Improving image quality of sparse-view lung tumor CT images with U-Net.

BACKGROUND: We aimed to improve the image quality (IQ) of sparse-view computed tomography (CT) images using a U-Net for lung metastasis detection and determine the best tradeoff between number of views, IQ, and diagnostic confidence. METHODS: CT images from 41 subjects aged 62.8 ± 10.6 years (mean ± standard deviation, 23 men), 34 with lung metastasis, 7 healthy, were retrospectively selected (2016-2018) and forward projected onto 2,048-view sinograms. Six corresponding sparse-view CT data subsets at varying levels of undersampling were reconstructed from sinograms using filtered backprojection with 16, 32, 64, 128, 256, and 512 views. A dual-frame U-Net was trained and evaluated for each subsampling level on 8,658 images from 22 diseased subjects. A representative image per scan was selected from 19 subjects (12 diseased, 7 healthy) for a single-blinded multireader study. These slices, for all levels of subsampling, with and without U-Net postprocessing, were presented to three readers. IQ and diagnostic confidence were ranked using predefined scales. Subjective nodule segmentation was evaluated using sensitivity and Dice similarity coefficient (DSC); clustered Wilcoxon signed-rank test was used. RESULTS: The 64-projection sparse-view images resulted in 0.89 sensitivity and 0.81 DSC, while their counterparts, postprocessed with the U-Net, had improved metrics (0.94 sensitivity and 0.85 DSC) (p = 0.400). Fewer views led to insufficient IQ for diagnosis. For increased views, no substantial discrepancies were noted between sparse-view and postprocessed images. CONCLUSIONS: Projection views can be reduced from 2,048 to 64 while maintaining IQ and the confidence of the radiologists on a satisfactory level. RELEVANCE STATEMENT: Our reader study demonstrates the benefit of U-Net postprocessing for regular CT screenings of patients with lung metastasis to increase the IQ and diagnostic confidence while reducing the dose. KEY POINTS: • Sparse-projection-view streak artifacts reduce the quality and usability of sparse-view CT images. • U-Net-based postprocessing removes sparse-view artifacts while maintaining diagnostically accurate IQ. • Postprocessed sparse-view CTs drastically increase radiologists' confidence in diagnosing lung metastasis.

Physical strain while wearing personal radiation protection systems in interventional radiology.

OBJECTIVES: Multiple studies show orthopedic health problems for medical staff due to wearing radiation protection aprons. The aim of this study was to evaluate the weight pressure on the shoulder as a marker of physical strain caused by different radiation-protection devices. METHODS: For the weight pressure measurement, a pressure sensor (OMD-30-SE-100N, OptoForce, Budapest, Hungary) placed on the left and right shoulder was used. Wearing different radiation protection systems the force measurement system was used to quantify the weight pressure. Measurements were acquired in still standing position and during various movements. RESULTS: A mean significant decreasing weight pressure on the shoulder between 74% and 84% (p<0.001) was measured, when the free-hanging radiation protection system was used in comparison to one-piece and two-piece radiation protection aprons and coats. Using two-piece radiation protection aprons, the weight pressure was significantly lower than that of one-piece radiation protection coats. If a belt was used for the one-piece radiation protection coat, the weight pressure on the shoulder was reduced by 32.5% (p = 0.003). For a two-piece radiation protection apron and a one-piece radiation protection coat (with and without belt) a significant different weight pressure distribution between the right and left shoulder could be measured. CONCLUSIONS: The free-hanging radiation protection system showed a significant lower weight pressure in comparison to the other radiation protection devices. Apart from this, use of a two-piece radiation protection apron or addition of a belt to a radiation protection coat proved to be further effective options to reduce weight pressure.

Lung nodule detection in chest X-rays using synthetic ground-truth data comparing CNN-based diagnosis to human performance.

