False-positive Cases Research Articles

A Decision Tree-Driven IoT systems for improved pre-natal diagnostic accuracy

Prenatal diagnostics are vital for the woman as well as her unborn baby. The diagnostics help in the early identification of the possibility of complication and the initial measures that help to ameliorate the mother and the fetus health status are taken. Over the year’s various techniques have been employed in diagnosing genetic disorders before birth that lack effectiveness in terms of cost, time, and places to access ultra-modern health facilities. To overcome these problems, this paper puts forward a diagnostic model that integrates Internet of Things innovation with a Machine Learning approach which is the Decision Tree Algorithms. First, it implies the application of IOT devices in the collection of vital information like heart rate, blood pressure, glucose levels, and fetal movement. The data is structured in the form of a dataset and transmitted to a Big Data storage for warehousing and processing. Secondly, the DTA is employed to analyze the data and look for patterns and possibilities of future health complications. The DTA operates in that it divides the dataset into subsets considering specific features and formulates a tree-like model of decisions. At every node, the algorithm chooses the attribute which has the highest information gain, to partition the data into different classes. This process goes on until it reaches a decision node through which, it can decide probable health problems from the input data. To increase the reliability of the developed model this study fine-tunes the model by using a large database of pre-natal health records. The system is capable of collecting data in real-time and flagging data that needs attention in the case of any abnormality to the health professional. The above methodology was tested on a 1000-record database of pre-natal health records where the proposal achieved 95% possibility of potential health problems as against 85% by classical statistical analysis. Furthermore, the system scaled down the number of false positive cases by 20 percent and false negatives by 15 percent thus the efficacy of the system.

Limitations of using artificial intelligence services to analyze chest x-ray imaging

BACKGROUND: Chest X-ray examination is one of the first radiology areas that started applying artificial intelligence, and it is still used to the present. However, when interpreting X-ray scans using artificial intelligence, radiologists still experience several routine restrictions that should be considered in issuing a medical report and require the attention of artificial intelligence developers to further improve the algorithms and increase their efficiency. AIM: To identify restrictions of artificial intelligence services for analyzing chest X-ray images and assesses the clinical significance of these restrictions. MATERIALS AND METHODS: A retrospective analysis was performed for 155 cases of discrepancies between the conclusions of artificial intelligence services and medical reports when analyzing chest X-ray images. All cases included in the study were obtained from the Unified Radiological Information Service of the Unified Medical Information and Analytical System of Moscow. RESULTS: Of the 155 analyzed difference cases, 48 (31.0%) were false-positive and 78 (50.3%) were false-negative cases. The remaining 29 (18.7%) cases were removed from further studies because they were true positive (27) or true negative (2) in the expert review. Most (93.8%) of the 48 false-positive cases were due to the artificial intelligence service mistaking normal chest anatomy (97.8% of cases) or catheter shadow (2.2% of cases) for pneumothorax signs. Overlooked clinically significant pathologies accounted for 22.0% of false-negative scans. Nearly half of these cases (44.4%) were overlooked lung nodules. Lung calcifications (60.9%) were the most common clinically insignificant pathology. CONCLUSIONS: Artificial intelligence services demonstrate a tendency toward over diagnosis. All false-positive cases were associated with erroneous detection of clinically significant pathology: pneumothorax, lung nodules, and pulmonary consolidation. Among false-negative cases, the rate of overlooked clinically significant pathology was low, which accounted for less than one-fourth.

Identifying Patients with Heart Failure Eligible for Guideline-Directed Medical Therapy.

