Search

Central serous chorioretinopathy (CSC) is the fourth most common retinopathy and can reduce quality of life. CSC is assessed using optical coherence tomography (OCT), but deep learning systems have not been used to classify CSC subtypes. This study aimed to build a deep learning system model to distinguish CSC subtypes using a convolutional neural network (CNN). We enrolled 435 patients with CSC from a single tertiary center between January 2015 and January 2020. Data from spectral domain OCT (SD-OCT) images of the patients were analyzed using a deep CNN. Five-fold cross-validation was employed to evaluate the model’s ability to discriminate acute, non-resolving, inactive, and chronic atrophic CSC. We compared the performances of the proposed model, Resnet-50, Inception-V3, and eight ophthalmologists. Overall, 3209 SD-OCT images were included. The proposed model showed an average cross-validation accuracy of 70.0% (95% confidence interval [CI], 0.676–0.718) and the highest test accuracy was 73.5%. Additional evaluation in an independent set of 104 patients demonstrated the reliable performance of the proposed model (accuracy: 76.8%). Our model could classify CSC subtypes with high accuracy. Thus, automated deep learning systems could be useful in the classification and management of CSC.

Quantitative single-photon emission computed tomography/computed tomography (SPECT/CT) using Tc-99m pertechnetate aids in evaluating salivary gland function. However, gland segmentation and quantitation of gland uptake is challenging. We develop a salivary gland SPECT/CT with automated segmentation using a deep convolutional neural network (CNN). The protocol comprises SPECT/CT at 20 min, sialagogue stimulation, and SPECT at 40 min post-injection of Tc-99m pertechnetate (555 MBq). The 40-min SPECT was reconstructed using the 20-min CT after misregistration correction. Manual salivary gland segmentation for %injected dose (%ID) by human experts proved highly reproducible, but took 15 min per scan. An automatic salivary segmentation method was developed using a modified 3D U-Net for end-to-end learning from the human experts (n = 333). The automatic segmentation performed comparably with human experts in voxel-wise comparison (mean Dice similarity coefficient of 0.81 for parotid and 0.79 for submandibular, respectively) and gland %ID correlation (R2 = 0.93 parotid, R2 = 0.95 submandibular) with an operating time less than 1 min. The algorithm generated results that were comparable to the reference data. In conclusion, with the aid of a CNN, we developed a quantitative salivary gland SPECT/CT protocol feasible for clinical applications. The method saves analysis time and manual effort while reducing patients’ radiation exposure.

The objective of this study was to examine the dependence of the pressure level in the wound area on the height of the syringe needle from the wound, the gauge of the needle, and the flow rate using the Bernoulli equation. This study was the control-volume analysis using the Bernoulli equation. At a given height of the syringe needle from the wound, the gauge of the syringe needle was fixed, and the pressure in the wound area, which depended on the flow rate of the irrigation solution discharged from the tip of the needle, was calculated according to the Bernoulli equation and the definition of the flow rate. At a constant flow rate of the irrigation solution, the velocity of the irrigation solution discharged through the syringe needle decreased (7.80 → 0.80) with an increase in the diameter of the needle (18G → 14G). At a constant inner diameter of the needle, the velocity of the irrigation solution increased with a reduction in the flow rate of the solution. As the velocity of the irrigation solution increased, the pressure in the wound area increased. As the height of the syringe needle from the wound area increased, the pressure in the wound area increased. In order to maintain the pressure of 8–15 psi when nurses perform syringe-based irrigation, it is necessary to set the flow rate of the cleaning solution from 3.5 cc/s to less than 4.8 cc/s for 19G. In addition, 20G maintains the flow rate of the solution at 2.6 cc/s or more and less than 3.5 cc/s, 22G maintains the flow rate of solution at 1.3 cc/s or more and less than 1.8 cc/s, and 25G maintains the flow rate of solution at 0.5 cc/s. This study provides nurses with a reference for the flow rate at which syringe-based irrigation can be performed while maintaining the appropriate pressure based on fluid dynamics, which can be used as the basis for wound nursing standards.

