The development of an eye movement-based deep learning system for laparoscopic surgical skills assessment.

The development of valid, reliable, and objective methods of skills assessment is central to modern surgical training. Numerous rating scales have been developed and validated for quantifying surgical performance. However, many of these scoring systems are potentially flawed in their design in terms of reliability. Eye-tracking techniques, which provide a more objective investigation of the visual-cognitive aspects of the decision-making process, recently have been utilized in surgery domains for skill assessment and training, and their use has been focused on investigating differences between expert and novice surgeons to understand task performance, identify experienced surgeons, and establish training approaches. Ten graduate students at the National Taiwan University of Science and Technology with no prior laparoscopic surgical skills were recruited to perform the FLS peg transfer task. Then k-means clustering algorithm was used to split 500 trials into three dissimilar clusters, grouped as novice, intermediate, and expert levels, by an objective performance assessment parameter incorporating task duration with error score. Two types of data sets, namely, time series data extracted from coordinates of eye fixation and image data from videos, were used to implement and test our proposed skill level detection system with ensemble learning and a CNN algorithm. Results indicated that ensemble learning and the CNN were able to correctly classify skill levels with accuracies of 76.0% and 81.2%, respectively. Furthermore, the incorporation of coordinates of eye fixation and image data allowed the discrimination of skill levels with a classification accuracy of 82.5%. We examined more levels of training experience and further integrated an eye tracking technique and deep learning algorithms to develop a tool for objective assessment of laparoscopic surgical skill. With a relatively unbalanced sample, our results have demonstrated that the approach combining the features of visual fixation coordinates and images achieved a very promising level of performance for classifying skill levels of trainees.

Foreign objects detection using deep learning techniques for graphic card assembly line.

An assembly is a process in which operators and machines manufacture products from semi-finished components into finished goods. It is important to conduct quality control at the end of the assembly line and ensure that no foreign object is put on the conveyor. This study uses a case of foreign object detection in graphics card assembly line to create models which is capable of detecting and marking foreign objects using convolutional neural network (CNN) models. This study uses Inception Resnet v2 to conduct the foreign object classification and Attention Residual U-net++ for the foreign object segmentation. Both benchmark datasets and case study dataset are employed for model evaluation. The result shows that the proposed models can have more promising result than some existing models.

Recurrent epistaxis: predicting risk of 30-day readmission, derivation and validation of RHINO-ooze score.

BACKGROUND: To derive and validate a predictive scoring tool (RHINO-ooze score) with good sensitivity and specificity in identifying patients with epistaxis at high risk of 30 day readmission and to enable risk stratification for possible definitive intervention. METHODS: Using medical databases, we searched for factors influencing recurrent epistaxis. The information ascertained together with our analysis of retrospective data on patients admitted with epistaxis between October 2013 and September 2014, was used as the derivation cohort to develop the predictive scoring model (RHINO-ooze score). The tool was validated by performing statistical analysis on the validation cohort of patients admitted with epistaxis between October 2014 and October 2015. Multiple linear regressions with backwards elimination was used to derive the predictive model. The area under the curve (AUC), sensitivity and specificity were calculated. RESULTS: 834 admissions were encountered within the study period. Using the derivative cohort (n= 302) the RHINO-ooze score with a maximum score of 8 from five variables (Recent admission, Haemorrhage point unidentified, Increasing age over 70, posterior Nasal packing, Oral anticoagulant) was developed. The RHINO-ooze score had a chi-square value of 99.72 with a significance level of smaller than 0.0001 and hence an overall good model fit. Comparison between the derivative and validation groups revealed similar rates of 30-day readmission between the cohorts. The sensitivity and specificity of predicting 30-day readmission in high risk patients with recurrent epistaxis (RHINO-ooze score equal/larger than 6) was 81% and 84%, respectively. CONCLUSIONS: The RHINO-ooze scoring tool demonstrates good specificity and sensitivity in predicting the risk of 30 day readmission in patients with epistaxis and can be used as an adjunct to clinical decision making with regards to timing of operative intervention in order to reduce readmission rates.

Commercial chicken breeds exhibit highly divergent patterns of linkage disequilibrium.

