Percentage Of False Negatives Research Articles

Self-Determination of Eligibility for Medication Abortion Without Ultrasonography Using a History-Based Tool: LMP-SURE.

To evaluate a self-screening eligibility tool for medication abortion without an ultrasonogram. We designed a patient-administered, five-question screening tool (LMP-SURE) that assesses gestational age plus factors associated with misdating or ectopic pregnancy. We recruited participants without prior ultrasonograms from family planning clinics in Alaska, Hawai'i, Idaho, and Utah to complete a brief survey including LMP-SURE and then obtained ultrasound dating by chart review. We compared eligibility for medication abortion by ultrasonogram with eligibility by the LMP-SURE screening tool. We consented 1,026 participants; 781 met eligibility requirements and completed the tool. Using the LMP-SURE tool, we identified 493 participants (65.1%) eligible for medication abortion without an ultrasonogram. The LMP-SURE tool sensitivity (ability to correctly identify a patient ineligible for medication abortion) was 83.8% (95% CI, 73.1-90.8), specificity (ability to correctly identify a patient eligible for medication abortion) was 70.0% (95% CI, 66.4-73.3), likelihood ratio (-) (probability of someone eligible by LMP-SURE to be ineligible by ultrasonogram vs eligible by ultrasonogram) was 0.23 (95% CI, 0.13-0.40), and percentage of false-negatives was 1.5%. Only 11 patients (1.5%) who met eligibility for medication abortion without an ultrasonogram by the LMP-SURE tool were found ineligible for medication abortion by their ultrasonogram. Of those with conflicts, six (0.8%) had a gestational age beyond 77 days. The two participants (0.3%) diagnosed with ectopic pregnancies both required ultrasonograms by LMP-SURE. This patient-facing, brief, history-based screening tool can safely minimize the need for ultrasonogram before medication abortion.

Evaluating Machine Learning Techniques for Web Robot Detection

A web robot, also known as a web crawler, is an automated script that systematically browses the World Wide Web in a methodical manner. Web robots are commonly used by search engines to index web pages, gather information, and update search engine databases. They follow hyperlinks from one web page to another, collecting data and information for various purposes such as indexing, data mining, and website monitoring. Web robot detection is the process of identifying and distinguishing between human users and web robots on websites, which are a major source of web traffic. This process is vital to prevent malicious web robots from having a negative effect on web servers’ traffic and their users’ privacy. Companies such as Imperva, Inc. use machine learning models to identify malicious bots and prevent them from having unauthorized access to an organization’s server. To understand the benefit that machine learning models bring to web robot detection, we used pre-published server log data to construct machine learning models on Orange (a Data Mining Software) and Python that can distinguish between malicious and benign web robots. We evaluated the performance of three well-known machine learning algorithms: kNN, neural network, and decision tree. Based on our experimental results, the neural network gains the highest precision and the lowest false-positive and false-negative percentage of web robots. However, the neural network requires more time to generate the desired output.

A Multi-Stage Method for Logo Detection in Scanned Official Documents Based on Image Processing

A logotype is a rectangular region defined by a set of characteristics, which come from the pixel information and region shape, that differ from those of the text. In this paper, a new method for automatic logo detection is proposed and tested using the public Tobacco800 database. Our method outputs a set of regions from an official document with a high probability to contain a logo using a new approach based on the variation of the feature rectangles method available in the literature. Candidate regions were computed using the longest increasing run algorithm over the document blank lines’ indices. Those regions were further refined by using a feature-rectangle-expansion method with forward checking, where the rectangle expansion can occur in parallel in each region. Finally, a C4.5 decision tree was trained and tested against a set of 1291 official documents to evaluate its performance. The strategic combination of the three previous steps offers a precision and recall for logo detention of 98.9% and 89.9%, respectively, being also resistant to noise and low-quality documents. The method is also able to reduce the processing area of the document while maintaining a low percentage of false negatives.

Digital Pathology in the Detection of Infectious Microorganisms: An Evaluation of Its Strengths and Weaknesses Across a Panel of Immunohistochemical and Histochemical Stains Routinely Used in Diagnostic Surgical Pathology.

