Usefulness of longitudinal nodule-matching algorithm in computer-aided diagnosis of new pulmonary metastases on cancer surveillance CT scans.

Background: Detecting new pulmonary metastases by comparing serial computed tomography (CT) scans is crucial, but a repetitive and time-consuming task that burdens the radiologists' workload. This study aimed to evaluate the usefulness of a nodule-matching algorithm with deep learning-based computer-aided detection (DL-CAD) in diagnosing new pulmonary metastases on cancer surveillance CT scans. Methods: Among patients who underwent pulmonary metastasectomy between 2014 and 2018, 65 new pulmonary metastases missed by interpreting radiologists on cancer surveillance CT (Time 2) were identified after a retrospective comparison with the previous CT (Time 1). First, DL-CAD detected nodules in Time 1 and Time 2 CT images. All nodules detected at Time 2 were initially considered metastasis candidates. Second, the nodule-matching algorithm was used to assess the correlation between the nodules from the two CT scans and to classify the nodules at Time 2 as "new" or "pre-existing". Pre-existing nodules were excluded from metastasis candidates. We evaluated the performance of DL-CAD with the nodule-matching algorithm, based on its sensitivity, false-metastasis candidates per scan, and positive predictive value (PPV). Results: A total of 475 lesions were detected by DL-CAD at Time 2. Following a radiologist review, the lesions were categorized as metastases (n=54), benign nodules (n=392), and non-nodules (n=29). Upon comparison of nodules at Time 1 and 2 using the nodule-matching algorithm, all metastases were classified as new nodules without any matching errors. Out of 421 benign lesions, 202 (48.0%) were identified as pre-existing and subsequently excluded from the pool of metastasis candidates through the nodule-matching algorithm. As a result, false-metastasis candidates per CT scan decreased by 47.9% (from 7.1 to 3.7, P<0.001) and the PPV increased from 11.4% to 19.8% (P<0.001), while maintaining sensitivity. Conclusions: The nodule-matching algorithm improves the diagnostic performance of DL-CAD for new pulmonary metastases, by lowering the number of false-metastasis candidates without compromising sensitivity.

Bone Suppression on Chest Radiographs for Pulmonary Nodule Detection: Comparison between a Generative Adversarial Network and Dual-Energy Subtraction.

OBJECTIVE: To compare the effects of bone suppression imaging using deep learning (BSp-DL) based on a generative adversarial network (GAN) and bone subtraction imaging using a dual energy technique (BSt-DE) on radiologists' performance for pulmonary nodule detection on chest radiographs (CXRs). MATERIALS AND METHODS: A total of 111 adults, including 49 patients with 83 pulmonary nodules, who underwent both CXR using the dual energy technique and chest CT, were enrolled. Using CT as a reference, two independent radiologists evaluated CXR images for the presence or absence of pulmonary nodules in three reading sessions (standard CXR, BSt-DE CXR, and BSp-DL CXR). Person-wise and nodule-wise performances were assessed using receiver-operating characteristic (ROC) and alternative free-response ROC (AFROC) curve analyses, respectively. Subgroup analyses based on nodule size, location, and the presence of overlapping bones were performed. RESULTS: BSt-DE with an area under the AFROC curve (AUAFROC) of 0.996 and 0.976 for readers 1 and 2, respectively, and BSp-DL with AUAFROC of 0.981 and 0.958, respectively, showed better nodule-wise performance than standard CXR (AUAFROC of 0.907 and 0.808, respectively; p ≤ 0.005). In the person-wise analysis, BSp-DL with an area under the ROC curve (AUROC) of 0.984 and 0.931 for readers 1 and 2, respectively, showed better performance than standard CXR (AUROC of 0.915 and 0.798, respectively; p ≤ 0.011) and comparable performance to BSt-DE (AUROC of 0.988 and 0.974; p ≥ 0.064). BSt-DE and BSp-DL were superior to standard CXR for detecting nodules overlapping with bones (p < 0.017) or in the upper/middle lung zone (p < 0.017). BSt-DE was superior (p < 0.017) to BSp-DL in detecting peripheral and sub-centimeter nodules. CONCLUSION: BSp-DL (GAN-based bone suppression) showed comparable performance to BSt-DE and can improve radiologists' performance in detecting pulmonary nodules on CXRs. Nevertheless, for better delineation of small and peripheral nodules, further technical improvements are required.