Engineered nucleases are widely used for creating frameshift or nonsense mutations in the target genes to eliminate gene functions. The resulting mRNAs carrying premature termination codons can be eliminated by nonsense-mediated mRNA decay. However, it is unclear how effective this process would be in vivo. Here, we found that the nonsense-mediated decay was unable to remove the mutant mRNAs in twelve out of sixteen homozygous mutant mice with frameshift mutations generated using engineered nucleases, which is far beyond what we expected. The frameshift mutant proteins translated by a single nucleotide deletion within the coding region were also detected in the p53 mutant mice. Furthermore, we showed that targeting the exons present downstream of the exons with a start codon or distant from ATG is relatively effective for eliminating mutant mRNAs in vivo, whereas the exons with a start codon are targeted to express the mutant mRNAs. Of the sixteen mutant mice generated, only four mutant mice targeting the downstream exons exhibited over 80% clearance of mutant mRNAs. Since the abnormal products, either mutant RNAs or mutant proteins, expressed by the target alleles might obscure the outcome of genome editing, these findings will provide insights in the improved performance of engineered nucleases when they are applied in vivo.

Neovascular age-related macular degeneration (nAMD) is among the main causes of visual impairment worldwide. We built a deep learning model to distinguish the subtypes of nAMD using spectral domain optical coherence tomography (SD-OCT) images. Data from SD-OCT images of nAMD (polypoidal choroidal vasculopathy, retinal angiomatous proliferation, and typical nAMD) and normal healthy patients were analyzed using a convolutional neural network (CNN). The model was trained and validated based on 4749 SD-OCT images from 347 patients and 50 healthy controls. To adopt an accurate and robust image classification architecture, we evaluated three well-known CNN structures (VGG-16, VGG-19, and ResNet) and two customized classification layers (fully connected layer with dropout vs. global average pooling). Following the test set performance, the model with the highest classification accuracy was used. Transfer learning and data augmentation were applied to improve the robustness and accuracy of the model. Our proposed model showed an accuracy of 87.4% on the test data (920 images), scoring higher than ten ophthalmologists, for the same data. Additionally, the part that our model judged to be important in classification was confirmed through Grad-CAM images, and consequently, it has a similar judgment criteria to that of ophthalmologists. Thus, we believe that our model can be used as an auxiliary tool in clinical practice.

Accurate measurement of glomerular filtration rate (GFR) is essential for optimal decision making in many clinical settings of renal failure. We aimed to show that GFR can be accurately measured using compartmental tracer kinetic analysis of 18F-fluoride dynamic PET/CT. Twenty-three male Sprague-Dawley rats of three experimental groups (cyclosporine-administered [n = 8], unilaterally nephrectomized [n = 8], and control [n = 7]) underwent simultaneous 18F-fluoride dynamic PET/CT and reference 51Cr-EDTA GFR (GFRCrEDTA) test at day 0 and post-intervention day 3. 18F-fluoride PET GFR (GFRF-PET) was calculated by multiplying the influx rate and functional kidney volume in a single-tissue-compartmental kinetic model. Within-test repeatability and between-test agreement were evaluated by intraclass correlation coefficient (ICC) and Bland-Altman analysis. In the control group, repeatability of GFRF-PET was excellent (ICC = 0.9901, repeatability coefficient = 12.5%). GFRF-PET significantly decreased in the renally impaired rats in accordance with respective GFRCrEDTA changes. In the pooled population, GFRF-PET agreed well with GFRCrEDTA with minimal bias (−2.4%) and narrow 95% limits of agreement (−25.0% to 20.1%). These data suggest that the single-compartmental kinetic analysis of 18F-fluoride dynamic PET/CT is an accurate method for GFR measurement. Further studies in humans are warranted.