The diagnosis of some infectious diseases requires their identification in tissue specimens. As institutions adopt digital pathology for primary diagnosis, the limits of microorganism detection from digital images must be delineated. To assess the reliability of microorganism detection from digitized images of histochemical and immunohistochemical stains commonly used in pathology. Original glass slides from 620 surgical pathology cases evaluated for the presence of infectious microorganisms were digitized. Immunohistochemical stains included those for herpes simplex virus (n = 100), cytomegalovirus (n = 100), Helicobacter pylori (n = 100), and spirochetes (n = 80). Histochemical stains included mucicarmine for Cryptococcus spp (n = 20), Grocott methenamine silver for fungi (n = 100), Giemsa for H pylori (n = 100), and Ziehl-Neelsen for acid-fast bacilli (n = 20). The original diagnosis based on the glass slides was regarded as the reference standard. Six pathologists reviewed the digital images. Digital review was generally associated with high (ie, ≥90%) specificity and positive predictive value owing to a low percentage of false positive reads, whereas a high percentage of false negatives contributed to low sensitivity and negative predictive value for many stains. Fleiss κ showed substantial interobserver agreement in the interpretation of Grocott methenamine silver and immunostains for herpes simplex virus, H pylori, and cytomegalovirus; moderate agreement for spirochete, Ziehl-Neelsen, and mucicarmine; and poor agreement for Giemsa. Digital immunohistochemistry generally outperforms histochemical stains for microorganism detection. Digital interpretation of Ziehl-Neelsen and mucicarmine stains is associated with low scores for interrater reliability, accuracy, sensitivity, and negative predictive value such that it should not substitute for conventional review of glass slides.

Value of repeated US-guided fine-needle aspiration (US-FNAB) in the follow-up of benign thyroid nodules: a real-life study based on the MoCyThy (Modena's Cytology of the Thyroid) DATABASE with a revision of the literature.

The utility of repeating ultrasound-guided fine-needle aspiration (US-FNAB) in the follow-up of benign (THY2) thyroid nodules is still debated. The aim of this study was to retrospectively investigate the diagnostic value of re-biopsy of thyroid nodules following an initially benign result. We retrospectively analyzed US-FNABs performed at the Unit of Endocrinology of Modena from 2006 to 2009. The firstly benign cytological result was compared with the cytological results of subsequent US-FNABs (2nd and/or 3rd) executed on the same nodule. Among 10449 US-FNABs, 6270 (60%) received a THY2 cytological categorization. Of them, 278 (4.43%) underwent a subsequent US-FNAB: 86.7% maintained the same cytology, 32 (11.5%) changed to THY3 (indeterminate) and 5 (1.8%) to THY4 (suspicious of malignancy). Among the 24 nodules addressed to surgery, 9 (37%) were histologically malignant, with an overall miss rate of 3.2%. Male patients had higher risk of discordant results at subsequent US-FNAB (p = 0.005, OR:3.59, 95%CI:1.453-7.769) while dimensional increase above 5 mm was predictive of concordant benign cytology (p = 0.036, OR:0.249, 95%CI:0.068-0.915). Age, suspicious US characteristics, and distance between US-FNABs resulted not predictive. Re-biopsy of benign nodules confirmed the benign nature in most cases. In case of discordant cytology, relocation in indeterminate category was the most common. The histological diagnosis of cancer occurred in one quarter of nodules surgically removed, with a low overall clinically significant miss rate. Thus, a small percentage of false negatives exists; males and subjects with US suspicious nodules should be carefully followed-up, considering case by case re-biopsy possibility.

Optical detection of infiltration during peripheral intravenous infusion in neonates.