The analysis of linkage disequilibrium (LD) underpins the development of effective genotyping technologies, trait mapping and understanding of biological mechanisms such as those driving recombination and the impact of selection. We apply the Malécot-Morton model of LD to create additive LD maps that describe the high-resolution LD landscape of commercial chickens. We investigated LD in chickens (Gallus gallus) at the highest resolution to date for broiler, white egg and brown egg layer commercial lines. There is minimal concordance between breeds of fine-scale LD patterns (correlation coefficient <0.21), and even between discrete broiler lines. Regions of LD breakdown, which may align with recombination hot spots, are enriched near CpG islands and transcription start sites (P<2.2 × 10-16), consistent with recent evidence described in finches, but concordance in hot spot locations between commercial breeds is only marginally greater than random. As in other birds, functional elements in the chicken genome are associated with recombination but, unlike evidence from other bird species, the LD landscape is not stable in the populations studied. The development of optimal genotyping panels for genome-led selection programmes will depend on careful analysis of the LD structure of each line of interest. Further study is required to fully elucidate the mechanisms underlying highly divergent LD patterns found in commercial chickens.

The genomic architecture of resistance to Campylobacter jejuni intestinal colonisation in chickens.

BACKGROUND: Campylobacter is the leading cause of foodborne diarrhoeal illness in humans and is mostly acquired from consumption or handling of contaminated poultry meat. In the absence of effective licensed vaccines and inhibitors, selection for chickens with increased resistance to Campylobacter could potentially reduce its subsequent entry into the food chain. Campylobacter intestinal colonisation levels are influenced by the host genetics of the chicken. In the present study, two chicken populations were used to investigate the genetic architecture of avian resistance to colonisation: (i) a back-cross of two White Leghorn derived inbred lines [(61 x N) x N] known to differ in resistance to Campylobacter colonisation and (ii) a 9(th) generation advanced intercross (61 x N) line. RESULTS: The level of colonisation with Campylobacter jejuni following experimental infection was found to be a quantitative trait. A back-cross experiment using 1,243 fully informative single nucleotide polymorphism (SNP) markers revealed quantitative trait loci (QTL) on chromosomes 7, 11 and 14. In the advanced intercross line study, the location of the QTL on chromosome 14 was confirmed and refined and two new QTLs were identified located on chromosomes 4 and 16. Pathway and re-sequencing data analysis of the genes located in the QTL candidate regions identified potential pathways, networks and candidate resistance genes. Finally, gene expression analyses were performed for some of the candidate resistance genes to support the results. CONCLUSION: Campylobacter resistance in chickens is a complex trait, possibly involving the Major Histocompatibility Complex, innate and adaptive immune responses, cadherins and other factors. Two of the QTLs for Campylobacter resistance are co-located with Salmonella resistance loci, indicating that it may be possible to breed simultaneously for enhanced resistance to both zoonoses.

A medical cost estimation with fuzzy neural network of acute hepatitis patients in emergency room.

Taiwan is an area where chronic hepatitis is endemic. Liver cancer is so common that it has been ranked first among cancer mortality rates since the early 1980s in Taiwan. Besides, liver cirrhosis and chronic liver diseases are the sixth or seventh in the causes of death. Therefore, as shown by the active research on hepatitis, it is not only a health threat, but also a huge medical cost for the government. The estimated total number of hepatitis B carriers in the general population aged more than 20 years old is 3,067,307. Thus, a case record review was conducted from all patients with diagnosis of acute hepatitis admitted to the Emergency Department (ED) of a well-known teaching-oriented hospital in Taipei. The cost of medical resource utilization is defined as the total medical fee. In this study, a fuzzy neural network is employed to develop the cost forecasting model. A total of 110 patients met the inclusion criteria. The computational results indicate that the FNN model can provide more accurate forecasts than the support vector regression (SVR) or artificial neural network (ANN). In addition, unlike SVR and ANN, FNN can also provide fuzzy IF-THEN rules for interpretation.

Lentivirus delivery of IL-10 to promote and sustain macrophage polarization towards an anti-inflammatory phenotype.