Deep learning-based denoising algorithm in comparison to iterative reconstruction and filtered back projection: a 12-reader phantom study.

OBJECTIVES: (1) To compare low-contrast detectability of a deep learning-based denoising algorithm (DLA) with ADMIRE and FBP, and (2) to compare image quality parameters of DLA with those of reconstruction methods from two different CT vendors (ADMIRE, IMR, and FBP). MATERIALS AND METHODS: Using abdominal CT images of 100 patients reconstructed via ADMIRE and FBP, we trained DLA by feeding FBP images as input and ADMIRE images as the ground truth. To measure the low-contrast detectability, the randomized repeat scans of Catphan® phantom were performed under various conditions of radiation exposures. Twelve radiologists evaluated the presence/absence of a target on a five-point confidence scale. The multi-reader multi-case area under the receiver operating characteristic curve (AUC) was calculated, and non-inferiority tests were performed. Using American College of Radiology CT accreditation phantom, contrast-to-noise ratio, target transfer function, noise magnitude, and detectability index (d') of DLA, ADMIRE, IMR, and FBPs were computed. RESULTS: The AUC of DLA in low-contrast detectability was non-inferior to that of ADMIRE (p < .001) and superior to that of FBP (p < .001). DLA improved the image quality in terms of all physical measurements compared to FBPs from both CT vendors and showed profiles of physical measurements similar to those of ADMIRE. CONCLUSIONS: The low-contrast detectability of the proposed deep learning-based denoising algorithm was non-inferior to that of ADMIRE and superior to that of FBP. The DLA could successfully improve image quality compared with FBP while showing the similar physical profiles of ADMIRE. KEY POINTS: • Low-contrast detectability in the images denoised using the deep learning algorithm was non-inferior to that in the images reconstructed using standard algorithms. • The proposed deep learning algorithm showed similar profiles of physical measurements to advanced iterative reconstruction algorithm (ADMIRE).

Robustness of Deep Learning Algorithm to Varying Imaging Conditions in Detecting Low Contrast Objects In Computed Tomography Phantom Images: in Comparison to 12 Radiologists.

The performance of deep learning algorithm (DLA) to that of radiologists was compared in detecting low contrast objects in CT phantom images under various imaging conditions. For training, 10,000 images were created using American College of Radiology CT phantom as the background. In half of the images, objects of 3-20 mm size and 5-30 HU contrast difference were generated in random locations. Binary responses were used as the ground truth. For testing, 640 images of Catphan® phantom were used, half of which had objects of either 5 or 9 mm size with 10 HU contrast difference. Twelve radiologists evaluated the presence of objects on a five-point scale. The performances of the DLA and radiologists were compared across different imaging conditions in terms of area under receiver operating characteristics curve (AUC). Multi-reader multi-case AUC and Hanley and McNeil tests were used. We performed post-hoc analysis using bootstrapping and verified that the DLA is less affected by the changing imaging conditions. The AUC of DLA was consistently higher than those of the radiologists across different imaging conditions (p < 0.0001), and it was less affected by varying imaging conditions. The DLA outperformed the radiologists and showed more robust performance under varying imaging conditions.

Deep Learning for Diagnosis of Paranasal Sinusitis Using Multi-View Radiographs.