We present a method to generate synthetic thorax radiographs with realistic nodules from CT scans, and a perfect ground truth knowledge. We evaluated the detection performance of nine radiologists and two convolutional neural networks in a reader study. Nodules were artificially inserted into the lung of a CT volume and synthetic radiographs were obtained by forward-projecting the volume. Hence, our framework allowed for a detailed evaluation of CAD systems' and radiologists' performance due to the availability of accurate ground-truth labels for nodules from synthetic data. Radiographs for network training (U-Net and RetinaNet) were generated from 855 CT scans of a public dataset. For the reader study, 201 radiographs were generated from 21 nodule-free CT scans with altering nodule positions, sizes and nodule counts of inserted nodules. Average true positive detections by nine radiologists were 248.8 nodules, 51.7 false positive predicted nodules and 121.2 false negative predicted nodules. The best performing CAD system achieved 268 true positives, 66 false positives and 102 false negatives. Corresponding weighted alternative free response operating characteristic figure-of-merits (wAFROC FOM) for the radiologists range from 0.54 to 0.87 compared to a value of 0.81 (CI 0.75-0.87) for the best performing CNN. The CNN did not perform significantly better against the combined average of the 9 readers (p = 0.49). Paramediastinal nodules accounted for most false positive and false negative detections by readers, which can be explained by the presence of more tissue in this area.

Efficient, high-performance semantic segmentation using multi-scale feature extraction.

The success of deep learning in recent years has arguably been driven by the availability of large datasets for training powerful predictive algorithms. In medical applications however, the sensitive nature of the data limits the collection and exchange of large-scale datasets. Privacy-preserving and collaborative learning systems can enable the successful application of machine learning in medicine. However, collaborative protocols such as federated learning require the frequent transfer of parameter updates over a network. To enable the deployment of such protocols to a wide range of systems with varying computational performance, efficient deep learning architectures for resource-constrained environments are required. Here we present MoNet, a small, highly optimized neural-network-based segmentation algorithm leveraging efficient multi-scale image features. MoNet is a shallow, U-Net-like architecture based on repeated, dilated convolutions with decreasing dilation rates. We apply and test our architecture on the challenging clinical tasks of pancreatic segmentation in computed tomography (CT) images as well as brain tumor segmentation in magnetic resonance imaging (MRI) data. We assess our model's segmentation performance and demonstrate that it provides performance on par with compared architectures while providing superior out-of-sample generalization performance, outperforming larger architectures on an independent validation set, while utilizing significantly fewer parameters. We furthermore confirm the suitability of our architecture for federated learning applications by demonstrating a substantial reduction in serialized model storage requirement as a surrogate for network data transfer. Finally, we evaluate MoNet's inference latency on the central processing unit (CPU) to determine its utility in environments without access to graphics processing units. Our implementation is publicly available as free and open-source software.

Treatment of intestinal tuberculosis with small bowel perforation: a case report.

BACKGROUND: Diagnosis of intestinal tuberculosis poses a dilemma to physicians due to nonspecific symptoms like abdominal pain, fever, nausea, and a change in bowel habit. In particular, the distinction between inflammatory bowel disease and intestinal tuberculosis remains challenging. CASE PRESENTATION: A 27-year-old man from Colombia presented with fever, night sweats, and progressive lower abdominal pain. Computed tomography revealed a thickening of the bowel wall with a mesenterial lymphadenopathy, ascites ,and a pleural tumor mass. Histology of intestinal and pleural biopsy specimens showed a granulomatous inflammation. Although microscopy and polymerase chain reaction (PCR) for Mycobacterium tuberculosis (MTB) were negative, empirical MTB treatment was initiated on suspicion. Due to a massive post-stenotic atrophied intestinal bowel, MTB medications were administered parenterally in the initial phase of treatment to guarantee adequate systemic resorption. The complicated and critical further course included an intra-abdominal abscess and bowel perforation requiring a split stoma, before the patient could be discharged in good condition after 3 months of in-hospital care. CONCLUSIONS: This case highlights the clinical complexity and diagnostic challenges of intestinal MTB infection. A multidisciplinary team of physicians should be sensitized to a timely diagnosis of this disease, which often mimics inflammation similar to inflammatory bowel disease, other infections, or malignancies. In our case, radiological findings, histological results, and migratory background underpinned the suspected diagnosis and allowed early initiation of tuberculostatic treatment.

Image-Based Molecular Phenotyping of Pancreatic Ductal Adenocarcinoma.

