Development of Pericardial Fat Count Images Using a Combination of Three Different Deep-Learning Models: Image Translation Model From Chest Radiograph Image to Projection Image of Three-Dimensional Computed Tomography.

RATIONALE AND OBJECTIVES: Pericardial fat (PF)-the thoracic visceral fat surrounding the heart-promotes the development of coronary artery disease by inducing inflammation of the coronary arteries. To evaluate PF, we generated pericardial fat count images (PFCIs) from chest radiographs (CXRs) using a dedicated deep-learning model. MATERIALS AND METHODS: We reviewed data of 269 consecutive patients who underwent coronary computed tomography (CT). We excluded patients with metal implants, pleural effusion, history of thoracic surgery, or malignancy. Thus, the data of 191 patients were used. We generated PFCIs from the projection of three-dimensional CT images, wherein fat accumulation was represented by a high pixel value. Three different deep-learning models, including CycleGAN were combined in the proposed method to generate PFCIs from CXRs. A single CycleGAN-based model was used to generate PFCIs from CXRs for comparison with the proposed method. To evaluate the image quality of the generated PFCIs, structural similarity index measure (SSIM), mean squared error (MSE), and mean absolute error (MAE) of (i) the PFCI generated using the proposed method and (ii) the PFCI generated using the single model were compared. RESULTS: The mean SSIM, MSE, and MAE were 8.56 × 10-1, 1.28 × 10-2, and 3.57 × 10-2, respectively, for the proposed model, and 7.62 × 10-1, 1.98 × 10-2, and 5.04 × 10-2, respectively, for the single CycleGAN-based model. CONCLUSION: PFCIs generated from CXRs with the proposed model showed better performance than those generated with the single model. The evaluation of PF without CT may be possible using the proposed method.

[Derivation of Conversion Formula of Image Quality Figure (IQFinv) Values by Old and New Analysis Software for Mammographic Contrast-detail Phantom].

PURPOSE: This study deals with the conversion of the image quality figure (IQFinv) values with CDMAM Analyser ver. 1 (old analysis software) used to analyse the images of CDMAM 3.4 phantom into the IQFinv values with a new analysis software, that is, CDMAM Analyser (ver. 2 and ver. 2.3), which is a software improved to be closer to human visual evaluation. METHODS: The IQFinv values were calculated using four mammography units and three analysis software. The IQFinv values using the old and new software were compared. RESULTS: The IQFinv values had no difference according to mammography units and beam quality. The conversion formula of IQFinv from CDMAM Analyser ver. 1 to CDMAM Analyser ver. 2 was derived. Furthermore, the conversion formula of IQFinv from CDMAM Analyser ver. 1 to CDMAM Analysis ver. 2.3 was also derived. CONCLUSION: Using each conversion formula, the IQFinv values using the old software can be converted to those using the new software. There was a slight difference between the IQFinv values using the new software.

[Comparison between Variance and Edge Methods for Measuring Image Resolution Properties in Digital Mammography].

We studied the image resolution properties in digital mammography using the variance and edge response function methods. The associated measurements were made using a polymethylmethacrylate phantom of 0-40 mm thickness while maintaining the incident dose to the detector constant, as well as with and without using the phantom but by varying the level of exposure to the detector. The results obtained using the two methods were compared on the object plane of the mammography unit with compensation for the magnification and rejection of the scattered radiation. In this unit, the imaging process of resolution enhancement may be performed according to the phantom thickness, shape of test device, exposure level, and so on. The modulation transfer function values of the variance method were slightly higher than those of the edge method, approximately 6% at 2 cycles/mm and 8% at 4 cycles/mm without resolution enhancement processing. We concluded that the variance method is more suitable than the edge method for the constancy test or quality control because of its lower coefficient of variation.

Effects of wind and rain on thermal responses of humans in a mildly cold environment.

The purpose of the present study was to clarify the effects of wind and rain on peripheral heat loss by non-exercising minimally clothed humans in a mildly cold environment. Seven healthy young male subjects wearing only shorts rested in a standing position for 20 min at an ambient temperature of 15 degrees C under three conditions: without exposure to wind or rain (CON), with exposure to wind (3 m/s) (WIND) and with exposure to wind (3 m/s) and rain (40 mm/h) (WIND + RAIN). Mean heat loss measured using a heat flux transducer was significantly greater in the subjects exposed to WIND + RAIN compared to those exposed to CON and WIND conditions (p < 0.01). Metabolic heat production was significantly greater under WIND + RAIN than under CON and WIND (p < 0.01). Decrease in heat storage was significantly larger at WIND + RAIN compared with CON and WIND (p < 0.01). Mean skin temperature was significantly lower under WIND + RAIN than under CON and WIND conditions (p < 0.01). These results indicate that peripheral heat loss significantly increases when humans are exposed to wind and rain for a short period (20 min) under a mildly cold condition.