Infiltration and extravasation are common complications during peripheral intravenous infusion in the neonatal intensive care unit, and diagnosis is usually clinical, by inspection and palpation. Delay in diagnosis due to poor surveillance or misinterpretation of clinical signs may carry serious damage to the tissues of the neonate. Recently, a novel technology based on optical detection of infiltration has become available. We have studied two groups of term and preterm infants receiving non-vesicant intravenous infusions by the peripheral route (24G short peripheral cannulas), and we evaluated the incidence of infiltration. In the first group, we have compared the clinical detection of infiltration versus the detection obtained by a novel optical device, blinding the alarms of the device. In the second group, the comparison was carried out without blinding the alarms. Of the neonates included in this study, 60% were female, 86% had a gestational age <37 weeks (34 ± 2.5 weeks) and a mean birth weight of 2.08 ± 0.4 g. Total recorded monitoring time was 1318 h and average monitoring time for each short peripheral cannula was 26.4 h. The incidence of infiltration was 80%, most of them having a Millam score <2. The novel device showed an overall sensitivity of 88.9% in detecting infiltration. Specificity was 84.4%, as evaluated assuming as standard the clinical diagnosis. Interestingly, in cases of documented infiltration, the event was diagnosed by the optical device approximately 6 h before the clinical diagnosis. Continuous monitoring of the insertion site, as automatically ensured by this novel optical device, may play a complementary role in early detection of infiltration, even if the percentage of false positives and false negatives suggests that periodic clinical assessment by expert nurses cannot be omitted.

Invasive terrestrial invertebrate detection in water and soil using a targeted eDNA approach

Terrestrial invasive invertebrates can rapidly colonise new areas, causing detrimental effects on biodiversity, economy and lifestyle. Targeted environmental DNA (eDNA) methods could constitute an early detection tool given their sensitivity to small numbers of individuals. We hypothesised that terrestrial runoff would transport eDNA from the land into adjacent waterbodies and used the invasive yellow crazy ant (Anoplolepis gracilipes) as a model species to test this hypothesis. We collected water samples from four waterbodies adjacent (&amp;lt; 10 m from the creek edge) to infestations following rainfall events for eDNA analysis. We also collected soil samples from areas of known infestations and tested five eDNA extraction methods to determine their efficiency to extract eDNA from soil. Water samples resulted in positive yellow crazy ant eDNA amplification (20–100% field replicates across all sites), even at one site located 300 m away from where ants had been detected visually. Soil samples resulted in a higher percentage of false negatives when sampled from ant transit areas than from nest entrances. Unpurified DNA extracts from soil also resulted in false negative detections and only after applying a purification step of DNA extracts, did we detect yellow crazy ant eDNA in 40–100% of field replicates across all methods and sites. This is the first study to empirically show that eDNA from a terrestrial invertebrate can be successfully isolated and amplified from adjacent or downstream waterbodies. Our results indicate that eDNA has the potential to be a useful method for detecting terrestrial invertebrates from soil and water.

The Role of C- Reactive Protein (CRP) in the Diagnosis of Neonatal Sepsis

Background: Neonatal sepsis or septicemia is the term that has been used to describe the systemic response to infection in a newborn infant during the first 28 days of life. Neonatal mortality is increasingly recognized as an important global public health challenge that must be addressed if we wish to reduce child death disparities between rich and poor countries. Most of the estimated 4 million neonatal deaths per year occur in low and middle-income countries. Objective: The aim of this study is to evaluate the role of C-reactive protein (CRP) in the diagnosis of neonatal sepsis. Method of the Study: This is a cross-sectional study carried out in the Department of Neonatology, BSMMU, Dhaka over a period of 6 months between January 2013 to June 2013. Newborns with suspected sepsis admitted to BSMMU were the study population. Clinically diagnosed cases of neonatal sepsis aged &lt; 28 days of both sexes whose parents or guardians provided informed consent were eligible for enrollment in the study. The subjects were selected consecutively from the study population. Data were collected using a structured questionnaire containing all the variables of interest. Data were processed and analyzed using the computer software SPSS (Statistical Package for Social Sciences). The test statistics used to analyze the data were Chi-square (χ2) or Fisher’s Exact Probability Test &amp; Student’s t-Test. Result: The sensitivity of CRP in correctly detecting neonatal sepsis of those who have the disease is 94.44%, while the specificity of the test in correctly differentiating neonates who do not have the disease is 13.72%. The positive predictive value (PPV) of the test is 27.86% and the negative predictive value of the test is 87.5%. The percentages of false positives and false negatives as yielded by the test are 72.13% and 12.5% respectively. Conclusion: The sensitivity of CRP was higher at 94.44% but specificity was low 13.72%. A combination of tests may increase the sensitivity, ........