To bridge the translational gap between recent discoveries of distinct molecular phenotypes of pancreatic cancer and tangible improvements in patient outcome, there is an urgent need to develop strategies and tools informing and improving the clinical decision process. Radiomics and machine learning approaches can offer non-invasive whole tumor analytics for clinical imaging data-based classification. The retrospective study assessed baseline computed tomography (CT) from 207 patients with proven pancreatic ductal adenocarcinoma (PDAC). Following expert level manual annotation, Pyradiomics was used for the extraction of 1474 radiomic features. The molecular tumor subtype was defined by immunohistochemical staining for KRT81 and HNF1a as quasi-mesenchymal (QM) vs. non-quasi-mesenchymal (non-QM). A Random Forest machine learning algorithm was developed to predict the molecular subtype from the radiomic features. The algorithm was then applied to an independent cohort of histopathologically unclassifiable tumors with distinct clinical outcomes. The classification algorithm achieved a sensitivity, specificity and ROC-AUC (area under the receiver operating characteristic curve) of 0.84 ± 0.05, 0.92 ± 0.01 and 0.93 ± 0.01, respectively. The median overall survival for predicted QM and non-QM tumors was 16.1 and 20.9 months, respectively, log-rank-test p = 0.02, harzard ratio (HR) 1.59. The application of the algorithm to histopathologically unclassifiable tumors revealed two groups with significantly different survival (8.9 and 39.8 months, log-rank-test p < 0.001, HR 4.33). The machine learning-based analysis of preoperative (CT) imaging allows the prediction of molecular PDAC subtypes highly relevant for patient survival, allowing advanced pre-operative patient stratification for precision medicine applications.

Multiple internalization pathways of polyelectrolyte multilayer capsules into mammalian cells.

Polyelectrolyte multilayer (PEM) capsules are carrier vehicles with great potential for biomedical applications. With the future aim of designing biocompatible, effective therapeutic delivery systems (e.g., for cancer), the pathway of internalization (uptake and fate) of PEM capsules was investigated. In particular the following experiments were performed: (i) the study of capsule co-localization with established endocytic markers, (ii) switching-off endocytotic pathways with pharmaceutical/chemical inhibitors, and (iii) characterization and quantification of capsule uptake with confocal and electron microscopy. As result, capsules co-localized with lipid rafts and with phagolysosomes, but not with other endocytic vesicles. Chemical interference of endocytosis with chemical blockers indicated that PEM capsules enter the investigated cell lines through a mechanism slightly sensitive to electrostatic interactions, independent of clathrin and caveolae, and strongly dependent on cholesterol-rich domains and organelle acidification. Microscopic characterization of cells during capsule uptake showed the formation of phagocytic cups (vesicles) to engulf the capsules, an increased number of mitochondria, and a final localization in the perinuclear cytoplasma. Combining all these indicators we conclude that PEM capsule internalization in general occurs as a combination of different sequential mechanisms. Initially, an adsorptive mechanism due to strong electrostatic interactions governs the stabilization of the capsules at the cell surface. Membrane ruffling and filopodia extensions are responsible for capsule engulfing through the formation of a phagocytic cup. Co-localization with lipid raft domains activates the cell to initiate a lipid-raft-mediated macropinocytosis. Internalization vesicles are very acidic and co-localize only with phagolysosome markers, excluding caveolin-mediated pathways and indicating that upon phagocytosis the capsules are sorted to heterophagolysosomes.

Nanopharmacy: Inorganic nanoscale devices as vectors and active compounds.

In this review we would like to aim at pharmaceuticals engineered on the nanoscale, i.e. pharmaceuticals where the nanomaterial plays the pivotal therapeutic role or adds additional functionality to the previous compound. Those cases would be considered as nanopharmaceuticals. The development of inorganic systems is opening the pharmaceutical nanotechnology novel horizons for diagnosis, imaging and therapy mainly because of their nanometer-size and their high surface area to volume ratios which allow for specific functions that are not possible in the micrometer-size particles. This review will focus on pharmaceutical forms that are based on inorganic nanoparticles where the nanosize of the inorganic component provides unique characteristics to the pharmaceutical form. Several examples of these systems that are either in pre-clinical investigation and under examination by the Food and Drug Administration (FDA) or that have been already approved by the FDA and are in clinical practice today like Gastromark, NanoTherm, Colloidal Gold for Lateral Flow tests, HfO-NPs, BioVant will be described and reviewed.