Are deep learning classification results obtained on CT scans fair and interpretable?

Following the great success of various deep learning methods in image and object classification, the biomedical image processing society is also overwhelmed with their applications to various automatic diagnosis cases. Unfortunately, most of the deep learning-based classification attempts in the literature solely focus on the aim of extreme accuracy scores, without considering interpretability, or patient-wise separation of training and test data. For example, most lung nodule classification papers using deep learning randomly shuffle data and split it into training, validation, and test sets, causing certain images from the Computed Tomography (CT) scan of a person to be in the training set, while other images of the same person to be in the validation or testing image sets. This can result in reporting misleading accuracy rates and the learning of irrelevant features, ultimately reducing the real-life usability of these models. When the deep neural networks trained on the traditional, unfair data shuffling method are challenged with new patient images, it is observed that the trained models perform poorly. In contrast, deep neural networks trained with strict patient-level separation maintain their accuracy rates even when new patient images are tested. Heat map visualizations of the activations of the deep neural networks trained with strict patient-level separation indicate a higher degree of focus on the relevant nodules. We argue that the research question posed in the title has a positive answer only if the deep neural networks are trained with images of patients that are strictly isolated from the validation and testing patient sets.

Effectiveness of uterine tamponade devices for refractory postpartum haemorrhage after vaginal birth: a systematic review.

OBJECTIVES: To evaluate uterine tamponade devices' effectiveness for atonic refractory postpartum haemorrhage (PPH) after vaginal birth and the effect of including them in institutional protocols. SEARCH STRATEGY: PubMed, EMBASE, CINAHL, LILACS, POPLINE, from inception to January 2021. STUDY SELECTION: Randomised and non-randomised comparative studies. OUTCOMES: Composite outcome including surgical interventions (artery ligations, compressive sutures or hysterectomy) or maternal death, and hysterectomy. RESULTS: All included studies were at high risk of bias. The certainty of the evidence was rated as very low to low. One randomised study measured the effect of the condom-catheter balloon compared with standard care and found unclear results for the composite outcome (relative risk [RR] 2.33, 95% CI 0.76-7.14) and hysterectomy (RR 4.14, 95% CI 0.48-35.93). Three comparative studies assessed the effect of including uterine balloon tamponade in institutional protocols. A stepped wedge cluster randomised controlled trial suggested an increase in the composite outcome (RR 4.08, 95% CI 1.07-15.58) and unclear results for hysterectomy (RR 4.38, 95% CI 0.47-41.09) with the use of the condom-catheter or surgical glove balloon. One non-randomised study showed unclear effects on the composite outcome (RR 0.33, 95% CI 0.11-1.03) and hysterectomy (RR 0.49, 95% CI 0.04-5.38) after the inclusion of the Bakri balloon. The second non-randomised study found unclear effects on the composite outcome (RR 0.95, 95% CI 0.32-2.81) and hysterectomy (RR 1.84, 95% CI 0.44-7.69) after the inclusion of Ebb or Bakri balloon. CONCLUSIONS: The effect of uterine tamponade devices for the management of atonic refractory PPH after vaginal delivery is unclear, as is the role of the type of device and the setting. TWEETABLE ABSTRACT: Unclear effects of uterine tamponade devices and their inclusion in institutional protocols for atonic refractory PPH after vaginal delivery.

Medical versus surgical methods for first trimester termination of pregnancy.

BACKGROUND: Induced abortions are very commonly practiced interventions worldwide. A variety of medical abortion methods have been introduced during the last decade in addition to existing surgical methods. In this review we systematically searched for and combined all evidence from randomised controlled trials comparing surgical with medical abortion. OBJECTIVES: To evaluate medical methods in comparison to surgical methods for first-trimester abortion with respect to efficacy, side effects and acceptability. SEARCH STRATEGY: The Cochrane Controlled Trials Register, MEDLINE (with the Cochrane 3-stage search strategy)(1966-2000) and Popline (1970-2000) were systematically searched. There were no language preferences in searching. Reference lists of retrieved papers were searched. Experts in WHO/HRP were contacted. SELECTION CRITERIA: Randomised trials of any surgical abortion method compared with any medical abortion method in the first trimester. DATA COLLECTION AND ANALYSIS: Trial quality was assessed and data extraction was made independently by two reviewers. MAIN RESULTS: Six studies mostly with small sample sizes, comparing 4 different interventions (prostaglandins alone, mifepristone alone, and mifepristone/misoprostol and methotrexate/misoprostol versus vacuum aspiration) were included. Results are sometimes based on one trial only. Prostaglandins vs vacuum aspiration: the rate of abortions not completed with the intended method was statistically significant higher in the prostaglandin group (2.7, 95% CI 1.1 to 6.8) compared to surgery. There are no data on the most commonly medical (mifepristone/misoprostol) and surgical abortion available to be included in the review. Duration of bleeding was longer in the medical abortion groups compared to vacuum aspiration. There was only one major complication (uterine perforation) in one trial in the surgical group. There was no difference between the groups for ongoing pregnancies at the time of follow-up or pelvic infections. No data on acceptability, side effects or women's satisfaction with the procedure were available for inclusion in the review. AUTHORS' CONCLUSIONS: The results are derived from relatively small trials. Prostaglandins used alone seems to be less effective and more painful compared to surgical first-trimester abortion. However, there is inadequate evidence to comment on the acceptability and side effects of medical compared to surgical first-trimester abortions. There is a need for trials to address the efficacy of currently used methods and women's preferences more reliably.