Gene delivery from biomaterials can create an environment that promotes and guides tissue formation. However, the immune response induced upon biomaterial implantation can be detrimental to tissue regeneration. Macrophages play a central role in mediating early phases of this response, and functional "polarization" of macrophages towards M1 (inflammatory) or M2 (anti-inflammatory) phenotypes may bias the local immune state at the implant site. Since gene delivery from biomaterial scaffolds can confer transgene expression in macrophages in vivo, we investigated whether transduction of macrophages with an IL-10 encoding lentivirus can (1) induce macrophage polarization toward an M2 phenotype even in an pro-inflammatory environment, and (2) prevent a shift in polarization from M2 to M1 following exposure to pro-inflammatory stimuli. IL-10 lentivirus delivery to pre-polarized M1 macrophages reduced TNF-α production 1.5-fold when compared to cells treated with either a control virus or a bolus delivery of recombinant IL-10 protein. IL-10 lentivirus delivery to naïve macrophages reduced the amount of TNF-α produced following an inflammatory challenge by 2.5-fold compared to cells treated with both the control virus and recombinant IL-10. At a mechanistic level, IL-10 lentivirus delivery mediated sustained reduction in NF-κB activation and, accordingly, reduced transcription of TNF-α. In sum, lentiviral delivery of IL-10 to macrophages represents a promising strategy for directing and sustaining macrophage polarization towards an M2 phenotype in order to promote local immune responses that facilitate tissue engineering.

Fluke (Spirorchiidae) infections in sea turtles stranded on Taiwan: prevalence and pathology.

The prevalence of spirorchiid fluke infections of marine turtles is high and may cause the death of the hosts throughout their ranges. Virtually nothing has been reported regarding the infective status of sea turtles stranded on Taiwan. Between 2007 and 2010, 30 green turtles (Chelonia mydas) and 2 loggerhead turtles ( Caretta caretta ), stranded and dead, were examined for spirorchiid flukes and their eggs. Twenty-four of the green turtles were juveniles, and the stranded loggerhead turtles were subadults. Adult spirorchiid flukes were found in 13 green turtles but not in the loggerheads. Four species of flukes were identified, namely, Leardius learedi , Hapalotrema postorchis , H. mehrai , and Carettacola hawaiiensis . The main infection sites were the major arteries and heart. Seventy percent of the green turtles harbored spirorchiid eggs, but no eggs were found in loggerheads. The largest eggs with bipolar spines, type I eggs, were found in every case. Although more than half of the stranded turtles were infected, parasite infections were not the main cause of death in the green turtles. Fishery by-catch is probably responsible for the mortality of these stranded turtles.

Leiomyosarcoma in the nasopharynx: MR imaging findings.

The MR imaging findings of a leiomyosarcoma arising from the nasopharynx are presented. To our knowledge, this is the first MR imaging description of this entity.

Non-calcifying variant of calcifying epithelial odontogenic tumor with Langerhans cells.

Calcifying epithelial odontogenic tumor (CEOT) is a rare type of odontogenic tumor. The most characteristic feature of the classical CEOT is the presence of amyloid globules and Liesegang ring calcification in the tumor tissue. Here, we present a non-calcifying variant of intraosseous CEOT with the presence of Langerhans cells within tumor epithelial nests in a 52-year-old Taiwanese woman. The patient was referred from a local dentist to our hospital for treatment of a unilocular radiolucent lesion at the right anterior region of the maxilla. The lesion was excised. Microscopically, the tumor was composed of small nests or strands of odontogenic epithelial cells and amorphous eosinophilic globules of amyloid-like materials in a loose fibrous connective tissue stroma. The tumor epithelial cells were positive for pan-cytokeratins (AE1 and AE3). Langerhans cells demonstrated by anti-CD1a staining were found in nests or strands of tumor epithelial cells. The eosinophilic globules were positive for Congo red and showed green birefringence when subjected to polarized light. Review of the English literature revealed two cases of non-calcifying variant of intraosseous CEOT with Langerhans cells in the anterior and premolar regions of the maxilla. Taken together, we suggest that the non-calcifying, Langerhan cell-rich variant of CEOT may have a distinct predilection for occurrence in the anterior and premolar region of the maxilla in contrast to the classical CEOTs that usually occur in the molar and ascending ramus area of the mandible.

Cure of orthopaedic infection with Scedosporium prolificans, using voriconazole plus terbinafine, without the need for radical surgery.

Scedosporium prolificans infections of normal hosts usually require extensive debridement and sometimes amputation to effect cure, due to the intrinsic resistance of this species to available antifungal agents. Newer agents have not tested favourably. Variable results are obtained with voriconazole, and 100% resistance is described with echinocandins. Itraconazole and terbinafine has offered synergy against various moulds including S. prolificans. In vivo success is reported with the azole/terbinafine combination in S. apiospermum pulmonary infection and Pythium insidiosum periorbital cellulitis. We report a case of orthopaedic infection in a non-immunocompromised host with S. prolificans, in which the combinations of itraconazole/terbinafine and voriconazole/terbinafine showed synergy in vitro, and success was achieved without radical surgery, using voriconazole and terbinafine.

