Lung Pattern Classification for Interstitial Lung Diseases Using a Deep Convolutional Neural Network

Automated tissue characterization is one of the most crucial components of a computer aided diagnosis (CAD) system for interstitial lung diseases (ILDs). Although much research has been conducted in this field, the problem remains challenging. Deep learning techniques have recently achieved impressive results in a variety of computer vision problems, raising expectations that they might be applied in other domains, such as medical image analysis. In this paper, we propose and evaluate a convolutional neural network (CNN), designed for the classification of ILD patterns. The proposed network consists of 5 convolutional layers with 2×2 kernels and LeakyReLU activations, followed by average pooling with size equal to the size of the final feature maps and three dense layers. The last dense layer has 7 outputs, equivalent to the classes considered: healthy, ground glass opacity (GGO), micronodules, consolidation, reticulation, honeycombing and a combination of GGO/reticulation. To train and evaluate the CNN, we used a dataset of 14696 image patches, derived by 120 CT scans from different scanners and hospitals. To the best of our knowledge, this is the first deep CNN designed for the specific problem. A comparative analysis proved the effectiveness of the proposed CNN against previous methods in a challenging dataset. The classification performance (~85.5%) demonstrated the potential of CNNs in analyzing lung patterns. Future work includes, extending the CNN to three-dimensional data provided by CT volume scans and integrating the proposed method into a CAD system that aims to provide differential diagnosis for ILDs as a supportive tool for radiologists.

Hierarchical Markov Random Fields Applied to Model Soft Tissue Deformations on Graphics Hardware

Many methodologies dealing with prediction or simulation of soft tissue deformations on medical image data require preprocessing of the data in order to produce a different shape representation that complies with standard methodologies, such as mass–spring networks, finite element method s (FEM). On the other hand, methodologies working directly on the image space normally do not take into account mechanical behavior of tissues and tend to lack physics foundations driving soft tissue deformations. This chapter presents a method to simulate soft tissue deformations based on coupled concepts from image analysis and mechanics theory. The proposed methodology is based on a robust stochastic approach that takes into account material properties retrieved directly from the image, concepts from continuum mechanics and FEM. The optimization framework is solved within a hierarchical Markov random field (HMRF) which is implemented on the graphics processor unit (GPU See Graphics processing unit ).

Minimally invasive surgery when treating endometriosis has a positive effect on health and on quality of work life of affected women

STUDY QUESTION: What is the effect of the minimally invasive surgical treatment of endometriosis on health and on quality of work life (e.g. working performance) of affected women? SUMMARY ANSWER: Absence from work, performance loss and the general negative impact of endometriosis on the job are reduced significantly by the laparoscopic surgery. WHAT IS KNOWN ALREADY: The benefits of surgery overall and of the laparoscopic method in particular for treating endometriosis have been described before. However, previous studies focus on medical benchmarks without including the patient's perspective in a quantitative manner. STUDY DESIGN, SIZE, DURATION: A retrospective questionnaire-based survey covering 211 women with endometriosis and a history of specific laparoscopic surgery in a Swiss university hospital, tertiary care center. Data were returned anonymously and were collected from the beginning of 2012 until March 2013. PARTICIPANTS/MATERIALS, SETTING, METHODS: Women diagnosed with endometriosis and with at least one specific laparoscopic surgery in the past were enrolled in the study. The study investigated the effect of the minimally invasive surgery on health and on quality of work life of affected women. Questions used were obtained from the World Endometriosis Research Foundation (WERF) Global Study on Women's Health (GSWH) instrument. The questionnaire was shortened and adapted for the purpose of the present study. MAIN RESULTS AND THE ROLE OF CHANCE: Of the 587 women invited to participate in the study, 232 (232/587 = 40%) returned the questionnaires. Twenty-one questionnaires were excluded due to incomplete data and 211 sets (211/587 = 36%) were included in the study. Our data show that 62% (n = 130) of the study population declared endometriosis as influencing the job during the period prior to surgery, compared with 28% after surgery (P < 0.001). The mean (maximal) absence from work due to endometriosis was reduced from 2.0 (4.9) to 0.5 (1.4) hours per week (P < 0.001). The mean (maximal) loss in working performance after the surgery averaged out at 5.7% (12.6%) compared with 17.5% (30.5%) before this treatment (P < 0.001). LIMITATIONS, REASONS FOR CAUTION: The mediocre response rate of the study weakens the representativeness of the investigated population. Considering the anonymous setting a non-responder investigation was not performed. A bias due to selection, information and negativity effects within a retrospective survey cannot be excluded, although study-sensitive questions were provided in multiple ways. The absence of a control group (sham group; e.g. patients undergoing specific diagnostic laparoscopy without treatment) is a further limitation of the study. WIDER IMPLICATIONS OF THE FINDINGS: Our study shows that indicated minimally invasive surgery has a clear positive effect on the wellbeing and working performance of women suffering from moderate to severe endometriosis. Furthermore, national net savings in indirect costs with the present number of surgeries is estimated to be €10.7 million per year. In an idealized setting (i.e. without any diagnosis delay) this figure could be more than doubled. STUDY FUNDING/COMPETING INTERESTS: The study was performed on behalf of the University Hospital of Bern (Inselspital) as one of the leading Swiss tertiary care centers. The authors do not declare any competing interests.

