Multiphase convolutional dense network for the classification of focal liver lesions on dynamic contrast-enhanced computed tomography

BACKGROUND The accurate classification of focal liver lesions(FLLs)is essential to properly guide treatment options and predict prognosis.Dynamic contrast-enhanced computed tomography(DCE-CT)is still the cornerstone in the exact classification of FLLs due to its noninvasive nature,high scanning speed,and high-density resolution.Since their recent development,convolutional neural network-based deep learning techniques has been recognized to have high potential for image recognition tasks.AIM To develop and evaluate an automated multiphase convolutional dense network(MP-CDN)to classify FLLs on multiphase CT.METHODS A total of 517 FLLs scanned on a 320-detector CT scanner using a four-phase DCECT imaging protocol(including precontrast phase,arterial phase,portal venous phase,and delayed phase)from 2012 to 2017 were retrospectively enrolled.FLLs were classified into four categories:Category A,hepatocellular carcinoma(HCC);category B,liver metastases;category C,benign non-inflammatory FLLs including hemangiomas,focal nodular hyperplasias and adenomas;and category D,hepatic abscesses.Each category was split into a training set and test set in an approximate 8:2 ratio.An MP-CDN classifier with a sequential input of the fourphase CT images was developed to automatically classify FLLs.The classification performance of the model was evaluated on the test set;the accuracy and specificity were calculated from the confusion matrix,and the area under the receiver operating characteristic curve(AUC)was calculated from the SoftMax probability outputted from the last layer of the MP-CDN.RESULTS A total of 410 FLLs were used for training and 107 FLLs were used for testing.The mean classification accuracy of the test set was 81.3%(87/107).The accuracy/specificity of distinguishing each category from the others were 0.916/0.964,0.925/0.905,0.860/0.918,and 0.925/0.963 for HCC,metastases,benign non-inflammatory FLLs,and abscesses on the test set,respectively.The AUC(95%confidence interval)for differentiating each category from the others was 0.92(0.837-0.992),0.99(0.967-1.00),0.88(0.795-0.955)and 0.96(0.914-0.996)for HCC,metastases,benign non-inflammatory FLLs,and abscesses on the test set,respectively.CONCLUSION MP-CDN accurately classified FLLs detected on four-phase CT as HCC,metastases,benign non-inflammatory FLLs and hepatic abscesses and may assist radiologists in identifying the different types of FLLs.

Findings on intraprocedural non-contrast computed tomographic imaging following hepatic artery embolization are associated with development of contrast-induced nephropathy

BACKGROUND Contrast-induced nephropathy(CIN)is a reversible form of acute kidney injury that occurs within 48-72 h of exposure to intravascular contrast material.CIN is the third leading cause of hospital-acquired acute kidney injury and accounts for 12%of such cases.Risk factors for CIN development can be divided into patientand procedure-related.The former includes pre-existing chronic renal insufficiency and diabetes mellitus.The latter includes high contrast volume and repeated exposure over 72 h.The incidence of CIN is relatively low(up to 5%)in patients with intact renal function.However,in patients with known chronic renal insufficiency,the incidence can reach up to 27%.AIM To examine the association between renal enhancement pattern on non-contrast enhanced computed tomographic(CT)images obtained immediately following hepatic artery embolization with development of CIN.METHODS Retrospective review of all patients who underwent hepatic artery embolization between 01/2010 and 01/2011(n=162)was performed.Patients without intraprocedural CT imaging(n=51),combined embolization/ablation(n=6)and those with chronic kidney disease(n=21)were excluded.The study group comprised of 84 patients with 106 procedures.CIN was defined as 25%increase above baseline serum creatinine or absolute increase≥0.5 mg/dL within 72 h post-embolization.Post-embolization CT was reviewed for renal enhancement patterns and presence of renal artery calcifications.The association between noncontrast CT findings and CIN development was examined by Fisher’s Exact Test.RESULTS CIN occurred in 11/106(10.3%)procedures(Group A,n=10).The renal enhancement pattern in patients who did not experience CIN(Group B,n=74 with 95/106 procedures)was late excretory in 93/95(98%)and early excretory(EE)in 2/95(2%).However,in Group A,there was a significantly higher rate of EE pattern(6/11,55%)compared to late excretory pattern(5/11)(P<0.001).A significantly higher percentage of patients that developed CIN had renal artery calcifications(6/11 vs 20/95,55%vs 21%,P=0.02).CONCLUSION A hyperdense renal parenchyma relative to surrounding skeletal muscle(EE pattern)and presence of renal artery calcifications on immediate post-HAE noncontrast CT images in patients with low risk for CIN are independently associated with CIN development.