A majority of patients with heart failure (HF) do not receive adequate medical therapy as recommended by clinical guidelines. One major obstacle encountered by population health management (PHM) programs to improve medication usage is the substantial burden placed on clinical staff who must manually sift through electronic health records (EHRs) to ascertain patients' eligibility for the guidelines. As a potential solution, the study team developed a rule-based system (RBS) that automatically parses the EHR for identifying patients with HF who may be eligible for guideline-directed therapy. The RBS was deployed to streamline a PHM program at Brigham and Women's Hospital wherein the RBS was executed every other month to identify potentially eligible patients for further screening by the program staff. The study team evaluated the performance of the system and performed an error analysis to identify areas for improving the system. Of approximately 161,000 patients who have an echocardiogram in the health system, each execution of the RBS typically identified around 4200 patients. A total 5460 patients were manually screened, of which 1754 were found to be truly eligible with an accuracy of 32.1%. An analysis of the false-positive cases showed that over 38% of the false positives were due to incorrect determination of symptomatic HF and medication history of the patients. The system's performance can be potentially improved by integrating information from clinical notes. The RBS provided a systematic way to narrow down the patient population to a subset that is enriched for eligible patients. However, there is a need to further optimize the system by integrating processing of clinical notes. This study highlights the practical challenges of implementing automated tools to facilitate guideline-directed care.

Mid-Trimester Uterine Artery Doppler for Aspirin Discontinuation in Pregnancies at High Risk for Preterm Preeclampsia: Post Hoc Analysis of StopPRE Trial

(BJOG. 2024;131(3):334–342. doi: 10.1111/1471-0528.17631) Preeclampsia (PEC) is characterized by hypertension and proteinuria or organ dysfunction after 20 weeks of gestation. The cause of the condition has not been fully elucidated; however, PEC is associated with a defect in the development of the placenta in early-onset patients. In unaffected individuals, uterine artery flow impedance (UtA) decreases during early gestation through week 24. However, if pulsatility index (PI) in UtA remains high, it is a sign of maladaptation and is commonly found in patients with preterm PEC. An additional consequence of impaired placentation is uteroplacental ischemia, during which the levels of placental growth factor (PIGF) are decreased and soluble fms-like tyrosine kinase-1 (sFlt-1) increases. In screening first trimester patients for risk of PEC, UtAPI and PIGF may be used. Clinicians prescribe aspirin daily as a preventative measure in patients who screen at risk for PEC. Although the incidence of PEC greatly decreases with this treatment, it is also noteworthy that screening for risk of PEC during the first trimester results in a high false positive rate. Thus, most women pre-emptively receiving aspirin for PEC are false positives. The authors previously performed a study where aspirin was discontinued in patients between 24 and 28 weeks of gestation with a normal sFlt-1/PlGF ratio of ≤38, identifying false positives cases. Although this study showed promising results, the ratio may not be available for use in every clinic. Thus, in this study, the authors examine whether UtAPI can be a correlate for these angiogenic factors and act as an alternative screening measure at 24 to 28 weeks of gestation with the same purpose to discontinue aspirin and assess incidence of preterm PEC.

Case Ascertainment in Pediatric Heart Failure Using International Classification of Disease Clinical Modification (ICD-CM) Codes.

Most epidemiological studies in pediatric heart failure (HF) use administrative database sources, defining patient cohorts by presence of a single HF ICD code. However, the ability of ICD codes to identify true HF patients is unknown in pediatrics.Here we describe the accuracy of HF ICD-10-CM code search algorithms, in identifying pediatric patients with HF from electronic data sources. Based on the adult HF literature, search algorithms were designed to incorporate HF ICD codes, imaging, and medications. Sensitivity, specificity, positive and negative predictive value and accuracy of the algorithms were tested among children in an advanced HF clinic ("Clinic cohort"). Top-performing algorithms were then tested in a large-scale regional electronic data warehouse (EDW), 01/2017 to 01/2020, generating the "EDW Cohort". False positive cases were identified and characterized by chart review. Within the Clinic Cohort, 78/378 patients (21%) had gold standard HF diagnoses. A search algorithm with one HF ICD coded visit was more sensitive but less specific than > 1 HF ICD coded visit, (sensitivity 94% and specificity 89% versus 69% and 97%, respectively). Correspondingly, > 1 ICD coded visit had a higher PPV than one ICD coded visit; 84% vs. 69%. Accuracy was similar (90% vs 91%). Presence of 1 HF ICD code combined with HF medication had high sensitivity, specificity, PPV, NPV and accuracy, all higher than the single ICD code algorithm. The "1 HF coded visit + any medication" algorithm resulted in highest accuracy (93%). Top-performing algorithms were tested in the EDW: the algorithm with > 1 HF ICD coded visit, and the algorithm with one HF ICD coded visit combined with HF medication. In the EDW Cohort, 133/248 (53.6%) patients had gold standard HF diagnoses though 115/248 (46.3%) were false positive cases; 41% of those had pulmonary over-circulation from congenital heart disease. Excluding children < 30days old and those with a history of an isolated VSD repair, complete AVSD repair, or PDA closure further reduced the proportion of false positives to 50/248 (20%). A search algorithm using a single HF ICD code can have acceptable sensitivity, specificity, PPV, NPV and accuracy in identifying children with HF from within electronic medical records. Similar to adult HF literature, specificity improves by including HF medication. When applied to large data sources, however, the search algorithms result in a high proportion of patients with pulmonary overcirculation related to congenital heart disease. To narrow the population to those with myocardial dysfunction, case identification may require use of ICD codes with linked of administrative, surgical, and/or imaging databases.