Central serous chorioretinopathy (CSC) is one of the most common macular diseases that can reduce the quality of life of patients. This study aimed to build a deep learning-based classification model using multiple spectral domain optical coherence tomography (SD-OCT) images together to diagnose CSC. Our proposed system contains two modules: single-image prediction (SIP) and a final decision (FD) classifier. A total of 7425 SD-OCT images from 297 participants (109 acute CSC, 106 chronic CSC, 82 normal) were included. In the fivefold cross validation test, our model showed an average accuracy of 94.2%. Compared to other end-to-end models, for example, a 3D convolutional neural network (CNN) model and a CNN-long short-term memory (CNN-LSTM) model, the proposed system showed more than 10% higher accuracy. In the experiments comparing the proposed model and ophthalmologists, our model showed higher accuracy than experts in distinguishing between acute, chronic, and normal cases. Our results show that an automated deep learning-based model could play a supplementary role alongside ophthalmologists in the diagnosis and management of CSC. In particular, the proposed model seems clinically applicable because it can classify CSCs using multiple OCT images simultaneously.

Much in vivo evidence indicates that cyclooxygenase-2 (COX-2) is deeply involved in tumorigenesis. Although it has been proposed that COX-2-derived pro-inflammatory prostanoids mediate the tumorigenic activity of COX-2, the tumorigenic mechanisms of COX-2 are not yet fully understood. Here, we investigated the mechanism by which COX-2 causes transformation from normal cells to malignant cells by using normal murine or human cells. We found that COX-2 inhibits the pro-senescent function of p53 under oncogenic RAS activation, by which it prevents oncogene-induced senescence (OIS) and induces neoplastic transformation. We also found that COX-2 physically interacts with p53 in the nucleus under oncogenic RAS activation, and that this COX-2-p53 interaction rather than the catalytic activity is involved in the COX-2-mediated inhibition of the pro-senescent function of p53 and OIS, and induction of neoplastic transformation. These findings strongly suggest that the oncogenic property of COX-2 is closely related to its ability to inactivate p53 under strong mitogenic signals, and that aberrant activation of the COX-2/a mitogenic oncogene combination can be a potent driving force for tumorigenesis. This study might contribute to our understanding of the molecular basis for the tumorigenic activity of COX-2 and the development of novel anti-tumor drugs targeting COX-2-p53 interactions.

This cross-sectional study aimed to build a deep learning model for detecting neovascular age-related macular degeneration (AMD) and to distinguish retinal angiomatous proliferation (RAP) from polypoidal choroidal vasculopathy (PCV) using a convolutional neural network (CNN). Patients from a single tertiary center were enrolled from January 2014 to January 2020. Spectral-domain optical coherence tomography (SD-OCT) images of patients with RAP or PCV and a control group were analyzed with a deep CNN. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUROC) were used to evaluate the model’s ability to distinguish RAP from PCV. The performances of the new model, the VGG-16, Resnet-50, Inception, and eight ophthalmologists were compared. A total of 3951 SD-OCT images from 314 participants (229 AMD, 85 normal controls) were analyzed. In distinguishing the PCV and RAP cases, the proposed model showed an accuracy, sensitivity, and specificity of 89.1%, 89.4%, and 88.8%, respectively, with an AUROC of 95.3% (95% CI 0.727–0.852). The proposed model showed better diagnostic performance than VGG-16, Resnet-50, and Inception-V3 and comparable performance with the eight ophthalmologists. The novel model performed well when distinguishing between PCV and RAP. Thus, automated deep learning systems may support ophthalmologists in distinguishing RAP from PCV.

Central serous chorioretinopathy (CSC) is a common condition characterized by serous detachment of the neurosensory retina at the posterior pole. We built a deep learning system model to diagnose CSC, and distinguish chronic from acute CSC using spectral domain optical coherence tomography (SD-OCT) images. Data from SD-OCT images of patients with CSC and a control group were analyzed with a convolutional neural network. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUROC) were used to evaluate the model. For CSC diagnosis, our model showed an accuracy, sensitivity, and specificity of 93.8%, 90.0%, and 99.1%, respectively; AUROC was 98.9% (95% CI, 0.983–0.995); and its diagnostic performance was comparable with VGG-16, Resnet-50, and the diagnoses of five different ophthalmologists. For distinguishing chronic from acute cases, the accuracy, sensitivity, and specificity were 97.6%, 100.0%, and 92.6%, respectively; AUROC was 99.4% (95% CI, 0.985–1.000); performance was better than VGG-16 and Resnet-50, and was as good as the ophthalmologists. Our model performed well when diagnosing CSC and yielded highly accurate results when distinguishing between acute and chronic cases. Thus, automated deep learning system algorithms could play a role independent of human experts in the diagnosis of CSC.