Accurate image interpretation of Waters' and Caldwell view radiographs used for sinusitis screening is challenging. Therefore, we developed a deep learning algorithm for diagnosing frontal, ethmoid, and maxillary sinusitis on both Waters' and Caldwell views. The datasets were selected for the training and validation set (n = 1403, sinusitis% = 34.3%) and the test set (n = 132, sinusitis% = 29.5%) by temporal separation. The algorithm can simultaneously detect and classify each paranasal sinus using both Waters' and Caldwell views without manual cropping. Single- and multi-view models were compared. Our proposed algorithm satisfactorily diagnosed frontal, ethmoid, and maxillary sinusitis on both Waters' and Caldwell views (area under the curve (AUC), 0.71 (95% confidence interval, 0.62-0.80), 0.78 (0.72-0.85), and 0.88 (0.84-0.92), respectively). The one-sided DeLong's test was used to compare the AUCs, and the Obuchowski-Rockette model was used to pool the AUCs of the radiologists. The algorithm yielded a higher AUC than radiologists for ethmoid and maxillary sinusitis (p = 0.012 and 0.013, respectively). The multi-view model also exhibited a higher AUC than the single Waters' view model for maxillary sinusitis (p = 0.038). Therefore, our algorithm showed diagnostic performances comparable to radiologists and enhanced the value of radiography as a first-line imaging modality in assessing multiple sinusitis.

Low-Dose Abdominal CT Using a Deep Learning-Based Denoising Algorithm: A Comparison with CT Reconstructed with Filtered Back Projection or Iterative Reconstruction Algorithm.

OBJECTIVE: To compare the image quality of low-dose (LD) computed tomography (CT) obtained using a deep learning-based denoising algorithm (DLA) with LD CT images reconstructed with a filtered back projection (FBP) and advanced modeled iterative reconstruction (ADMIRE). MATERIALS AND METHODS: One hundred routine-dose (RD) abdominal CT studies reconstructed using FBP were used to train the DLA. Simulated CT images were made at dose levels of 13%, 25%, and 50% of the RD (DLA-1, -2, and -3) and reconstructed using FBP. We trained DLAs using the simulated CT images as input data and the RD CT images as ground truth. To test the DLA, the American College of Radiology CT phantom was used together with 18 patients who underwent abdominal LD CT. LD CT images of the phantom and patients were processed using FBP, ADMIRE, and DLAs (LD-FBP, LD-ADMIRE, and LD-DLA images, respectively). To compare the image quality, we measured the noise power spectrum and modulation transfer function (MTF) of phantom images. For patient data, we measured the mean image noise and performed qualitative image analysis. We evaluated the presence of additional artifacts in the LD-DLA images. RESULTS: LD-DLAs achieved lower noise levels than LD-FBP and LD-ADMIRE for both phantom and patient data (all p < 0.001). LD-DLAs trained with a lower radiation dose showed less image noise. However, the MTFs of the LD-DLAs were lower than those of LD-ADMIRE and LD-FBP (all p < 0.001) and decreased with decreasing training image dose. In the qualitative image analysis, the overall image quality of LD-DLAs was best for DLA-3 (50% simulated radiation dose) and not significantly different from LD-ADMIRE. There were no additional artifacts in LD-DLA images. CONCLUSION: DLAs achieved less noise than FBP and ADMIRE in LD CT images, but did not maintain spatial resolution. The DLA trained with 50% simulated radiation dose showed the best overall image quality.

Residual Error Based Anomaly Detection Using Auto-Encoder in SMD Machine Sound.

Detecting an anomaly or an abnormal situation from given noise is highly useful in an environment where constantly verifying and monitoring a machine is required. As deep learning algorithms are further developed, current studies have focused on this problem. However, there are too many variables to define anomalies, and the human annotation for a large collection of abnormal data labeled at the class-level is very labor-intensive. In this paper, we propose to detect abnormal operation sounds or outliers in a very complex machine along with reducing the data-driven annotation cost. The architecture of the proposed model is based on an auto-encoder, and it uses the residual error, which stands for its reconstruction quality, to identify the anomaly. We assess our model using Surface-Mounted Device (SMD) machine sound, which is very complex, as experimental data, and state-of-the-art performance is successfully achieved for anomaly detection.

Changes in the expression of transthyretin and protein kinase Cγ genes in the prefrontal cortex in response to naltrexone.