False-Positive Diagnosis of Congenital Heart Defects at First-Trimester Ultrasound: An Italian Multicentric Study.

Objective. Our objective was to assess the proportion of false-positive CHD cases at the first-trimester evaluation of the fetal heart, performed by experienced operators. Methods. This multicenter retrospective study included of pregnant women with suspicion of CHDs during first-trimester screening for aneuploidies. In all cases, the fetal heart assessments were performed by obstetricians with extensive experience in first-trimester scanning, following an extended protocol proposed by SIEOG national guidelines, which included an axial view of the fetal abdomen and chest to assess visceral situs and evaluation of the four-chamber view (4CV) and three-vessel trachea view (3VTV) with color Doppler. In all suspected cases, fetal echocardiography was offered within 16 and/or at 19-22 weeks' gestation. Results. From a population of 4300 fetuses, 46 CHDs were suspected. Twenty-four cases were excluded from this analysis because the parents opted for early termination of the pregnancies due to associated structural and/or genetic anomalies. For the remaining 22, echocardiography was performed by 16 weeks in 14 cases (64%) and after 16 weeks in 8 cases. In 19 cases (86.4%), a fetal cardiologist confirmed the presence of a CHD. In three cases (13%), the cardiac anatomy was found to be normal at the fetal echocardiography and postnatally. Conclusions. This study shows that the proportion of false-positive cases at the first-trimester ultrasound examination of the fetal heart, performed by experienced operators, may carry a higher risk of false-positive diagnosis than expected. Therefore, this issue must be discussed in instances where a CHD is suspected at the first-trimester screening.

False-positive XXY results by interphase FISH in cytogenetically normal XX individuals: two cases highlighting the necessity of additional laboratory follow-up

BackgroundInterphase fluorescence in situ hybridization (FISH) is commonly used for rapid aneuploidy detection in clinical settings. While FISH-based aneuploidy detection provides rapid results desirable for patient management, it usually only utilizes one probe per chromosome, which may lead to rare false-positive findings.Case presentationHere we report two interphase FISH results, which were false-positive for XXY in cytogenetically normal XX individuals. Both false-positive cases were due to hybridization of the Y chromosome centromeric probe DYZ3 to the pericentromeric region of chromosome 15. In both cases, chromosomal microarray revealed no detectable Y chromosome material, suggesting the hybridizations of the DYZ3 probe to chromosome 15 likely represent benign heterochromatic variants of no clinical significance. In one case, the DYZ3 hybridization was also identified in the phenotypically unaffected mother, further suggesting this is likely a rare variant of no clinical significance.ConclusionsThis report marks the first documentation of hybridization of the DYZ3 probe to another chromosome in cytogenetically normal individuals. Our report has important clinical implications, because DYZ3 is widely used by clinical laboratories for Y chromosome detection. Our findings underscore the necessity of confirming abnormal aneuploidy detection FISH results with additional laboratory methods such as chromosomal microarray analysis.