Light element identification is necessary in materials research to obtain detailed insight into various material properties. However, reported techniques, such as scanning transmission electron microscopy (STEM)-energy dispersive X-ray spectroscopy (EDS) have inadequate detection limits, which impairs identification. In this study, we achieved light element identification with nanoscale spatial resolution in a multi-component metal alloy through unsupervised machine learning algorithms of singular value decomposition (SVD) and independent component analysis (ICA). Improvement of the signal-to-noise ratio (SNR) in the STEM-EDS spectrum images was achieved by combining SVD and ICA, leading to the identification of a nanoscale N-depleted region that was not observed in as-measured STEM-EDS. Additionally, the formation of the nanoscale N-depleted region was validated using STEM–electron energy loss spectroscopy and multicomponent diffusional transformation simulation. The enhancement of SNR in STEM-EDS spectrum images by machine learning algorithms can provide an efficient, economical chemical analysis method to identify light elements at the nanoscale.

Cell survival is essential for all living organisms to cope against multiple environmental insults. Intercellular signaling between dying and surviving cells plays an important role to ensure compensatory proliferation, preventing tissue loss after environmental stresses. Here, we show that Sol narae (Sona), a Disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) in Drosophila is required for cell survival. sona exhibited a positive genetic interaction with Death-associated inhibitor of apoptosis 1 (Diap1), and a negative genetic interaction with reaper (rpr). Transcription patterns of sona, Diap1, and rpr genes in the pouch region of wing discs were coordinately changed after irradiation. Interestingly, there was a negative correlation in the expression levels of Sona and DIAP1, and both cell types, one with high Sona level and the other with high Diap1 level, were resistant to irradiation-induced cell death. The sona-expressing cells rarely entered into cell cycle themselves but promoted the nearby cells to proliferate in irradiation conditions. We found that these sona-expressing cells are able to upregulate Cyclin D (Cyc D) and increase tissue size. Furthermore, transient Sona overexpression increased survival rate and promoted development of flies in irradiation conditions. We propose that the two types of radiation-resistant cells, one with high Sona level and the other with high Diap1 level, communicate with dying cells and between each other for cell survival and proliferation in response to irradiation.

In this study, a hydrogel impregnated with powder activated carbon (PAC), MAA-PAC, was synthesized through the polymerization of acrylic acid (AA) and PB was immobilized using the carboxyl group of AA. In this process, an adsorbent with an enhancement of PB content and stability of immobilization was developed through the additional supply of Fe3+ ions by the layer by layer (LBL) assembly. XRD, FT-IR, SEM (EDS), TEM (EDS, mapping), and TG analyzes of the LBL and non-LBL groups were performed to confirm the change of PB content in the adsorbent as the LBL assembly was applied. The stability of PB immobilization was confirmed during the washing process after the synthesis of the adsorbent. When the LBL assembly process was applied as a PB immobilization strategy, the PB content in the adsorbent was improved and PB leakage was not observed during the washing process. The maximum adsorption (qm) for cesium in the MAA-PAC-PB LBL group that showed high PB content was 40.03 mg/g, and the adsorption isotherm was more suitable for the Langmuir model than the Freundlich model. The LBL group showed a high removal efficiency of 99.81% and a high DF value (525.88) for radioactive cesium (120 Bq/g). These results demonstrate the potential efficiency of the MAA-PAC-PB LBL group for the decontamination of radioactive cesium-contaminated water systems. Furthermore, it was verified that the LBL group of MAA-PAC-PB could be used as an adsorbent without an additional design of the existing water treatment facility. This can an economical decontamination method for removing radioactive cesium.