Naltrexone, an opioid receptor antagonist, has been approved for clinical use in the treatment of alcohol dependence. In the present study, we examined the underlying mechanisms of naltrexone by investigating the pharmacogenomic variations in the brain regions associated with alcohol consumption. A complementary DNA microarray analysis was used to profile gene expression changes in the hippocampus and prefrontal cortex (PFC) of C57BL/6 mice injected with naltrexone following ethanol treatment. Intraperitoneal administration of 200µl (16mg/kg) of naltrexone for 4 weeks caused alterations in the expression of a wide range of hippocampal (394) and PFC (566) genes in ethanol-treated mice. Ingenuity Pathway Analysis (IPA) software was used to search for biological pathways and interrelationships between gene networks in the subsets of candidate genes that were altered in the naltrexone-treated PFC and hippocampus. We found gene networks associated with cell morphology, cell death, nervous system development and function, and neurological disease. Confirmation studies using quantitative reverse transcription-polymerase chain reaction (qRT-PCR) revealed that the expression of transthyretin (TTR) and protein kinase C (PKC)γ were increased in the PFC but not in the hippocampus of naltrexone-treated mice. In conclusion, the present study demonstrates a pharmacogenomic response to naltrexone in the brains of ethanol-consuming mice. These findings provide a basis for conducting pharmacogenetic research on the effect of naltrexone in specific brain areas, which would enhance our understanding of the underlying causes and possible treatments of alcohol use disorders.

Alcohol Use during Pregnancy and Related Risk Factors in Korea.

OBJECTIVE: The number of Korean women of childbearing age who drink alcohol and binge drink has increased remarkably in recent years. In the present study, we examined self-reported rates of alcohol use before and during pregnancy and identified maternal characteristics associated with drinking in pregnancy. METHODS: One thousand pregnant Korean women who visited the Department of Obstetrics and Gynecology (OB/GYN) completed a self-administered questionnaire that sought information on their demographic characteristics and incorporated features of the Alcohol Use Disorder Identification Test (AUDIT)-C to investigate their use of alcohol, including binge drinking, during three time periods ("in the year before this pregnancy," "during this pregnancy," and "in the previous 30 days"). RESULTS: Of these participants, 16.4% reported using alcohol during their pregnancy, 12.2% had used alcohol in the previous 30 days, and 1.7% reported binge drinking during their pregnancy. In the year before pregnancy, 77.1% had used alcohol, and 22.3% had binge drunk. The group using any amount of any alcohol during pregnancy showed a lower educational level, a lower rate of planned pregnancy, a lower level of knowledge relating to the risks of drinking alcohol during pregnancy, and a higher frequency of alcohol drinking in the year before pregnancy when compared with the abstinent group. Low educational level and unplanned pregnancy were revealed to be significant risk factors for alcohol consumption in pregnant women. CONCLUSION: This is the first study to examine any alcohol and binge alcohol drinking during pregnancy in Korea. Clinical attention and monitoring system on alcohol use during pregnancy are necessary in Korea.

Artificial microRNA-based neurokinin-1 receptor gene silencing reduces alcohol consumption in mice.

In the brain, the stress system plays an important role in motivating continued alcohol use and relapse. The neuropeptide substance P and the neurokinin-1 receptor (NK1R) are involved in the stress response and drug reward systems. Recent findings have shown that the binding of ligands to NK1Rs decreases the self-administration of alcohol in mice. We examined the effect of an artificial microRNA (amiRNA) on the functional expression of NK1R in mouse brains. Lentiviruses expressing either an amiRNA targeting the NK1R (amiNK1R) or a negative control amiRNA (amiNC) were injected into mouse brains. Four weeks after amiRNA injection, we found that amiNK1R decreased the voluntary alcohol consumption compared to mice injected with amiNC. We also observed that NK1R expression was reduced in the hippocampus. RNA interference is an effective approach to regulate the expression of specific behavior-related genes. Our results support the potential use of amiRNA as a therapeutic agent for the treatment of alcohol dependence.

Relation between plasma brain-derived neurotrophic factor and nerve growth factor in the male patients with alcohol dependence.

Brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) are thought to be related to neuroprotection in cell culture and animal studies. Our aim was to verify the changes in human plasma BDNF and NGF concentrations induced by chronic alcohol use. Forty-one male patients with alcohol dependence were sampled the next morning of admission and compared with 41 healthy male subjects. Plasma BDNF and NGF were assayed using an enzyme-linked immunosorbent assay (ELISA). Mean plasma BDNF level was significantly higher in the patients with alcohol dependence (3502.21+/-1726.9 pg/mL) compared with the healthy subjects (861.75+/-478.9 pg/mL) (P=.000). Mean plasma NGF level was also significantly higher in patients with alcohol dependence (137.64+/-32.7 pg/mL) than in healthy subjects (112.61+/-90.2 pg/mL) (P=.012). Plasma BDNF and NGF levels showed significant negative correlation in alcohol dependence group (r=-0.388, P=.012). Increased plasma BDNF and NGF with negative correlation in alcohol-dependent patients may have some role in the regeneration of damage done by chronic alcohol use.

Association between Dopamine D4 Receptor Gene Polymorphism and Scores on a Continuous Performance Test in Korean Children with Attention Deficit Hyperactivity Disorder.

OBJECTIVE: The aim of this study was to evaluate the association between a variable number of tandem repeats polymorphism at the dopamine D4 receptor gene (DRD4) and the performance of children with attention deficit hyperactivity disorder (ADHD) in a continuous performance test (CPT). METHODS: This study included 72 ADHD children (mean age=9.39+/-2.05 years) who were recruited from one child psychiatric clinic. The omission errors, commission errors, reaction time and reaction standardization in the CPT were computed. The number of 48-base pairs tandem repeats in the exon III of DRD4 was analyzed in a blind manner. RESULTS: The homozygosity of the 4-repeat allele at DRD4 was significantly associated with fewer commission errors (t=2.364, df=28.685, p=0.025) and standard deviation of reaction time (t=2.351, df=24.648, p=0.027) even after adjusting for age. The results of analyses of CPT measured values among three groups showed that the group with higher frequency of the 4-repeat allele showed a lower mean score of commission errors (F=4.268, df=2, p=0.018). CONCLUSION: These results suggest a protective role of 4-repeat allele of the DRD4 polymorphisms on commission errors in the CPT in children with ADHD.

Mirtazapine for patients with alcohol dependence and comorbid depressive disorders: a multicentre, open label study.

Major depressive disorder and alcohol dependence are common and serious mental illnesses. There is a great interest in discovering useful treatments for both mood symptoms and alcohol abuse in those patients with depressive disorders and comorbid alcohol dependence. The primary purpose of this study was to evaluate the effectiveness and tolerability of mirtazapine for the treatment of patients with alcohol dependence comorbid with a depressive disorder in an open label, naturalistic multicentre treatment setting. The 17-item Hamilton Depression Rating Scale (HDRS), the Hamilton Anxiety Rating Scale (HARS) and the Clinical Global Impression-Severity (CGI-S) scale were measured at baseline and at weeks 4 and 8 for the assessment of treatment effectiveness. Alcohol craving was measured using the Obsessive Compulsive Drinking Scale (OCDS) and the Visual Analog Scale for Craving (VAS). This study showed a statistically significant reduction of the scores on the HDRS (13.9+/-7.3, p<0.0001), HARS (10.8+/-7.2, p<0.0001) and the CGI-S (1.7+/-1.0, p<0.0001) from baseline to the endpoint (week 8). The OCDS and VAS scores were also decreased significantly by 42.3% and 53.2% (9.0+/-10.0, p<0.0001; 2.5+/-2.4, p<0.0001, respectively). The number of patients with a 50% reduction or more in the HDRS and HARS scores was 103 (72.0%) and 106 (74.1%) at the endpoint, respectively. Adverse events related to mirtazapine were observed in 10% or more of the patients in this study. In conclusion, the results from this naturalistic study suggest that the use of mirtazapine for the patients with alcohol dependence comorbid with depressive disorder is accompanied by clinical improvement in their mood and alcohol craving.

Naltrexone influences protein kinase Ce and integrin alpha7 activity in SH-SY5Y neuroblastoma cells.

