Performance Evaluation of Age Estimation from T1-Weighted Images Using Brain Local Features and CNN.

The age of a subject can be estimated from the brain MR image by evaluating morphological changes in healthy aging. We consider using two-types of local features to estimate the age from T1-weighted images: handcrafted and automatically extracted features in this paper. The handcrafted brain local features are defined by volumes of brain tissues parcellated into 90 or 1,024 local regions defined by the automated anatomical labeling atlas. The automatically extracted features are obtained by using the convolutional neural network (CNN). This paper explores the difference between the handcrafted features and the automatically extracted features. Through a set of experiments using 1,099 T1-weighted images from a Japanese MR image database, we demonstrate the effectiveness of the proposed methods, analyze the effectiveness of each local region for age estimation and discuss its medical implication.

Design and on-orbit operation of the Soft X-ray Spectrometer ADR on the Hitomi Observatory.

The Soft X-ray Spectrometer (SXS) instrument that flew on the Astro-H observatory was designed to perform imaging and spectroscopy of x-rays in the energy range of 0.2 to 13 keV with a resolution requirement of 7 eV or better. This was accomplished using a 6x6 array of x-ray microcalorimeters cooled to an operating temperature of 50 mK by an adiabatic demagnetization refrigerator (ADR). The ADR consisted of three stages in order to operate using either a 1.2 K superfluid helium bath or a 4.5 K Joule-Thomson (JT) cryocooler as its heat sink. The design was based on the following operating strategy. After launch, while liquid helium was present (cryogen mode), two of the ADR's stages would be used to single-shot cool the detectors, using the helium as a heat sink. When the helium was eventually depleted (cryogen-free mode), all three ADR stages would be used to continuously cool the helium tank to about 1.5 K, and to single-shot cool the detectors (to 50 mK), using the JT cryocooler as a heat sink. The Astro-H observatory, renamed Hitomi after its successful launch in February 2016, carried approximately 36 liters of helium into orbit. Based on measurements during ground testing, the average heat load on the helium was projected to be 0.66 mW, giving a lifetime of more than 4 years. On day 5, the helium had cooled to <1.4 K and ADR operation began, successfully cooling the detector array to 50 mK. The ADR's hold time steadily increased to 48 hours as the helium cooled to a temperature of 1.12 K. As the commissioning phase progressed, the ADR was recycled (requiring approximately 45 minutes) periodically, either in preparation for science observations or whenever the 50 mK stage approached the end of its hold time. In total, 18 cycles were completed by the time an attitude control anomaly led to an unrecoverable failure of the satellite on day 38. This paper presents the design, operation and on-orbit performance of the ADR in cryogen mode as the foreshortened mission did not provide an opportunity to test cryogen-free mode.

Hitomi X-ray studies of Giant Radio Pulses from the Crab pulsar.

To search for giant X-ray pulses correlated with the giant radio pulses (GRPs) from the Crab pulsar, we performed a simultaneous observation of the Crab pulsar with the X-ray satellite Hitomi in the 2 - 300 keV band and the Kashima NICT radio observatory in the 1.4 - 1.7 GHz band with a net exposure of about 2 ks on 25 March 2016, just before the loss of the Hitomi mission. The timing performance of the Hitomi instruments was confirmed to meet the timing requirement and about 1,000 and 100 GRPs were simultaneously observed at the main and inter-pulse phases, respectively, and we found no apparent correlation between the giant radio pulses and the X-ray emission in either the main or inter-pulse phases. All variations are within the 2 sigma fluctuations of the X-ray fluxes at the pulse peaks, and the 3 sigma upper limits of variations of main- or inter-pulse GRPs are 22% or 80% of the peak flux in a 0.20 phase width, respectively, in the 2 - 300 keV band. The values become 25% or 110% for main or inter-pulse GRPs, respectively, when the phase width is restricted into the 0.03 phase. Among the upper limits from the Hitomi satellite, those in the 4.5-10 keV and the 70-300 keV are obtained for the first time, and those in other bands are consistent with previous reports. Numerically, the upper limits of main- and inter-pulse GRPs in the 0.20 phase width are about (2.4 and 9.3) ×10-11 erg cm-2, respectively. No significant variability in pulse profiles implies that the GRPs originated from a local place within the magnetosphere and the number of photon-emitting particles temporally increases. However, the results do not statistically rule out variations correlated with the GRPs, because the possible X-ray enhancement may appear due to a > 0.02% brightening of the pulse-peak flux under such conditions.

Brain age estimation from T1-weighted images using effective local features.

Statistical analysis using large-scale brain magnetic resonance (MR) image databases has examined that brain tissues have age-related morphological changes. The age of a subject can be estimated from the brain MR image by evaluating morphological changes with healthy aging. This paper proposes an age estimation method using local features of T1-weighted MR images. The brain local features are defined by volumes of brain tissues parcellated into 1,024 local regions defined by the automated anatomical labeling atlas. This paper also proposes the effective local feature selection method to improve the accuracy of age estimation. We evaluate the accuracy of the proposed method using 1,099 T1-weighted images from a Japanese MR image database. We also analyze effectiveness of each local region for age estimation and discuss its medical implication.

Age estimation using effective brain local features from T1-weighted images.

This paper proposes a simple method of selecting effective brain local features for age estimation from T1-weighted MR images. We also employ the high-resolution AAL atlas, which is defined by 1,024 local regions, to improve the accuracy of age estimation. We evaluate performance of the proposed method using 1,099 T1-weighted images from a large-scale brain MR image database of healthy Japanese, and demonstrate that the proposed method exhibits efficient performance of age estimation compared with conventional methods.