Liver disease classification from ultrasound using multi-scale CNN.

PURPOSE: Ultrasound (US) is the preferred modality for fatty liver disease diagnosis due to its noninvasive, real-time, and cost-effective imaging capabilities. However, traditional B-mode US is qualitative, and therefore, the assessment is very subjective. Computer-aided diagnostic tools can improve the specificity and sensitivity of US and help clinicians to perform uniform diagnoses. METHODS: In this work, we propose a novel deep learning model for nonalcoholic fatty liver disease classification from US data. We design a multi-feature guided multi-scale residual convolutional neural network (CNN) architecture to capture features of different receptive fields. B-mode US images are combined with their corresponding local phase filtered images and radial symmetry transformed images as multi-feature inputs for the network. Various fusion strategies are studied to improve prediction accuracy. We evaluate the designed network architectures on B-mode in vivo liver US images collected from 55 subjects. We also provide quantitative results by comparing our proposed multi-feature CNN architecture against traditional CNN designs and machine learning methods. RESULTS: Quantitative results show an average classification accuracy above 90% over tenfold cross-validation. Our proposed method achieves a 97.8% area under the ROC curve (AUC) for the patient-specific leave-one-out cross-validation (LOOCV) evaluation. Comprehensive validation results further demonstrate that our proposed approaches achieve significant improvements compared to training mono-feature CNN architectures ([Formula: see text]). CONCLUSIONS: Feature combination is valuable for the traditional classification methods, and the use of multi-scale CNN can improve liver classification accuracy. Based on the promising performance, the proposed method has the potential in practical applications to help radiologists diagnose nonalcoholic fatty liver disease.

Chest X-ray image phase features for improved diagnosis of COVID-19 using convolutional neural network.

PURPOSE: Recently, the outbreak of the novel coronavirus disease 2019 (COVID-19) pandemic has seriously endangered human health and life. In fighting against COVID-19, effective diagnosis of infected patient is critical for preventing the spread of diseases. Due to limited availability of test kits, the need for auxiliary diagnostic approach has increased. Recent research has shown radiography of COVID-19 patient, such as CT and X-ray, contains salient information about the COVID-19 virus and could be used as an alternative diagnosis method. Chest X-ray (CXR) due to its faster imaging time, wide availability, low cost, and portability gains much attention and becomes very promising. In order to reduce intra- and inter-observer variability, during radiological assessment, computer-aided diagnostic tools have been used in order to supplement medical decision making and subsequent management. Computational methods with high accuracy and robustness are required for rapid triaging of patients and aiding radiologist in the interpretation of the collected data. METHOD: In this study, we design a novel multi-feature convolutional neural network (CNN) architecture for multi-class improved classification of COVID-19 from CXR images. CXR images are enhanced using a local phase-based image enhancement method. The enhanced images, together with the original CXR data, are used as an input to our proposed CNN architecture. Using ablation studies, we show the effectiveness of the enhanced images in improving the diagnostic accuracy. We provide quantitative evaluation on two datasets and qualitative results for visual inspection. Quantitative evaluation is performed on data consisting of 8851 normal (healthy), 6045 pneumonia, and 3323 COVID-19 CXR scans. RESULTS: In Dataset-1, our model achieves 95.57% average accuracy for a three classes classification, 99% precision, recall, and F1-scores for COVID-19 cases. For Dataset-2, we have obtained 94.44% average accuracy, and 95% precision, recall, and F1-scores for detection of COVID-19. CONCLUSIONS: Our proposed multi-feature-guided CNN achieves improved results compared to single-feature CNN proving the importance of the local phase-based CXR image enhancement. Future work will involve further evaluation of the proposed method on a larger-size COVID-19 dataset as they become available.

Liver biopsy in assessment of extended criteria donors.

The safety and liver utilization with prerecovery liver biopsy (PLB) in extended criteria liver donors are unclear. We conducted a retrospective cohort study in 1323 brain death donors (PLB = 496) from 3 organ procurement organizations (OPOs). Outcomes were complications, preempted liver recovery (PLR), and liver transplantation (LT). Additional analyses included liver-only and propensity score-matched multiorgan donor subgroups. PLB donors were older (57 versus 53 years; P < 0.001). Hepatitis C antibody positivity (14.3% versus 9.6%, P = 0.01) and liver-only donors (42.6% versus 17.5%; P < 0.001) were more prevalent. The PLB cohort had fewer complications (31.9% versus 42.3%; P < 0.001). In the PLB cohort, PLR was significantly higher (odds ratio [OR], 3.45; 95% confidence interval [CI], 2.42-4.92) and LT lower (OR, 0.69; 95% CI, 0.52-0.91). In liver-only and propensity score-matched multiorgan donor subgroups, PLR was significantly higher (OR, 1.76; 95% CI, 1.06-2.94 and OR, 2.29; 95% CI, 1.37-3.82, respectively) without a decrease in LT (OR, 0.71; 95% CI, 0.43-1.18 and OR, 0.91; 95% CI, 0.63-1.33, respectively) in PLB subgroups. In conclusion, in extended criteria liver donors, PLB is safe and decreases futile liver recovery without decreasing LT. Increased use of PLB, especially in liver-only donors, is likely to save costs to OPOs and transplant centers and improve efficiencies in organ allocation. Liver Transplantation 24 182-191 2018 AASLD.