Alcohol influences the neuroadaptation of brain cells where receptors and enzymes like protein kinase C (PKC) exist. Naltrexone acts on opioid receptors. However, other mechanisms of action remain unknown. We prepared SH-SY5Y neuroblastoma cells, and fed them with 150 mM ethanol for 72 h followed by treatment with naltrexone for 24 h. We performed microarray analysis and reverse transcriptase-polymerase chain reaction. Our results showed that PKCepsilon increased 1.90 times and showed an overall decreasing pattern as time increased. Phosphorylated ERK also increased 2.0 times according to the change of PKCepsilon. Integrin alpha7 increased 2.32 times and showed an increasing pattern as time increased. In conclusion, naltrexone influences PKCepsilon neuronal signaling system and endothelial adhesion molecule integrin alpha7 in addition to the well-known opioid system.

Alcohol and aldehyde dehydrogenase polymorphisms in men with type I and Type II alcoholism.

OBJECTIVE: The authors examined the genetic polymorphisms of alcohol dehydrogenase 2 and 3 (ADH2 and ADH3) and aldehyde dehydrogenase (ALDH2) in patients diagnosed as having Cloninger's type I or type II alcoholism. METHOD: Seventy-two alcoholic men and 38 nonalcoholic, healthy men were tested for the distribution of genotypes and alleles of ADH2, ADH3, and ALDH2. Forty-eight of the alcoholic men had type I alcoholism, and 24 had type II alcoholism. RESULTS: The frequencies of ADH2*1 and ADH3*2 alleles were significantly higher in men with type II alcoholism than in men with type I alcoholism and healthy men. The frequency of the ALDH2*1 allele was significantly higher in men with alcohol dependence than in healthy men. CONCLUSIONS: The genetic characteristics of alcohol dehydrogenases in men with type I alcoholism were similar to those of healthy men, and the genetic characteristics of aldehyde dehydrogenase in men with type I alcoholism were similar to those of men with type II alcoholism. These findings suggest that the genetic characteristics of alcohol metabolism in type I alcoholism fall between nonalcoholism and type II alcoholism.

Fluoxetine-induced up-regulation of 14-3-3zeta and tryptophan hydroxylase levels in RBL-2H3 cells.

The primary mechanisms of antidepressants are based on the monoamine depletion hypothesis. However, we do not yet know the full cascade of mechanisms responsible for the therapeutic effect of antidepressants. To identify the genes involved in the therapeutic mechanism of the selective serotonin reuptake inhibitor, fluoxetine, we used a cDNA microarray analysis with RBL-2H3 cells. We observed the transcriptional changes of several tens of genes containing the 14-3-3zeta gene in the fluoxetine-treated RBL-2H3 cells. Real-time RT-PCR and Western blotting confirmed changes in the expression of the gene and protein. The increase of 14-3-3zeta mRNA was observed at 72 h in the fluoxetine-treated RBL-2H3 cells. The increase of 14-3-3zeta protein was observed at 48 and 72 h. In this study, the expressions of the 14-3-3zeta gene and the protein were up-regulated at 72 h. In addition, the increase of TPH mRNA was observed at 12, 24 and 72 h in the fluoxetine-treated RBL-2H3 cells. We conclude that fluoxetine induces increases of 14-3-3zeta mRNA, 14-3-3zeta protein and TPH mRNA at 72 h in the RBL-2H3 cells. This suggests that the 14-3-3zeta and TPH genes may play a role in the molecular mechanism of fluoxetine. To date, no cases of 14-3-3zeta alterations by antidepressants and specifically by fluoxetine have been reported.

The functional domains of dopamine transporter for cocaine analog, CFT binding.

Cocaine analogue, CFT (2beta-carbomethoxy-3beta-(4-fluorophenyl) tropane) binding to dopamine transporter (DAT) in different species is quite heterogeneous. CFT is scarcely detected in bovine DAT whereas it is conspicuous in humans. To examine the structural basis for this functional discrepancy, we analyzed transporter chimeras of these two DATs. The CFT binding activities are avid in all of the chimeric DATs of which both of the 3rd and the 6-8th transmembrane domain (TM) are composed of human DAT sequences. On the contrary, CFT binding activities were scarcely detected if either or both of two regions are replaced with bovine sequences. These findings indicate that the CFT binding absolutely requires human DAT sequences, at least, in the regions encompassing the 3rd and 6-8th transmembrane domain (TM), and that these regions might contribute to form the 3-dimensional pocket for CFT binding.