Enhanced Feature Fusion Segmentation for Tumor Detection Using Intelligent Techniques

In thefield of diagnosis of medical images the challenge lies in tracking and identifying the defective cells and the extent of the defective region within the complex structure of a brain cavity.Locating the defective cells precisely during the diagnosis phase helps tofight the greatest exterminator of mankind.Early detec-tion of these defective cells requires an accurate computer-aided diagnostic system(CAD)that supports early treatment and promotes survival rates of patients.An ear-lier version of CAD systems relies greatly on the expertise of radiologist and it con-sumed more time to identify the defective region.The manuscript takes the efficacy of coalescing features like intensity,shape,and texture of the magnetic resonance image(MRI).In the Enhanced Feature Fusion Segmentation based classification method(EEFS)the image is enhanced and segmented to extract the prominent fea-tures.To bring out the desired effect the EEFS method uses Enhanced Local Binary Pattern(EnLBP),Partisan Gray Level Co-occurrence Matrix Histogram of Oriented Gradients(PGLCMHOG),and iGrab cut method to segment image.These prominent features along with deep features are coalesced to provide a single-dimensional fea-ture vector that is effectively used for prediction.The coalesced vector is used with the existing classifiers to compare the results of these classifiers with that of the gen-erated vector.The generated vector provides promising results with commendably less computatio nal time for pre-processing and classification of MR medical images.

Facile and robust strategy to antireflective photo-curing coating through self-wrinkling

We reported a facile and bio-inspired strategy for obtaining antireflective （AR） coating through polymerization-induced self-wrinkling. Upon irradiation of light, the complex wrinkle micro-patterns with different morphologies were generated spontaneously on the surface of coating during photo-cross- linking, which enables the photo-curing coating can decrease reflection. The resulting photo-curing coating exhibits a high transmittance over 90% and low reflection below 5% ～ 8%, with an efficiency anti- reflection of 4% ～ 7%; compared to the flat blank coating. The successful application of these AR coatings with wrinkles pattern to encapsulate the thin film solar cells results in appreciable photovoltaic performance improvement of more than 4% ～ 8%, which benefits from the decrease of the light reflection and increase of optical paths in the photoactive layer by the introduction of wrinkling pattern. Furthermore, the efficiency improvements of the solar cells are more obvious, with a remarkable increase of 8.5%, at oblique light incident angle than that with vertical light incident angle

Aerodynamic Comparison between Increasing Cascade Pitch and Turning Angle in the Highly-Loaded Design

Theoretical and numerical study was carried out based on a linear turbine cascade(the Basic cascade)to compare the influences of the increased cascade pitch and turning angle in this paper.On one hand,the two highly-loaded designs both reduced the stability of flow field through enhancing adverse pressure gradient and span-wise pressure gradient of the fluid near suction surface.Therefore,the two highly-loaded designs would both result in thicker boundary layer and stronger secondary flow,so the secondary loss would be increased and more difficult to suppress in the highly-loaded cascades.On the other hand,the two highly-loaded designs showed different influences on the pitch-wise migration of the fluid near the endwall(cross flow)because of the different load enhancing mechanisms.In other words,the increased cascade pitch(TCx highly-loaded design)would delay the pitch-wise migration of the horseshoe vortex because of the increased channel width,while the increased turning angle(Turn highly-loaded design)would do the opposite because of the increased pitch-wise pressure gradient.As a result,the enhancement of the interaction between the fluid near the suction surface and the cross flow would be much stronger in the Turn highly-loaded design than the TCx highly-loaded design,and the span-wise developing tendencies of vortexes and fluid near the suction surface would show much stronger enhancing tendency in the former than the latter.