Steroid-Mediated Decrease in Blood Mesenchymal Stem Cells in Liver Transplant could Impact Long-Term Recovery.

Orthotopic liver transplant (OLT) remains the standard of care for end stage liver disease. To circumvent allo-rejection, OLT subjects receive gluococorticoids (GC). We investigated the effects of GC on endogenous mesenchymal stem (stromal) cells (MSCs) in OLT. This question is relevant because MSCs have regenerative potential and immune suppressor function. Phenotypic analyses of blood samples from 12 OLT recipients, at pre-anhepatic, anhepatic and post-transplant (2 h, Days 1 and 5) indicated a significant decrease in MSCs after GC injection. The MSCs showed better recovery in the blood from subjects who started with relatively low MSCs as compared to those with high levels at the prehepatic phase. This drop in MSCs appeared to be linked to GC since similar change was not observed in liver resection subjects. In order to understand the effects of GC on decrease MSC migration, in vitro studies were performed in transwell cultures. Untreated MSCs could not migrate towards the GC-exposed liver tissue, despite CXCR4 expression and the production of inflammatory cytokines from the liver cells. GC-treated MSCs were inefficient with respect to migration towards CXCL12, and this correlated with retracted cytoskeleton and motility. These dysfunctions were partly explained by decreases in the CXCL12/receptor axis. GC-associated decrease in MSCs in OLT recipients recovered post-transplant, despite poor migratory ability towards GC-exposed liver. In total, the study indicated that GC usage in transplant needs to be examined to determine if this could be reduced or avoided with adjuvant cell therapy.

Organ Procurement Organization Survey of Practices and Beliefs Regarding Prerecovery Percutaneous Liver Biopsy in Donation After Neurologic Determination of Death.

BACKGROUND: Prerecovery liver biopsy (PLB) allows histological evaluation of the organ before procurement. The opinions and what factors might influence PLB use within Organ Procurement Organizations (OPOs) are unknown. METHODS: A survey instrument was distributed by the Association of OPOs to the clinical directors of all 58 OPOs. Descriptive statistics were calculated. Results were also stratified based on OPO characteristics. RESULTS: Forty-nine (84.5%) of 58 OPOs responded to the survey; 40 (81.6%) of 49 currently perform PLB. This did not vary based on land mass, population, livers discarded, transplanted, donor age, or recipient MELD scores. Donor age, obesity, alcohol abuse, hepatitis serology, liver only donor, imaging results, and transplant center request were the most common indications for PLB in over 80% of OPOs. The median rate of performance is 5% to 10% of donors. Most use interventional radiologists to perform and the donor hospital pathologist/s to interpret PLB. Most OPOs believe PLBs are safe, reliable, useful, and performed often enough. Most say they did not believe they are easy to obtain. Beliefs were mixed regarding accuracy. The topics likely to influence PLB use were utility and accuracy of PLB, and availability of staff to perform PLB. OPOs that perform PLB more often were more likely to have favorable opinions of safety and pathologist availability, and more influenced by safety, reliability, availability, and a national consensus on the use of PLB. CONCLUSIONS: Considerable variability exists in the use of PLB. Additional information on the utility, accuracy, and safety of PLB are needed to optimize its use.

A comparison of long-term outcomes of portal versus systemic venous drainage in pancreatic transplantation: a systematic review and meta-analysis.

