this paper, we developed a list-mode image reconstruction method using general-purpose computing on graphics processing units (GPGPUs) toward real-time image reconstruction. We used the dynamic row-action maximum likelihood algorithm (DRAMA). For GPU implementation, efficiency of acceleration depends on implementation methods; reduced conditional statements and efficient memory accesses are required. On the other hand, accurate system model is required to improve quality of reconstructed images. Therefore, we developed a new system model which was suited for the GPU implementation. In the new system model, the detector response functions (DRFs), which were calculated analytically to represent the probability distribution of each LOR, were modeled by sixth-order polynomial functions. The system model enabled us to calculate each element of the system matrix with reduced conditional statements. We applied the developed method to a small OpenPET prototype, which was developed for a proof-of-concept. We compared the proposed system model to the sub-LOR model, a geometrically-defined accurate system model which we had previously proposed. The difference between the reconstructed images with the new system model using GPU and the sub-LOR model using CPU was very small. Our new system model on GPU was 46.3 times faster than the sub-LOR model on CPU.

The "OpenPET" geometry has potential for real-time tracking of targeted tumors and in situ visualization of dose distribution in radiation therapy. For the real-time imaging system, we are intending to use one-pass list-mode dynamic row-action maximum likelihood algorithm (DRAMA) and implement it using general-purpose computation on graphics processing unit techniques. However, it is difficult to consistently reconstruct in real-time because the amount of list-mode data acquired in PET scan can be large depending on the activity and reconstruction speed depends on the amount of the list-mode data. In this study, we developed a system to control the data used in the reconstruction step keeping quantitative performance. In the proposed system, data transfer control system limits the event count to be used in the reconstruction step according to the reconstruction speed, and the reconstructed images are properly intensified by using ratio of the used count to the total count. We implemented the system on a small OpenPET prototype system and evaluated the performance in terms of the real-time tracking ability by displaying reconstructed images in which the intensity was compensated. The intensity of the displayed image was properly correlated with the original count rate and a frame rate of 2 frames per second was achieved without significant delay.

The X'tal cube is our new PET detector which is composed of a segmented crystal block and multi-pixel photon counters (MPPCs). It detects scintillation photons in three dimensions by arranging MPPCs on multiple surfaces of the crystal block. It is possible to collect scintillation photons efficiently for higher energy, spatial, and timing resolution. We have previously reported that 3 mm isotropic spatial resolution was obtained using the X'tal cube. In this study, we achieved 1 mm isotropic spatial resolution. The developed X'tal cube detector was composed of 1.0 x 1.0 x 1.0 mm(3) LYSO crystals, which were structured into 16 x 16 x 16 arrays. Each surface of the crystal block was covered with 4x4 MPPCs with 3.0 mm x 3.0 mm sensitive area. We evaluated crystal identification and energy performance, and the results showed that the X'tal cube has the expected 1 mm isotropic spatial resolution.

We have proposed a depth of interaction (DOI) PET detector named X'tal cube, in which a number of Multi-Pixel Photon Counters (MPPCs) are coupled on various positions of six surfaces of a segmented scintillation crystal block. There are no reflectors within the block, and the areas among MPPCs on the surface are covered with reflector. The MPPC is thin and light solid-state photo-detector so that it is possible to set the MPPCs on the subject side in the detector ring of a PET scanner and the crystal blocks can be closely placed each other in the PET detector ring.To study the characteristics of the X'tal cube, we constructed a segmented crystal block consisting of six layers of a 6 x 6 crystal array with Lu2xGd2(1-x)SiO5 : Ce (LGSO, x=0.9) crystals. Each crystal is 3.0 x 3.0 x 3.0 mm(3) in dimensions. We examined crystal identification performance for different arrangements of photo-detector elements on the block surfaces. The preliminary experiments showed the possibility of realizing a 3D detector having isotropic resolutions.

MRE methods deform the sample using an external vibration system. We have been using a transverse driver, which generates shear waves at the object surface. One of the problems is that shear waves rapidly attenuate at the surface of tissue and do not propagate into the body. In this study, we compared the shear waves generated by transverse and longitudinal drivers. The longitudinal driver was found to induce shear waves deep inside a porcine liver phantom. These results suggest that the longitudinal driver will allow measurement of the shear modulus deep inside the body. © 2007 IEEE.

This paper describes the measurement of various head images and the presentation of integrated images on an immersive projection system (IPS). We handled optical and magnetic method to obtain the information about face, head, brain and blood vessel. The system enables to observe the inside of a head interactively using a joystick, which gives the impression to travel in a human head freely. It is thought that the integrated images are available not only for diagnosis but also treatment of a brain in the future robot surgical operation with a manipulator or a micro-machining device.

This paper describes the measurement of various head images and the presentation of integrated images on an immersive projection system (IPS). We handled optical and magnetic method to obtain the information about face, head, brain and blood vessel. The system enables to observe the inside of a head interactively using a joystick, which gives the impression to travel in a human head freely. It is thought that the integrated images are available not only for diagnosis but also treatment of a brain in the future robot surgical operation with a manipulator or a micro-machining device.

To provide realistic surgical simulation, haptic feedback is important. In the existing surgical simulators, the fidelity of the deformation and haptic feedback is limited because they are based on the subjective evaluation of the expert-user and not on an objective model-based evaluation. To obtain elastic modulus of in-vivo human tissues, magnetic resonance elastography (MRE) was developed. MRE is a phase-contrast-based method that can visualize propagating strain waves in materials. The quantitative values of shear modulus can be calculated by estimating the local wavelength of the wave pattern. Low frequency mechanical motion must be used for soft tissue-like materials, because strain waves rapidly attenuate at higher frequency. Therefore, wavelength in MRE is long. It is difficult to estimate local wavelength with high spatial resolution especially from noisy MRE. In the MRE sequence, motion-sensitizing gradient (MSG) are synchronized with the mechanical cyclic motion. MRE with multiple initial phase offsets can be generated with increasing delays between the MSG and mechanical excitation, In this paper, we describe a method of measuring local wavelength with high spatial resolution by combining multiple phase offsets MRE. To confirm the reliability of this method, a computer simulation and phantom study were performed. The shear modulus measured with various elastic objects was well consistent with the value obtained by MRE and the mechanical method. The shear moduli of excised porcine liver and in vivo human calf muscle were also analyzed by this method.

This paper describes 3D (three dimensional) visualization of a brain blood vessel on a CRT (cathode ray tube) or an IFS (immersive projection system). For the visualization, MRA (magnetic resonance angiography) images were acquired, blood vessels were extracted from MRA images and a 3D image of blood vessels in a human brain was projected on an IFS. The 3D image allows us to understand the distribution of blood vessels and guides us into a human body. We concluded that the image is available not only for diagnosis but also treatment of a brain in the future robot surgical operation.

Magnetic resonance elastography (MRE) is a phase-contrast-based MR method that can visualize propagating strain waves in materials. The quantitative values of shear modulus can be calculated by estimating the local wavelength (LW) of the wave pattern. Low frequency mechanical motion must be used for soft tissue-like materials, because strain waves rapidly attenuate as higher frequency. Therefore, it is difficult to estimate LW with high spatial resolution especially from noisy MRE image. In the MRE sequence, motion-sensitizing gradient (MSG) is synchronized with the mechanical cyclic motion. MRE images with multiple phase offsets can be generated with increasing delays between MSG and the mechanical excitation. In this report, we describe the new algorithm in order to measure LW in higher spatial resolution using MRE images with multiple phase offsets. This method was evaluated by the computer simulation and the phantom study. The result shows LW was successively estimated.