An intelligent sales forecasting system through integration of artificial neural networks and fuzzy neural networks with fuzzy weight elimination.

Sales forecasting plays a very prominent role in business strategy. Numerous investigations addressing this problem have generally employed statistical methods, such as regression or autoregressive and moving average (ARMA). However, sales forecasting is very complicated owing to influence by internal and external environments. Recently, artificial neural networks (ANNs) have also been applied in sales forecasting since their promising performances in the areas of control and pattern recognition. However, further improvement is still necessary since unique circumstances, e.g. promotion, cause a sudden change in the sales pattern. Thus, this study utilizes a proposed fuzzy neural network (FNN), which is able to eliminate the unimportant weights, for the sake of learning fuzzy IF-THEN rules obtained from the marketing experts with respect to promotion. The result from FNN is further integrated with the time series data through an ANN. Both the simulated and real-world problem results show that FNN with weight elimination can have lower training error compared with the regular FNN. Besides, real-world problem results also indicate that the proposed estimation system outperforms the conventional statistical method and single ANN in accuracy.

An indicator cell assay for detection of human cytomegalovirus based on enhanced green fluorescent protein.

An indicator cell line (ML-UL54-EGFP) for the detection of human cytomegalovirus (HCMV) by a simple and direct method was developed. The stable line was constructed by introducing into mink lung cells an expression cassette that contains the enhanced green fluorescent protein (EGFP) reporter gene under the control of an HCMV-inducible promoter. The promoter was from the upstream region of the HCMV UL54 (pol) gene, an early gene promoter that is activated in the early phase of HCMV infection. Following infection with HCMV for 48 h, the stable line expressed well detectable level of the EGFP as observed under a fluorescence microscope. The sensitivity of the indicator cell assay is at least comparable with that of a plaque assay as assessed with a panel of HCMV strains. There were no detectable fluorescent cells after inoculations with several viruses other than HCMV, indicating high specificity. Analysis with flow cytometry revealed that the induced fluorescence from the infected cells was proportional to the titer of HCMV inoculated, making it possible to quantify HCMV infectious particles. In summary, the EGFP-based indicator cell line is of potential use for rapid detection and quantification of HCMV in clinical specimens.

Hydrachine A, a novel alkaloid from the roots of Hydrangea chinensis.

A novel alkaloid, hydrachine A (3), has been isolated, along with 15 known compounds, from the roots of Hydrangea chinensis. The structure and stereochemistry of the new alkaloid 3 was determined using extensive 2D NMR data.

Antiplatelet activity of N-methoxycarbonyl aporphines from Rollinia mucosa.

Bioassay-directed fractionation of the stems of Rollinia mucosa led to the isolation of new N-methoxycarbonyl aporphine alkaloids, romucosine A (1), romucosine B (2), romucosine C (3), and romucosine D (4), along with the known alkaloid, N-methoxylcarbonyl-nornuciferine (5). Alkaloids 1 and 4 exhibited significant inhibition of collagen, arachidonic acid, and platelet activating factor-induced platelet aggregation, and alkaloid 3 also showed an inhibitory effect on arachidonic acid induced platelet aggregation.

Virtual reality: current urologic applications and future developments.

The concept of virtual reality (VR) involves the computer generation of environments with which a user can interact directly. Virtual reality is now being used for medical applications, especially in the area of surgical simulation. As technology advances, VR simulation will play an important training role for both residents and urologists already in practice. This paper examines the history of VR, current developments, and its future implications for the field of urology.

Current urologic applications of digital imaging.

One of the most significant developments in imaging technology has been the process of digitalization. By incorporating currently available digital imaging equipment into surgical practice, urologists can be assured of obtaining real-time video images with optimal clarity and detail. In addition, one can efficiently capture and store still images that are crisper and sharper than their analog counterparts. These factors greatly improve the diagnostic capabilities and organization of today's endourologist.

Telemedicine: recent developments and future applications.

Telemedicine, or the delivery of health care from a distance, is an exciting field that has undergone rapid advancements over the past three decades. Previous studies have demonstrated that telemedicine can effectively assist in patient care. However, cost issues and the lack of resources to sustain telemedicine systems have previously limited the use of this technology. Herein, we explore the development, current applications, and future of telemedicine.