Cell survival is essential for all living organisms to cope against multiple environmental insults. Intercellular signaling between dying and surviving cells plays an important role to ensure compensatory proliferation, preventing tissue loss after environmental stresses. Here, we show that Sol narae (Sona), a Disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) in Drosophila is required for cell survival. sona exhibited a positive genetic interaction with Death-associated inhibitor of apoptosis 1 (Diap1), and a negative genetic interaction with reaper (rpr). Transcription patterns of sona, Diap1, and rpr genes in the pouch region of wing discs were coordinately changed after irradiation. Interestingly, there was a negative correlation in the expression levels of Sona and DIAP1, and both cell types, one with high Sona level and the other with high Diap1 level, were resistant to irradiation-induced cell death. The sona-expressing cells rarely entered into cell cycle themselves but promoted the nearby cells to proliferate in irradiation conditions. We found that these sona-expressing cells are able to upregulate Cyclin D (Cyc D) and increase tissue size. Furthermore, transient Sona overexpression increased survival rate and promoted development of flies in irradiation conditions. We propose that the two types of radiation-resistant cells, one with high Sona level and the other with high Diap1 level, communicate with dying cells and between each other for cell survival and proliferation in response to irradiation.

Quantitative SPECT/CT is potentially useful for more accurate and reliable measurement of glomerular filtration rate (GFR) than conventional planar scintigraphy. However, manual drawing of a volume of interest (VOI) on renal parenchyma in CT images is a labor-intensive and time-consuming task. The aim of this study is to develop a fully automated GFR quantification method based on a deep learning approach to the 3D segmentation of kidney parenchyma in CT. We automatically segmented the kidneys in CT images using the proposed method with remarkably high Dice similarity coefficient relative to the manual segmentation (mean = 0.89). The GFR values derived using manual and automatic segmentation methods were strongly correlated (R2 = 0.96). The absolute difference between the individual GFR values using manual and automatic methods was only 2.90%. Moreover, the two segmentation methods had comparable performance in the urolithiasis patients and kidney donors. Furthermore, both segmentation modalities showed significantly decreased individual GFR in symptomatic kidneys compared with the normal or asymptomatic kidney groups. The proposed approach enables fast and accurate GFR measurement.

Scientific Reports 5: Article number: 15993; published online: 06 January 2016; updated: 11 November 2016

ADAMTS (a disintegrin and metalloproteases with thrombospondin motif) family consists of secreted proteases, and is shown to cleave extracellular matrix proteins. Their malfunctions result in cancers and disorders in connective tissues. We report here that a Drosophila ADAMTS named Sol narae (Sona) promotes Wnt/Wingless (Wg) signaling. sona loss-of-function mutants are lethal and rare escapers had malformed appendages, indicating that sona is essential for fly development and survival. sona exhibited positive genetic interaction with wntless (wls) that encodes a cargo protein for Wg. Loss of sona decreased the level of extracellular Wg, and also reduced the expression level of Wg effector proteins such as Senseless (Sens), Distalless (Dll) and Vestigial (Vg). Sona and Wg colocalized in Golgi and endosomal vesicles, and were in the same protein complex. Furthermore, co-expression of Wg and Sona generated ectopic wing margin bristles. This study suggests that Sona is involved in Wg signaling by regulating the level of extracellular Wg.

The strain states of [111]-, [110]-, and [002]-oriented grains in poly-crystalline sputtered (Ba,Sr)TiO(3) thin films on highly [111]-oriented Pt electrode/Si substrates were carefully examined by X-ray diffraction techniques. Remarkably, [002]-oriented grains respond more while [110]- and [111]-oriented grains do less than the theoretically estimated responses, which is understandable from the arrangement of the TiO(6) octahedra with respect to the stress direction. Furthermore, such mechanical responses are completely independent of the degree of crystallization and film thickness. The transition growth temperature between the positive and negative strains was also different depending on the grain orientation. The unstrained lattice parameter for each type of grain was different suggesting that the oxygen vacancy concentration for each type of grain is different, too. The results reveal that polycrystalline (Ba,Sr)TiO(3) thin films are not an aggregation of differently oriented grains which simply follow the mechanical behavior of single crystal with different orientations.