Internal Quality Control Indicators in Cervical Smear Screening- Report From a Tertiary Care Center, India

Introduction: Quality indicators are one of the tools to monitor the Quality Control (QC) system and have revolutionised the field of laboratory medicine. Internal QC helps in identifying the nonconformities in lab from the moment sample reaches the lab and till the report is being despatched. This study was performed to evaluate the internal QC indicators of cervical smears with an intent to know where we stand, identify the lacunae and to improve performance of lab services. Aim: To evaluate the internal QC indicators of cervical smears in an effort to improve performance of lab services. Materials and Methods: This was a retrospective study conducted in the year 2021. Archived reports of females &gt;18 years of age who had undergone Papanicolaou (PAP) smears between August 2019 and August 2021 were collated from the Department of Pathology, AIIMS, Mangalagiri. Based on these reports, various internal quality indicators, including positivity rate, percentage of Atypical Squamous Cells (ASC) among satisfactory tests, percentage of ASC among abnormal tests {includes Atypical Squamous Cells of Undetermined Significance (ASCUS), ASC-H, Low-grade Squamous Intraepithelial Lesion (LSIL), High-grade Squamous Intraepithelial Lesion (HSIL), carcinoma} ASC/SIL ratio, ASCUS/SIL ratio, percentage of LSIL, percentage of HSIL, percentage of false negatives, percentage of unsatisfactory smears were calculated were calculated. Data was analysed by using the Statistical Package for the Social Sciences (SPSS, version 15.0; IBM, USA) and descriptive statistical analysis was calculated for the quality indicators. Results: A total of 1227 PAP smear cases were analysed in two-year duration, out of which 41 cases were unsatisfactory (3.34%). The annual smear positivity rate ranged from 1.19% to 1.31%, ASC percentage among the abnormal tests- 40%, ASC percentage among the satisfactory tests- 0.50%, percentage of tests with LSIL- 0.08%, percentage of tests with HSIL- 0.25% and false negative rate- 0.16%. ASC/SIL ratio and ASCUS/SIL ratio were 1.50% &amp; 1.25%, respectively. Conclusion: The internal QC indicators obtained in our study were lower than the recommended values by CAP/Bethesda. Thus, achieving benchmark in internal quality indicators is still far from reality as it depends on population screened incidence of cervical lesions and expertise of sampling team and cytopathologists. Regular audit improves screening ability of the test. Thus, every lab should try to achieve the internal quality indicator goals, which will ultimately result in building a good cervical screening system.

SARS-CoV-2 and omicron variant detection with a high selectivity, sensitivity, and low-cost silicon bio-nanosensor.

The recent SARS-CoV-2 pandemic has highlighted the urgent need for novel point-of-care devices to be promptly used for a rapid and reliable large screening analysis of several biomarkers like genetic sequences and antibodies. Currently, one of the main limitations of rapid tests is the high percentage of false negatives in the presence of variants and, in particular for the Omicron one. We demonstrate in this work the detection of SARS-CoV-2 and the Omicron variant with a cost-effective silicon nanosensor enabling high sensitivity, selectivity, and fast response. We have shown that a silicon (Si) nanowires (NW) platform detects both Sars-CoV-2 and its Omicron variant with a limit of detection (LoD) of four effective copies (cps), without any amplification of the genome, and with high selectivity. This ultrasensitive detection of 4cps allows to obtain an extremely early diagnosis paving the way for efficient and widespread tracking. The sensor is made with industrially compatible techniques, which in perspective may allow easy and cost-effective industrialization.

On the choice of negative examples for prediction of host-pathogen protein interactions.