Effect of the losses in the vocal tract on determination of the area function.

In this work, the cross-sectional areas of the vocal tract are determined for the lossy and lossless cases by using the pole-zero models obtained from the electrical equivalent circuit model of the vocal tract and the system identification method. The cross-sectional areas are used to compare the lossy and lossless cases. In the lossy case, the internal losses due to wall vibration, heat conduction, air friction and viscosity are considered, that is, the complex poles and zeros obtained from the models are used directly. Whereas, in the lossless case, only the imaginary parts of these poles and zeros are used. The vocal tract shapes obtained for the lossy case are close to the actual ones.

Medical versus surgical methods for first trimester termination of pregnancy.

BACKGROUND: Induced abortions are very commonly practiced interventions worldwide. A variety of medical abortion methods have been introduced during the last decade in addition to existing surgical methods. In this review we systematically searched for and combined all evidence from randomised controlled trials comparing surgical with medical abortion. OBJECTIVES: To evaluate medical methods in comparison to surgical methods for first-trimester abortion with respect to efficacy, side effects and acceptability. SEARCH STRATEGY: The Cochrane Controlled Trials Register, MEDLINE (with the Cochrane 3-stage search strategy)(1966-2000) and Popline (1970-2000) were systematically searched. There were no language preferences in searching. Reference lists of retrieved papers were searched. Experts in WHO/HRP were contacted. SELECTION CRITERIA: Randomised trials of any surgical abortion method compared with any medical abortion method in the first trimester. DATA COLLECTION AND ANALYSIS: Trial quality was assessed and data extraction was made independently by two reviewers. MAIN RESULTS: Five studies mostly with small sample sizes, comparing 4 different interventions (prostaglandins alone, mifepristone alone, and mifepristone/misoprostol and methotrexate/misoprostol versus vacuum aspiration) were included. Results are sometimes based on one trial only. Prostaglandins vs vacuum aspiration: the rate of abortions not completed with the intended method was statistically significant higher in the prostaglandin group (2.7, 95% CI 1.1 to 6.8) compared to surgery. There are no data on the most commonly medical (mifepristone/misoprostol) and surgical abortion available to be included in the review. Duration of bleeding was longer in the medical abortion groups compared to vacuum aspiration. There was only one major complication (uterine perforation) in one trial in the surgical group. There was no difference between the groups for ongoing pregnancies at the time of follow-up or pelvic infections. No data on acceptability, side effects or women's satisfaction with the procedure were available for inclusion in the review. REVIEWER'S CONCLUSIONS: The results are derived from small trials. Prostaglandins used alone seems to be less effective and more painful compared to surgical first-trimester abortion. However, there is inadequate evidence to comment on the acceptability and side effects of medical compared to surgical first-trimester abortions. There is a need for trials to address the efficacy of currently used methods and women's preferences more reliably.

Patterns of routine antenatal care for low-risk pregnancy.