Pancreas transplantation venous effluent can be drained via the portal vein or the systemic circulation; however, no recommendation exists for the ideal technique. A systematic review of the literature from 1989 through 2014 using PubMed, CINHAL, and Cochrane Library for portal versus systemic venous drainage was undertaken. Only studies on humans and published in English were considered. Measures of glycemic control and total cholesterol were synthesized for meta-analysis utilizing random-effects models. Of 166 articles retrieved, 15 articles were included for meta-analysis. Patient and graft survival were comparable in a large database study as well as in the only randomized control study. No differences in complications were seen when exocrine drainage was enteric for the systemic venous group. Fasting insulin (-34.13 pmol/mL, p < 0.001) was significantly lower within the portal drained group; however, fasting blood glucose levels (-3.4 mg/dL, p = 0.32) and hemoglobin A1C levels (mean difference 0.124%, p = 0.25) were comparable. Total cholesterol levels (-3.62 mg/dL, p = 0.447), as well as other measures of lipids, showed no difference. Based on this systematic review and meta-analysis, there is no evidence of differences in outcomes or metabolic control in patients undergoing pancreatic transplant with portal venous drainage compared to the systemic venous drainage.

Prerecovery liver biopsy in the brain-dead donor: a case-control study of logistics, safety, precision, and utility.

Prerecovery liver biopsy (PLB) can potentially to decrease futile recovery and increase utilization of marginal brain-dead donor (BDD) livers. A case-control study was conducted to examine the logistics, safety, histological precision, and liver utilization associated with PLB in BDDs. Twenty-three cases between January 2008 and January 2013 were compared to 2 groups: 48 sequential and 69 clinically matched controls. Compared to the sequential controls, the cases were older (53 versus 46 years), heavier (30.2 versus 25.8 kg/m2), had higher prevalences of hypertension (78.3% versus 44.7%) and alcohol use (56.5% versus 23.4%), and a lower United Network for Organ Sharing expected organ yield (0.73 versus 0.81 livers/donor; P < 0.05 for all). Baseline characteristics were similar between cases and clinical controls. Donor management time was longer for the cases (22.4 hours) versus sequential controls (16.5 hours, P = 0.01) and clinical controls (15.9 hours, P = 0.01). Complications for cases (8.7%) were not different from either group of controls (18.8% for sequential controls, P = 0.46; 17.4% for clinical controls, P = 0.50). The agreement between the donor hospital and study pathologists was substantial regarding evaluation of steatosis (κ = 0.623) and fibrosis (κ = 0.627) and moderate regarding inflammation (κ = 0.495). The proportions of livers that were transplanted were similar for the cases and the clinical controls (60.9% versus 59.4%). In contrast, the proportion of donors for whom liver recovery was not attempted was higher (30.4% versus 8.7%), and the proportion of attempted liver recoveries that did not result in transplantation was lower (8.7% versus 31.9%). These differences were significant at P = 0.009. Overall, PLB is logistically feasible with only a minimal delay and is safe, its interpretation at donor hospitals is reproducible, and it appears to decrease futile liver recovery.

Mini-incision right hepatic lobectomy with or without laparoscopic assistance for living donor hepatectomy.

Minimally invasive procedures are considered to be safe and effective approaches to the management of surgical liver disease. However, this indication remains controversial for living donor hepatectomy. Between 2000 and 2011, living donor right hepatectomy (LDRH) was performed 58 times. Standard right hepatectomy was performed in 30 patients via a subcostal incision with a midline extension. Minimally invasive procedures began to be used for LDRH in 2008. A hybrid technique (hand-assisted laparoscopic liver mobilization and minilaparotomy for parenchymal dissection) was developed and used in 19 patients. In 2010, an upper midline incision (10 cm) without laparoscopic assistance for LDRH was innovated, and this technique was used in 9 patients. The perioperative factors were compared, and the indications for minimally invasive LDRH were investigated. The operative blood loss was significantly less for the patients undergoing a minimally invasive procedure versus the patients undergoing the standard procedure (212 versus 316 mL, P = 0.001), and the operative times were comparable. The length of the hospital stay was significantly shorter for the minimally invasive technique group (5.9 versus 7.8 days, P < 0.001). The complication rates were 23% and 25% for the standard technique and minimally invasive technique groups, respectively (P = 0.88). Patients undergoing minilaparotomy LDRH had a body mass index (24.0 kg/m(2)) similar to that of the hybrid technique patients (25.8 kg/m(2), P = 0.36), but the graft size was smaller (780 versus 948 mL, P = 0.22). In conclusion, minimally invasive LDRH can be performed without safety being impaired. LDRH with a 10-cm upper midline incision and without laparoscopic assistance may be appropriate for donors with a smaller body mass. Laparoscopic assistance can be added as needed for larger donors. This type of LDRH with a 10-cm incision is innovative and is recommended for experienced centers.

IFN-alpha and bortezomib overcome Bcl-2 and Mcl-1 overexpression in melanoma cells by stimulating the extrinsic pathway of apoptosis.