As practitioners of machine learning in the area of bioinformatics we know that the quality of the results crucially depends on the quality of our labeled data. While there is a tendency to focus on the quality of positive examples, the negative examples are equally as important. In this opinion paper we revisit the problem of choosing negative examples for the task of predicting protein-protein interactions, either among proteins of a given species or for host-pathogen interactions and describe important issues that are prevalent in the current literature. The challenge in creating datasets for this task is the noisy nature of the experimentally derived interactions and the lack of information on non-interacting proteins. A standard approach is to choose random pairs of non-interacting proteins as negative examples. Since the interactomes of all species are only partially known, this leads to a very small percentage of false negatives. This is especially true for host-pathogen interactions. To address this perceived issue, some researchers have chosen to select negative examples as pairs of proteins whose sequence similarity to the positive examples is sufficiently low. This clearly reduces the chance for false negatives, but also makes the problem much easier than it really is, leading to over-optimistic accuracy estimates. We demonstrate the effect of this form of bias using a selection of recent protein interaction prediction methods of varying complexity, and urge researchers to pay attention to the details of generating their datasets for potential biases like this.

Accuracy of Dysphagia Standard Assessment (DSA®) bedside screening test: a flowchart for patient eligibility

BackgroundOropharyngeal dysphagia (OD) screening tests have improved patient management; however, the complex applicability and high percentage of false negatives do not allow these tests to be considered completely reliable if not supported by an instrumental investigation. The aim of the present study is to evaluate an OD screening test, the Dysphagia Standard Assessment (DSA®) with different volumes and viscosities.Materials and methodsProspective study of 72 patients evaluated for suspected OD through a double-blind methodology conducted by two operators. All patients underwent fiberoptic endoscopic evaluation of swallowing (FEES) as a reference test and a separate DSA® test. DSA® was performed by administering boluses with different viscosities, with the signal of interruption of the test being: onset of the cough reflex, wet voice after swallowing, and/or desaturation of O2 ≥ 5%. The Penetration-Aspiration Scale (PAS) was evaluated by FEES. The cut-off identified to diagnose OD was PAS ≥ 3.ResultsThe test showed an accuracy of 82%, a sensitivity of 0.93 (95% C.I. 0.84–0.97), and a specificity of 0.78 (95% C.I. 0.67–0.87); positive predictive value 0.55 (95% C.I. 0.43–0.67); negative predictive value 0.97 (95% C.I. 0.90–0.99), positive likelihood ratio 4.37 (95% C.I. 3.6–5.2); likelihood negative ratio 0.08 (95% C.I. 0.06–0.09).ConclusionsAccording to the preliminary results, the test showed good outcomes in determining the presence or absence of OD with a wide spectrum of applicability with some limitations that could be overcome by the selection of a target population. For this reason, a flowchart to address patient eligibility was developed.

DESC-1, a prospective study on epidemiologic factors influencing general population to engage with liquid biopsy based cancer screening.

e13630 Background: Liquid biopsy technology has delivered promising results for therapy monitoring and disease relapse detection. The simplicity of the procedure makes it attractive for early cancer detection despite its clinical value in this setting being as yet unknown. The objective of the study is to evaluate the acceptance of this new screening technology by the general population. Methods: Participants in a digital health study originally selected randomly from the general population of the canton of Geneva (Switzerland) were invited to participate in a survey investigating (1) their motivation and (2) what factors might affect their decision regarding whether to proceed to a liquid biopsy test for cancer screening. Results: 2’898 participants responded to the survey (participation rate: 35%), 1’568 are &gt; 50yo. Two thirds are female, 7.6% have a history of cancer, 80.8% have a family relative with a history of cancer. In the whole population: 97.7% are ready to use liquid biopsy as a cancer screening test. In the &gt; 50yo, 97.3% of women and 98.5% of men would do the test. In the &lt; 50yo, 97.2% of women and 98.6% of men would take it with X² = 0.018 (p = 0.89) for independence with sex. Age category, education level, income band, professional status, self-rated health, tobacco status, chronic disease, personal or familial history of cancer and COVID pandemic have no statistical impact on the incentive to do a liquid biopsy test. In the whole population, 94.6% would take the test knowing that there is a risk of a false positive. The median accepted false positive percentage error rate is 10% with interquartile rate (IQR) Q1: 5% and Q3: 15%. Ninety-six percent of respondents would take the test knowing that there is a risk of a false negative, accepting a median false negative percentage error rate of 10% with IQR Q1: 5% and Q3 : 20%. Most respondents (97.6%) will do the test knowing that additional tests or procedures could be required in case of positivity and 58.8% would take it yearly, while 24.6% only every two years. Fifty-nine percent believe that taking the test could influence their lifestyle, regardless of the result. Conclusions: This study shows that the surveyed population is ready to engage with liquid biopsy as a cancer screening tool despite its limitations and drawbacks. Willingness to take the test might depend on sex only in the &lt; 50yo population. Knowing the lack of demonstrated benefit, it is urgent to conduct trials assessing the clinical value, psychological impact and financial burden of liquid biopsy tests before incorporating this technology into cancer screening programs. Due to the very high level of interest demonstrated in our study, the possibility of direct-to-consumer availability should prompt health authorities and accreditation bodies to carefully weigh the impact of authorizing market access to liquid biopsy technologies for this purpose.