BACKGROUND: It has been suggested that reduced antenatal care packages or prenatal care managed by providers other than obstetricians for low risk women can be as effective as standard models of antenatal care. OBJECTIVES: The objective of this review was to assess the effects of antenatal care programmes for low-risk women. SEARCH STRATEGY: We searched the Cochrane Pregnancy and Childbirth Group trials register, reference lists of articles and we also contacted researchers in the field. Date of last search: May 2001 SELECTION CRITERIA: Randomised trials comparing programmes of antenatal care with varied frequency and timing of the visits and different types of care providers. DATA COLLECTION AND ANALYSIS: Trial quality was assessed and data were extracted by two reviewers independently. Study authors were contacted for additional information and they were provided with the final version of the review. MAIN RESULTS: Ten trials involving over 60,000 women were included. Seven trials evaluated the number of antenatal clinic visits, and three trials evaluated the type of care provider. Most trials were of acceptable quality. A reduction in the number of antenatal visits was not associated with an increase in any of the negative maternal and perinatal outcomes reviewed. However, trials from developed countries suggest that women can be less satisfied with the reduced number of visits and feel that their expectations with care are not fulfilled. Antenatal care provided by a midwife/general practitioner was associated with improved perception of care by women. Clinical effectiveness of midwife/general practitioner managed care was similar to that of obstetrician/gynaecologist led shared care. REVIEWER'S CONCLUSIONS: A reduction in the number of antenatal care visits with or without an increased emphasis on the content of the visits could be implemented without any increase in adverse biological maternal and perinatal outcomes. Women can be less satisfied with reduced visits. Lower costs for the mothers and providers could be achieved. While clinical effectiveness seemed similar, women appeared to be slightly more satisfied with midwife/general practitioner managed care compared with obstetrician/gynaecologist led shared care.

WHO systematic review of randomised controlled trials of routine antenatal care.

BACKGROUND: There is a lack of strong evidence on the effectiveness of the content, frequency, and timing of visits in standard antenatal-care programmes. We undertook a systematic review of randomised trials assessing the effectiveness of different models of antenatal care. The main hypothesis was that a model with a lower number of antenatal visits, with or without goal-oriented components, would be as effective as the standard antenatal-care model in terms of clinical outcomes, perceived satisfaction, and costs. METHODS: The interventions compared were the provision of a lower number of antenatal visits (new model) and a standard antenatal-visits programme. The selected outcomes were pre-eclampsia, urinary-tract infection, postpartum anaemia, maternal mortality, low birthweight, and perinatal mortality. We also selected measures of women's satisfaction with care and cost-effectiveness. This review drew on the search strategy developed for the Cochrane Pregnancy and Childbirth Group of the Cochrane Collaboration. FINDINGS: Seven eligible randomised controlled trials were identified. 57418 women participated in these studies: 30799 in the new-model groups (29870 with outcome data) and 26619 in the standard-model groups (25821 with outcome data). There was no clinically differential effect of the reduced number of antenatal visits when the results were pooled for pre-eclampsia (typical odds ratio 0.91 [95% CI 0.66-1.26]), urinary-tract infection (0.93 [0.79-1.10]). postpartum anaemia (1.01), maternal mortality (0.91 [0.55-1.51]), or low birthweight (1.04 [0.93-1.17]). The rates of perinatal mortality were similar, although the rarity of the outcome did not allow formal statistical equivalence to be attained. Some dissatisfaction with care, particularly among women in more developed countries, was observed with the new model. The cost of the new model was equal to or less than that of the standard model. INTERPRETATION: A model with a reduced number of antenatal visits, with or without goal-oriented components, could be introduced into clinical practice without risk to mother or baby, but some degree of dissatisfaction by the mother could be expected. Lower costs can be achieved.

Effect of the signal measured from the glottis on determination of the vocal tract shape.

All-pole and pole-zero models for the vocal tract are developed. First an impulse train, then the pressure signal measured from the glottis, is used as the input in the models. The models for eight Turkish vowels produced by one male subject are studied to determine the effects of the presumed impulse train and the pressure signal measured from the glottis on the estimation of the vocal tract shape. The motion of the tongue is also examined for a whole word.

Who should provide routine antenatal care for low-risk women, and how often? A systematic review of randomised controlled trials. WHO Antenatal Care Trial Research Group.