Streak detection algorithm for space debris detection on optical images.

Any image processing object detection algorithm somehow tries to integrate the object light (Recognition Step) and applies statistical criteria to distinguish objects of interest from other objects or from pure background (Decision Step). There are various possibilities how these two basic steps can be realized, as can be seen in the different proposed detection methods in the literature. An ideal detection algorithm should provide high recognition sensitiv ity with high decision accuracy and require a reasonable computation effort . In reality, a gain in sensitivity is usually only possible with a loss in decision accuracy and with a higher computational effort. So, automatic detection of faint streaks is still a challenge. This paper presents a detection algorithm using spatial filters simulating the geometrical form of possible streaks on a CCD image. This is realized by image convolution. The goal of this method is to generate a more or less perfect match between a streak and a filter by varying the length and orientation of the filters. The convolution answers are accepted or rejected according to an overall threshold given by the ackground statistics. This approach yields as a first result a huge amount of accepted answers due to filters partially covering streaks or remaining stars. To avoid this, a set of additional acceptance criteria has been included in the detection method. All criteria parameters are justified by background and streak statistics and they affect the detection sensitivity only marginally. Tests on images containing simulated streaks and on real images containing satellite streaks show a very promising sensitivity, reliability and running speed for this detection method. Since all method parameters are based on statistics, the true alarm, as well as the false alarm probability, are well controllable. Moreover, the proposed method does not pose any extraordinary demands on the computer hardware and on the image acquisition process.

Medical, psychological and socio-economic implications of chronic hand eczema: a cross-sectional study.

BACKGROUND Hand eczema (HE) is a common skin disease with major medical psychological and socio-economic implications. Onset and prognosis of HE are determined by individual as well as environmental factors. So far, most epidemiological data on HE have been reported from Scandinavian and recently German studies. OBJECTIVE To investigate the characteristics and medical care of patients with chronic HE (CHE) in Switzerland, and identify risk factors. METHODS In this cross-sectional study, data from patients with chronic HE were obtained by means of medical history, dermatological examination and patient questionnaires. Multiple logistic regression analysis was applied to identify risk factors for high severity and dermatology life quality index (DLQI). RESULTS In seven dermatology departments, 199 patients (mean age 40.4 years, 50.8% female) with CHE (mean duration 6.6 years) were enrolled. Moderate to severe HE was reported by 70.9% of patients, and was associated with age <30 or >50 years, localization of lesions and pruritus. Because of the CHE, 37.3% of patients were on sick leave over the past 12 months, 14.8% had changed or lost their job. Practically all patients applied topical therapy, 21% were treated with alitretinoin, and 21% with psoralen plus UVA light (PUVA). The effects on the health-related quality of life was moderate to large in 33.7% and 39.4% of CHE patients, respectively. Factors associated with a high impact on DLQI (mean 9.7 ± 5.8) were female sex, lesions on back of the hands and pruritus as well as mechanical skin irritation and wearing gloves. CONCLUSION In agreement with recent studies, the Swiss data demonstrate the high impact of CHE on medical well-being, patient quality of life and work ability. As it is associated with an intense use of health care services, high rate of sick leave, job loss and change, CHE may cause a high socio-economic burden.