USING A THREE-TIER MULTIPLE-CHOICE DIAGNOSTIC INSTRUMENT TOWARD ALTERNATIVE CONCEPTIONS AMONG LOWER-SECONDARY SCHOOL STUDENTS IN TAIWAN: TAKING ECOSYSTEMS UNIT AS AN EXAMPLE

This research is conducted to identify the scientific conceptual cognition of ecosystem and the corresponding alternative conceptions by lower-secondary school students in Taiwan. Concept mapping, interviewing, and two-tier diagnostic test cannot make explicit reasoning pathways that students may use. Therefore, its purpose is to develop, validate, and utilize a three-tier multiple-choice diagnostic instrument for the ecosystem concept (TDIEC). The instrument can assist teachers in analyzing their students’ reasoning. This does not only distinguish alternative conceptions from lack of knowledge but also determines the percentages of false positives and false negatives. One hundred and six students were tested with this instrument in Taiwan. The data analysis reveals common alternative conceptions shared by students, indicating that 35.59% of the samples had inaccurate conceptualizations of ecosystems. Common alternative conceptions of ecosystems include the following: (i) Only biological components constitute an ecosystem; (ii) The Venus flytrap and pitcher plants are preys; (iii) Competition is synonymous to predation; (iv) Ecological balance refers to the sustainable circulation of energy and matter in the environment; and (v) The simpler the species, the more stable the ecosystem. It concludes with the interpretation of the results, suggestions for the application of the TDIEC, and correct alternative conceptions in Ecology classes. Keywords: Three-Tier Diagnostic Instrument, ecosystem concepts, alternative conceptions, lower-secondary students

Validity of Methods to Identify Individuals With Lower Extremity Amputation Using Department of Veterans Affairs Electronic Medical Records

ObjectivesTo determine the positive predictive value (PPV) of algorithms to identify patients with major (at the ankle or more proximal) lower extremity amputation (LEA) using Department of Veterans Affairs electronic medical records (EMR) and to evaluate whether PPV varies by sex, age, and race. DesignWe conducted a validation study comparing EMR determined LEA status to self-reported LEA (criterion standard). SettingVeterans who receive care at the Department of Veterans Affairs. ParticipantsWe invited a national sample of patients (N=699) with at least 1 procedure or diagnosis code for major LEA to participate. We oversampled women, Black men, and men ≤40 years of age. InterventionsNot applicable. Main Outcome MeasureWe calculated PPV estimates and false negative percentages for 7 algorithms using EMR LEA procedure and diagnosis codes relative to self-reported major LEA. ResultsA total of 466 veterans self-reported their LEA status (68%). PPVs for the 7 algorithms ranged from 89% to 100%. The algorithm that required a single diagnosis or procedure code had the lowest PPV (89%). The algorithm that required at least 1 procedure code had the highest PPV (100%) but also had the highest proportion of false negatives (66%). Algorithms that required at least 1 procedure code or 2 or more diagnosis codes 1 month to 1 year apart had high PPVs (98%-99%) but varied in terms of false negative percentages. PPV estimates were higher among men than women but did not differ meaningfully by age or race, after accounting for sex. ConclusionPPVs were higher if 1 procedure or at least 2 diagnosis codes were required; the difference between algorithms was marked by sex. Investigators should consider trade-offs between PPV and false negatives to identify patients with LEA using EMRs.