Many activities, the timing and the frequency of visits of conventional antenatal care provided to low-risk pregnant women have most often been introduced without proper scientific evaluation. Few trials, to date, have been conducted to evaluate the effectiveness of antenatal care programmes for low-risk women with varied number of antenatal visits and type of care providers. We have performed a systematic review of these randomised controlled trials. Five randomised controlled trials were identified in which the effectiveness of a schedule of reduced number of antenatal visits (n = 12,883) was compared with the existing practice (n = 9438). Four of these trials were carried out in developed countries. The difference in the number of visits between intervention and control arms of the trials was moderate. Only one trial achieved a reduction in the median number of visits, from six in the standard care to four in the intervention group, that could be considered to be of health-care relevance for the study population. No significant differences were observed in the two arms of the trials when low birthweight, small-for-gestational-age, Caesarean section, induction of labour, antepartum haemorrhage and postpartum haemorrhage were considered as outcome measures. However, there was a tendency towards an increased rate of preterm delivery in the intervention group in three of the trials. Conversely, the largest trial in Harare, Zimbabwe, demonstrated a statistically significant reduction in preterm delivery in the intervention group (relative risk [RR] 0.88; 95% confidence interval [CI] 0.80, 0.96). Neither the individual studies nor the review had the statistical power to evaluate mortality outcome variables. When perception of care was assessed, women participating in two trials expressed less satisfaction with frequency of visits in the experimental group. In the London, UK, trial, some women in the reduced number of visits group felt that their expectations were not completely fulfilled. However, an opposite trend was reflected when women were asked for their preference of the type of care for any future pregnancy. We also identified three trials that compared midwife/general practitioner-managed care vs. obstetrician/gynaecologist-led shared care. The results were indicative of similar clinical efficacy of the two groups. However, women's response regarding the continuity of care favoured midwife-led care. From the health economics perspective, there is evidence of cost reduction if antenatal care was provided by staff other than the obstetrician/gynaecologist. The available data demonstrate no significant differences in selected perinatal outcomes for low-risk women receiving care according to a reduced frequency (approximately two visits fewer) of prenatal visits vs. those following the existing practice. However, there are differences in satisfaction with the prenatal care provider and the prenatal care system. There is evidence that a midwife's clinic for provision of antenatal care for low-risk pregnancies is feasible and therapy reduction in costs achievable.

Nutritional interventions to prevent intrauterine growth retardation: evidence from randomized controlled trials.

This report summarizes the evidence from systematic reviews of randomized controlled trials on the effectiveness of nutritional interventions aimed at reducing intrauterine growth retardation (IUGR). There were 12 interventions including protein-energy, vitamin, mineral, and fish oil supplementation, as well as the prevention and treatment of anemia and hypertensive disorders. A primary concern is the limited data supporting the effectiveness of recommended nutritional interventions during pregnancy, some of which are widely used even in women without nutritional deficiencies. Overall, with the exception of perhaps balanced protein/energy supplementation (typical odds ratio: 0.77; 95 per cent confidence interval: 0.58 to 1.01), no effective nutritional interventions for reducing the risk of IUGR have been demonstrated. Other interventions, such as zinc, folate and magnesium supplementation during gestation, merit further research which should be conducted among populations at risk of IUGR, using larger sample sizes, and addressing coexisting factors limiting fetal growth. Appropriate combinations of interventions should be a priority for evaluation as it is unlikely that a single factor will reduce a multicausal outcome like IUGR that is so dependent on socioeconomic disparities.

Effectiveness of interventions to prevent or treat impaired fetal growth.

This is an overview of 126 randomized controlled trials (RCTs) evaluating 36 prenatal interventions to prevent or treat impaired fetal growth (IFG). Results are based on systematic reviews including the meta-analyses of these RCTs. Most of the prenatal interventions do not show any significant effects on short-term perinatal outcomes. There are, however, a few interventions likely to be beneficial: smoking cessation, antimalarial chemoprophylaxis in primigravidae, and balanced protein/energy supplementation. Others merit further research: zinc, folate, and magnesium supplementation during gestation. Appropriate combinations of interventions should be a priority for evaluation because it is unlikely that a single intervention will reduce a multicausal outcome like IFG that is so dependent on socioeconomic disparities. Of concern is the discrepancy between the importance given in the epidemiological and clinical literature to the problem of IFG and the methodological quality and sample size of the RCTs conducted for the evaluation of preventive or treatment modalities.

[The distribution of Candida species isolated from vaginal discharge]. / Vajinal akintilardan izole edilen Candida türlerinin dagilimi.

In this study we performed swab cultures from vaginal discharge of 100 patients with various gynecological and obstetrical symptoms during visits to Ankara University Medical Faculty, Department of Gynecology and Obstetrics and Dr. Zekai Tahir Burak Women's Hospital Clinics. 23 Candida spp. were isolated and the distribution of the strains were as follows; C. albicans (16), C. glabrata (3), C. pseudotropicalis (1), C. parapsilosis (1). Two strains were untypable. Our findings indicate that C. albicans has an important role in patients with vaginal discharge.