The Contribution of the Activation Entropy to the Gas-Phase Stability of Modified Nucleic Acid Duplexes

Tricyclo-DNA (tcDNA) is a sugar-modified analogue of DNA currently tested for the treatment of Duchenne muscular dystrophy in an antisense approach. Tandem mass spectrometry plays a key role in modern medical diagnostics and has become a widespread technique for the structure elucidation and quantification of antisense oligonucleotides. Herein, mechanistic aspects of the fragmentation of tcDNA are discussed, which lay the basis for reliable sequencing and quantification of the antisense oligonucleotide. Excellent selectivity of tcDNA for complementary RNA is demonstrated in direct competition experiments. Moreover, the kinetic stability and fragmentation pattern of matched and mismatched tcDNA heteroduplexes were investigated and compared with non-modified DNA and RNA duplexes. Although the separation of the constituting strands is the entropy-favored fragmentation pathway of all nucleic acid duplexes, it was found to be only a minor pathway of tcDNA duplexes. The modified hybrid duplexes preferentially undergo neutral base loss and backbone cleavage. This difference is due to the low activation entropy for the strand dissociation of modified duplexes that arises from the conformational constraint of the tc-sugar-moiety. The low activation entropy results in a relatively high free activation enthalpy for the dissociation comparable to the free activation enthalpy of the alternative reaction pathway, the release of a nucleobase. The gas-phase behavior of tcDNA duplexes illustrates the impact of the activation entropy on the fragmentation kinetics and suggests that tandem mass spectrometric experiments are not suited to determine the relative stability of different types of nucleic acid duplexes.

[Microbiological point of care tests]

It is well known that the early initiation of a specific antiinfective therapy is crucial to reduce the mortality in severe infection. Procedures culturing pathogens are the diagnostic gold standard in such diseases. However, these methods yield results earliest between 24 to 48 hours. Therefore, severe infections such as sepsis need to be treated with an empirical antimicrobial therapy, which is ineffective in an unknown fraction of these patients. Today's microbiological point of care tests are pathogen specific and therefore not appropriate for an infection with a variety of possible pathogens. Molecular nucleic acid diagnostics such as polymerase chain reaction (PCR) allow the identification of pathogens and resistances. These methods are used routinely to speed up the analysis of positive blood cultures. The newest PCR based system allows the identification of the 25 most frequent sepsis pathogens by PCR in parallel without previous culture in less than 6 hours. Thereby, these systems might shorten the time of possibly insufficient antiinfective therapy. However, these extensive tools are not suitable as point of care diagnostics. Miniaturization and automating of the nucleic acid based method is pending, as well as an increase of detectable pathogens and resistance genes by these methods. It is assumed that molecular PCR techniques will have an increasing impact on microbiological diagnostics in the future.

Cost and mortality prediction using polymerase chain reaction pathogen detection in sepsis: evidence from three observational trials

Delays in adequate antimicrobial treatment contribute to high cost and mortality in sepsis. Polymerase chain reaction (PCR) assays are used alongside conventional cultures to accelerate the identification of microorganisms. We analyze the impact on medical outcomes and healthcare costs if improved adequacy of antimicrobial therapy is achieved by providing immediate coverage after positive PCR reports.

[The abdominal catastrophe]

Patients with an abdominal catastrophe are in urgent need of early, interdisciplinary medical help. The treatment plan should be based on medical priorities and clear leadership. First priority should be given to achieve optimal oxygenation of blood and stabilization of circulation during all treatment-phases. The sicker the patient, the less invasive the (surgical) treatment should to be, which means "damage control only". This short article describes 7 important, pragmatic rules that will help to increase the survival of a patient with an abdominal catastrophe. Preexisting morbidity and risk factors must be included in the overall risk-evaluation for every therapeutic intervention. The challenge in patients with an abdominal catastrophe is to carefully balance the therapeutic stress and the existing resistance of the individual patient. The best way to avoid abdominal disaster, however, is its prevention.