Low Accuracy of FIB-4 and NAFLD Fibrosis Scores for Screening for Liver Fibrosis in the Population

Fibrosis-4 (FIB-4) and the nonalcoholic fatty liver disease fibrosis score (NFS) are the 2 most popular noninvasive blood-based serum tests proposed for widespread fibrosis screening. We therefore aimed to describe the accuracy of FIB-4 and NFS to detect elevated liver stiffness as an indicator of hepatic fibrosis in low-prevalence populations. This study included a total of 5129 patients with concomitant measurement of FIB-4, NFS, and liver stiffness measurement (LSM) by Fibroscan (Echosens, France) from 5 independent population-based cohorts from Spain, Hong Kong, Denmark, England, and France; 3979 participants from the general population and 1150 from at-risk cohorts due to alcohol, diabetes, or obesity. We correlated LSM with FIB-4 and NFS, and calculated pre- and post-test predictive values of FIB-4 and NFS to detect elevated LSM at 8 kPa and 12 kPa cutoffs. The mean age was 53 ± 12 years, the mean body mass index was 27 ± 5 kg/m2, and 2439 (57%) were women. One in 10 patients (552; 11%) had liver stiffness ≥8 kPa, but 239 of those (43%) had a normal FIB-4, and 171 (31%) had normal NFS. The proportion of false-negatives was higher in at-risk patients than the general population. FIB-4 was false-negative in 11% of diabetic subjects, compared with 2.5% false-negatives with NFS. Waist circumference outperformed FIB-4 and NFS for detecting LSM ≥8 kPa in the general population. Almost one-third (28%-29%) of elevated FIB-4/NFS were false-positive in both the general population and at-risk cohorts. FIB-4 and NFS are suboptimal for screening purposes due to a high risk of overdiagnosis and a non-negligible percentage of false-negatives, especially in patients with risk factors for chronic liver disease. Waist circumference emerged as a potential first step to identify patients at risk for liver fibrosis in the general population.

Differences in lesion interpretation between radiologists in two countries: Lessons from a digital breast tomosynthesis training test set.

In many western countries, there is good evidence documenting the performance of radiologists reading digital breast tomosynthesis (DBT) images. However, the diagnostic efficiency of Chinese radiologists using DBT, particularly type of errors being made and type of cancers being missed, is understudied. This study aims to investigate the pattern of diagnostic errors across different lesion types produced by Chinese radiologists diagnosing from DBT images. Australian radiologists will be used as a benchmark. Twelve Chinese radiologists read a DBT test set and located each perceived cancer lesion. True positives, false positives (FP), true negatives and false negatives (FN) were generated. The same test set was also read by 14 Australian radiologists. Z-scores and Pearson correlations were used to compare interpretation of lesions and identification of normal appearances between two groups of radiologists. Architectural distortions (p<.001) and stellate masses (p=.02) were more difficult for Chinese radiologists to correctly diagnose compared to their Australian counterparts. Chinese readers categorised more FPs as discrete masses (p<.001) and fewer FPs as architectural distortions (p<.001) comparing with Australian radiologists. The percentages of FN for each cancer case were not correlated (r=0.37, p=.18) but the percentages of FP for each normal case were moderately correlated (r=0.52, p=.02) between two groups of readers. Architectural distortions and stellate masses were challenging to Chinese radiologists when reading DBT. Our findings proposed the need of development of training and education programs focussing on imaging cases tailored for specific groups of readers with certain interpretation patterns.

Overweight in the Dutch National Cohort of Long-Term Survivors of Childhood Cancer