We hypothesized that IFN-alpha would enhance the apoptotic activity of bortezomib on melanoma cells. Combined treatment with bortezomib and IFN-alpha induced synergistic apoptosis in melanoma and other solid tumor cell lines. Apoptosis was associated with processing of procaspase-3, procaspase-7, procaspase-8, and procaspase-9 and with cleavage of Bid and poly(ADP-ribose) polymerase. Bortezomib plus IFN-alpha was effective at inducing apoptosis in melanoma cells that overexpressed Bcl-2 or Mcl-1, suggesting that this treatment combination can overcome mitochondrial pathways of cell survival and resistance to apoptosis. The proapoptotic effects of this treatment combination were abrogated by a caspase-8 inhibitor, led to increased association of Fas and FADD before the onset of cell death, and were significantly reduced in cells transfected with a dominant-negative FADD construct or small interfering RNA targeting Fas. These data suggest that bortezomib and IFN-alpha act through the extrinsic pathway of apoptosis via FADD-induced caspase-8 activation to initiate cell death. Finally, bortezomib and IFN-alpha displayed statistically significant antitumor activity compared with either agent alone in both the B16 murine model of melanoma and in athymic mice bearing human A375 xenografts. These data support the future clinical development of bortezomib and IFN-alpha for malignant melanoma.

Interferon alpha and CPG oligodeoxynucleotides elicit additive immunostimulatory and antitumor effects.

BACKGROUND: We hypothesized that interferon alpha (IFN-alpha) and unmethylated cytosine-phosphothioate-guanine (CpG)-rich oligodexoynucleotides (CpG ODNs) would elicit potent antitumor activity due to the ability of this treatment combination to activate complimentary signal transduction intermediates. METHODS: Peripheral blood mononuclear cells treated with CpG ODNs, IFN-alpha, or both agents combined were evaluated for cytotoxicity against human melanoma cells, Jak-STAT signal transduction by flow cytometry, and ISG-15 gene expression by real-time polymerase chain reaction. The effects of CpG ODNs and IFN-alpha were evaluated in murine models of melanoma in wild-type, IFN-gamma-deficient, and STAT1-deficient mice. Negative controls in all experiments included treatment with control ODN or phosphate-buffered saline. RESULTS: Treatment of peripheral blood mononuclear cells with a combination of CpG ODNs and IFN-alpha resulted in enhanced cytotoxicity, activation of natural killer cells, IFN-alpha-induced STAT1 phosphorylation, and transcription levels of ISG-15. These immunostimulatory effects of CpG ODNs were associated with increased expression of STAT1 and STAT2 proteins. Administration of CpG ODNs plus IFN-alpha elicited superior antitumor activity in a murine model of B16 melanoma compared with either agent alone. The antitumor properties of CpG ODNs were dependent on STAT1-mediated signal transduction within the host but independent of endogenously produced IFN-gamma. CONCLUSIONS: CpG ODNs represent potent immune stimulants that elicit antitumor effects through STAT1-mediated signal transduction.

Mutations at positions 547-553 of rat glucocorticoid receptors reveal that hsp90 binding requires the presence, but not defined composition, of a seven-amino acid sequence at the amino terminus of the ligand binding domain.

Glucocorticoid receptors (GRs) must heterocomplex with hsp90 to have an open steroid binding cleft that can be accessed by steroid. We reported that a seven-amino acid sequence (547-553 of rat GR) overlapping the amino-terminal end of the ligand binding domain is required for hsp90 binding to GR. We have now conducted saturation mutagenesis of this sequence, which appears to be part of the surface where the ligand binding cleft merges with the surface of the ligand binding domain. No single point mutation causes significant changes in any of a variety of biochemical and biological properties in addition to hsp90 binding. A triple mutation (P548A/T549A/V551A) increases by >100-fold the steroid concentration required for half-maximal induction without affecting the level of maximal induction or coactivator response. Interestingly, this triple mutant displays reduced binding of steroid and hsp90 in whole cells, but it possesses wild type affinity for steroid and normal hsp90 binding capacity under cell-free conditions. This phenotype of a dramatic shift in the dose response for transactivation would be expected from an increase in the rate of disassembly of the triple mutant GR.hsp90 heterocomplex in the cell. Mutation of the entire seven-amino acid region to CAAAAAC maintains the presence of a critical alpha-helical structure and heterocomplex formation with hsp90 but eliminates steroid binding and transcriptional activation, thus disconnecting hsp90 binding from opening of the ligand binding cleft and steroid binding.