Similarities and differences between physician and non-physician acupuncturists in using diagnostic methods - an explorative study

Purpose: Acupuncture is one of the complementary medicine therapies with the greatest demand in Switzerland and many other countries in the West and in Asia. Over the past decades, the pool of scientific literature in acupuncture has markedly increased. The diagnostic methods upon which acupuncture treatment is based, have only been addressed sporadically in scientific journals. The goal of this study is to assess the use of different diagnostic methods in the acupuncture practices and to investigate similarities and differences in using these diagnostic methods between physician and non-physician acupuncturists. Methods: 44 physician acupuncturists with certificates of competence in acupuncture – traditional chinese medicine (TCM) from ASA (Assoziation Schweizer Ärztegesellschaften für Akupunktur und Chinesische Medizin: the Association of Swiss Medical Societies for Acupuncture and Chinese Medicine) and 33 non-physician acupuncturists listed in the EMR (Erfahrungsmedizinisches Register: a national register, which assigns a quality label for CAM therapists in complementary and alternative medicine) in the cantons Basel-Stadt and Basel-Land were asked to fill out a questionnaire on diagnostic methods. The responder rate was 46.8% (69.7% non-physician acupuncturists and 29, 5% physician acupuncturists). Results: The results show that both physician and non-physician acupuncturists take patients’ medical history (94%), use pulse diagnosis (89%), tongue diagnosis (83%) and palpation of body and ear acupuncture points (81%) as diagnostic methods to guide their acupuncture treatments. Between the two groups, there were significant differences in the diagnostic tools being used. Physician acupuncturists do examine their patients significantly more often with western medical methods (p<.05) than this is the case for nonphysician acupuncturists. Non-physician acupuncturists use pulse diagnosis more often than physicians (p<.05). A highly significant difference was observed in the length of time spent with collecting patients’ medical history, where nonphysician acupuncturists clearly spent more time (p<.001). Conclusion: Depending on the educational background of the acupuncturist, different diagnostic methods are used for making the diagnosis. Especially the more time consuming methods like a comprehensive anamnesis and pulse diagnosis are more frequently employed by non-physician practitioners. Further studies will clarify if these results are valid for Switzerland in general, and to what extent the differing use of diagnostic methods has an impact on the diagnosis itself and on the resulting treatment methods, as well as on the treatment success and the patients’ satisfaction.

Esophageal ECG: The challenge of electrode design

Two commercially available electrode catheters are examined for their suitability in esophageal long-term ECG recordings. Both, electrical sensing characteristics as well as clinical acceptance were investigated in a clinical study including inpatients with cardiovascular diseases. In total, 31 esophageal ECG were obtained in 36 patients. Results showed that esophageal electrodes were well tolerated by the patients. Hemispherical electrodes with higher diameter required more insertion attempts and were associated with increased failure rates as compared to cylindrical electrodes. In contrast, the higher surface area of hemispherical electrodes resulted in significantly higher signal-to-noise ratio. Contact impedance was equal for both electrode types, but esophageal electrodes had lower impedance if compared with skin electrodes.

Graph-Based Multi-Surface Segmentation of OCT Data Using Trained Hard and Soft Constraints

Optical coherence tomography (OCT) is a well-established image modality in ophthalmology and used daily in the clinic. Automatic evaluation of such datasets requires an accurate segmentation of the retinal cell layers. However, due to the naturally low signal to noise ratio and the resulting bad image quality, this task remains challenging. We propose an automatic graph-based multi-surface segmentation algorithm that internally uses soft constraints to add prior information from a learned model. This improves the accuracy of the segmentation and increase the robustness to noise. Furthermore, we show that the graph size can be greatly reduced by applying a smart segmentation scheme. This allows the segmentation to be computed in seconds instead of minutes, without deteriorating the segmentation accuracy, making it ideal for a clinical setup. An extensive evaluation on 20 OCT datasets of healthy eyes was performed and showed a mean unsigned segmentation error of 3.05 ±0.54 μm over all datasets when compared to the average observer, which is lower than the inter-observer variability. Similar performance was measured for the task of drusen segmentation, demonstrating the usefulness of using soft constraints as a tool to deal with pathologies.