BACKGROUNDOverweight is a common problem in the general population, but occurs more frequently among childhood cancer survivors (CCS) and is regarded as a late adverse effect. However, risk factors are not fully elucidated and it is often disguised in CCS because they can have normal weight but high fat percentage (fat%) on dual-energy X-ray absorptiometry (DXA, gold standard). We aimed to assess overweight prevalence in a nationwide survivor cohort, to clarify risk factors and to identify which measurement method captures overweight best.METHODSThe prevalence of overweight and obesity (body mass index (BMI) ≥25 and ≥30 kg/m 2) was assessed in the Dutch nationwide cohort of adult CCS treated between 1963 and 2002. Risk factors for overweight were analyzed using multivariable logistic regression models. In addition, overweight prevalence was calculated according to fat%, waist circumference (WC), waist/hip ratio (WHR) and waist/height ratio (WHtR). The validity of BMI, WC, WHR and WHtR for characterizing obesity, compared to fat% (expressed as false-negative percentage and in logistic regression models to identify treatment-related risk factors for disguised overweight) was studied.RESULTSA total of 2,338 (51.2% male) survivors (54.7% hematologic malignancies) participated, with mean age 35.5 (±9.3) years and 28.3 (±8.4) years follow-up. In men and women respectively, overweight prevalence was 45.9% and 43.8%, for obesity this was 11.2% and 15.5%.Risk factors for overweight included overweight at cancer diagnosis (adjusted odds ratio (aOR) 3.43, p<0.001), cranial radiotherapy (CRT, aOR 3.27, p<0.001) and growth hormone deficiency (GHD) (unadjusted OR 2.28, p<0.001, after adjustment the effect partially disappeared, aOR 1.60, p=0.072). Previous treatment with corticosteroids was not associated with overweight.Using BMI, WC, WHR and WHtR, similar overweight prevalence was observed. However, this was 58.4% in men and even 83.7% in women when measured with DXA. Disguised overweight was more frequent after treatment with abdominal radiotherapy, high dose anthracyclines and stem cell transplantation (SCT) (aOR up to 3.37).CONCLUSIONSOverweight occurs in almost half of all long-term CCS, and risk factors include overweight at cancer diagnosis, CRT and potentially GHD. DXA identified overweight in an additional 25% of survivors. In CCS treated with abdominal irradiation, anthracyclines and SCT, overweight is more often missed with conventional methods. Hence, in these risk groups DXA needs serious consideration in surveillance, to enable early intervention and prevent complications of overweight including diabetes and atherosclerotic disease. DisclosuresNo relevant conflicts of interest to declare.

Evaluation of Convolutional Neural Network Architecture for Feasibility Analysis on Fetal Abdomen and Brain Images

This research aims to evaluate the possibilities of fetus ultrasound image classification using machine learning algorithms as normal or abnormal. Most of the earlier research works have produced a high percentage of false-negative classification results—recent research work aimed to reduce the rate of false-negative diagnoses. Also, the number of sonologists for analyzing prenatal ultrasound worldwide is very less and solved by developing an efficient algorithm, which reduces the percentage of false negatives in the diagnosis output. Several earlier research works focused on analyzing fetal abdominal image or fetal head images, making the medical industry use two different diagnostic modules separately. This work aims to design and implement a convolution frame-work named as two Convolution Neural Network (tCNN) model for diagnosing any fetal images. The proposed tCNN model diagnoses the fetal abdominal and fetal brain images and classify them as normal or abnormal. CNN1of tCNN performs segmentation and classification based on the acceptance of abdomen circumference and stomach bubble, umbilical vein, and amniotic fluid measurements. CNN2shows based on head circumference and head and abdominal circumference, femur, crown-rump, and humerus lengths measured.With clinical validation, an extensive experiment carried out and the results compared with the experts in terms of segmentation accuracy and the obstetric measurements. This paper provides a foundation for future multi-classification research works on diagnosing fetal intracranial abnormalities and differential diagnosis using machine learning algorithms.

Evaluating Fairness of Algorithmic Risk Assessment Instruments: The Problem With Forcing Dichotomies

Researchers and stakeholders have developed many definitions to evaluate whether algorithmic pretrial risk assessment instruments are fair in terms of their error and accuracy. Error and accuracy are often operationalized using three sets of indicators: false-positive and false-negative percentages, false-positive and false-negative rates, and positive and negative predictive value. To calculate these indicators, a threshold must be set, and continuous risk scores must be dichotomized. We provide a data-driven examination of these three sets of indicators using data from three studies on the most widely used algorithmic pretrial risk assessment instruments: the Public Safety Assessment, the Virginia Pretrial Risk Assessment Instrument, and the Federal Pretrial Risk Assessment. Overall, our findings highlight how conclusions regarding fairness are affected by the limitations of these indicators. Future work should move toward examining whether there are biases in how the risk assessment scores are used to inform decision-making.