Purpose: In preoperative contrast-enhanced computed tomography (CT) of the stomach, an effervescent agent is taken internally to stretch the wall of the stomach to allow evaluation of tumor depth. However, there are no reports on the effects of the effervescent agent on the contrast effects of arteries, veins, portal veins, and hepatic veins. The purpose of this study was to clarify the effects of an effervescent agent on the contrast effects of blood vessels in abdominal contrast-enhanced CT. Methods: The subjects were 60 preoperative gastric or pancreatic CT patients who underwent contrast-enhanced CT either with or without the effervescent agent. CT attenuations of the vessels were measured. Results: CT attenuations of the left gastric artery, left gastroepiploic artery, right gastroepiploic artery, left gastric vein, left gastroepiploic vein, right gastroepiploic vein, portal vein, and hepatic vein were significantly higher (p<0.01) when the effervescent agent was taken. Conclusion: In abdominal contrast-enhancement CT, CT attenuations of the left gastric artery, left gastroepiploic artery, right gastroepiploic artery, left gastric vein, left gastroepiploic vein, right gastroepiploic vein, portal vein, and hepatic vein were found to be higher when the patient was administered the effervescent agent.

Purpose: The purposes of this study were to clarify the difference in image characteristics of EPI with compressed SENSE (EPICS) DWI and conventional EPI-SENSE DWI when the reduction factor is increased, and to investigate the optimal reduction factor setting for EPICS DWI. Methods: Using Philips MRI Ingenia Elition 3.0T and a phantom, we compared the SNR, CNR, and ADC values between the EPI-SENSE and EPICS methods with increasing reduction factor. The presence of deployment failure artifacts was verified by the dynamic noise scan method. P<0.05 was set as the significance level. Results: The EPICS method showed significantly higher SNR (1.1–1.4 times) and CNR (1.3–1.8 times) than the EPI-SENSE method at reduction factors 2–5 (P<0.05), with less deployment failure artifacts. The ADC of the EPICS method was 0.03–0.07×10⁻³ mm²/s lower at reduction factors 3–5. Conclusion: EPICS DWI is a useful imaging method that is highly effective in reducing image degradation in high reduction factor imaging.

Readout-segmented echo planar imaging has a reacquisition mode in case of body motion. The phantom was moved and imaged to characterize the reacquisition mode. The counts, angle, and terms of continuous body motion and the terms and counts of intermittent body motion were changed. Then, we investigated the effect on the image quality and the improvement effect of the reacquisition mode. We compared the signal intensity of each pixel in images without body motion and images with and without the reacquisition mode using Spearman’s correlation coefficient. The correlation coefficient decreased with increasing counts of body motion. There was no difference by angle. The correlation coefficient was high for body motion immediately after the start of imaging and decreased thereafter. The correlation coefficient was high when the counts of body motion were decreased and even when the terms of body motion increased. In all cases, the correlation coefficient was improved by the reacquisition mode.

In this study, we investigated the necessary phantom length for dose evaluation of kV-CBCT mounted on the linear accelerator using a Farmer-type ionization chamber and cylindrical PMMA phantom, which many radiotherapy facilities own. Furthermore, a phantom length correction factor was proposed to compensate for the lack of scattered radiation contributed from the craniocaudal direction due to the inadequate phantom length. The air-absorbed dose at the center of a cylindrical PMMA phantom measured by a Farmer-type ionization chamber indicated that the contribution of scattered radiation saturates when the phantom length is 300 mm or longer. The phantom length correction factor was calculated from an approximate curve showing the relationship of the charge ratio measured using phantoms of 300 mm and 150 mm in length to the X-ray beam width. The air-absorbed dose measured with the 150-mm length phantom, corrected by the phantom length correction factor, showed a 1.61% dose difference from the air-absorbed dose measured with the 300-mm length phantom. In this study, the air-absorbed dose at the center of the phantom could be estimated over a wide X-ray beam width only using a 150-mm length cylindrical PMMA phantom. The method proposed in this study could be used in any radiation therapy facility.

The effectiveness of a neck fixation device to improve the image quality of DWIBS was investigated. Healthy volunteers were examined while chewing with and without a neck fixation device using a 3-T MRI system. Distance of mandibular movement was measured using true-fast imaging of steady-state precession (true FISP). Signal-to-noise ratio (SNR) and apparent diffusion coefficient (ADC) of DWIBS were measured. Image quality of DWIBS was scored by visual evaluation. These values were compared with and without a neck fixation device. Regarding results, the mandibular displacement and ADC were decreased, and the SNR and visual score were increased by the use of the fixation device. There is a significant difference between with and without a neck fixation device in each measurement. The technique using a neck fixation device helps improve image quality of DWIBS in the head and neck region.

Purpose: Magnetic resonance (MR) images provide essential diagnostic information; however, it is also a very burdensome examination for patients. At our hospital, radiologists make imaging instructions for all MR examination orders, but this is a time-consuming task. If a natural language processing model can predict the imaging instructions, it will be possible to reduce the burden on radiologists and the instruction quality can be assured. The purpose of this study was to investigate the feasibility of using natural language processing to predict MR imaging instructions with the aim of assisting radiologists. Methods: Considering the uniqueness of the MR imaging protocols at each facility and the particularity of the test order text, we considered that the use of large datasets and pre-training models would be unsuitable. We focused on LSTM, which has been used for natural language processing, and built a 4-layer bi-LSTM model in combination with our own morphological preprocessing to predict MR imaging instructions. Results: The proposed method achieved macro-average precision, recall, and F1-score of 70.6%, 69.5%, and 68.9%, respectively. Compared to the previous studies, the proposed method achieved satisfactory performance in the natural language analysis task for Japanese. It is considered that the proposed method improved the prediction accuracy of the minority class through direct and indirect effects of vocabulary reduction, optimization, and similarity learning. Conclusion: It is suggested that the proposed method is effective and that the prediction of MR imaging instructions using natural language analysis in combination with the proposed method is feasible.