NIMG-73. FALSE POSITIVE INTRA-OPERATIVE MAGNETIC RESONANCE IMAGING DURING HIGH GRADE GLIOMA RESECTION

Abstract Intraoperative magnetic resonance imaging (iMRI) allows assessment of extent of resection during surgery for high-grade glioma (HGG). In most cases, contrast-enhancing areas noted on iMRI indicate residual tumor. However, false positives may occur, leading to inappropriate additional resection and patient morbidity. The rate and pattern of false positive findings in iMRI have not been fully described. We analyzed surgical and imaging findings in patients undergoing iMRI during resection of HGG to determine factors linked with potential false positive contrast enhancement. A single-center analysis retrospectively assessed consecutive patients undergoing resection of newly diagnosed high-grade glioma using iMRI guidance. Preoperative, intraoperative and postoperative imaging, as well as pathologic findings were assessed. One hundred seventy-four consecutive patients who underwent surgery from 2016-2023 were included. All patients had preoperative MRI showing contrast-enhancing tumor and all underwent one iMRI during surgery to assess extent of resection. 124/174 cases (71.3%) demonstrated contrast enhancement on iMRI that prompted inspection and additional resection using updated neuronavigation. Enhancement corresponding to residual tumor was identified with or without additional removal in 120/124 cases (96.8%). No residual tumor was found in 4/124 cases (3.2%, false positives). All false positive cases were the result of contrast accumulation in hematomas (2), hemostatic agents (1) or dependent portions of the resection cavity (1). Both hematomas were located in the walls of the resection cavity and confirmed as hematoma, rather than tumor, by separate pathology analysis. False positives were not associated with patient positioning, tumor sub-type, anatomic location of tumor or extent of resection. Gross total resection (GTR) was confirmed in 163/174 (93.7%) cases on postoperative imaging. False-positive iMRI during HGG surgery can mimic residual tumor. This occurs at a low rate due to contrast pooling in the cavity or accumulation within hemostatic agents or hematomas along the cavity wall. Meticulous hemostasis when preparing the patient for iMRI may help prevent false positives and prevent unnecessary additional resection.

The effect of an enhanced fetal growth ultrasound protocol on pregnancy outcomes: A retrospective service evaluation within a single UK National Health Service centre between 2014 and 2022.

Growth Assessment Protocol is a fetal growth initiative designed to improve antenatal detection of babies who are small-for-gestational-age and reduce stillbirths. However, its direct impact on pregnancy outcome and stillbirth rates is questioned. This service evaluation aimed to assess Growth Assessment Protocol's influence on pregnancy outcomes at a National Health Service hospital. Anonymous, maternity and ultrasound data, routinely acquired between 2014 and 2022 were extracted from clinical databases (Viewpoint, Euroking). Trends in maternity data and ultrasound scan volume were explored with descriptive statistics. Variables of stillbirth, antenatal small-for-gestational-age detection and scan volume were compared before and after Growth Assessment Protocol implementation. Associations between these variables were evaluated using Spearman's rho. The percentage of babies born small-for-gestational-age reduced by 0.3% across the evaluation period. Antenatal small-for-gestational-age diagnosis rose from 4.1% to 14.3%. However, the number of false-positive cases of antenatally diagnosed small-for-gestational-age increased fivefold from 2.2% to 11.5%. Although stillbirth rates remained consistent post-Growth Assessment Protocol, complex scan volume (e.g. number of growth scans using Doppler) increased annually. The peak incline coincided with the Growth Assessment Protocol implementation period (2016-2018). Complex scan volume was significantly associated with overall small-for-gestational-age detection (rho = 0.8, p =< 0.001), but not with stillbirth frequency (rho = -0.1, p = 0.4). Small-for-gestational-age detection increased following Growth Assessment Protocol implementation, although this was associated with a high false-positive rate and no reduction in stillbirths. The potential implications associated with clinical management, parent experiences and departmental workflow, alongside the benefits for stillbirth reduction, should be fully considered prior to the introduction of a new fetal growth initiative to the antenatal care pathway.

Multi-centre post-mortem study on the use of coronary inflammation in CCTA to improve the diagnosis of the coronary related cause of death with histopathology validation

Abstract Background/Introduction A limitation of post-mortem computed tomography (PMCT) is the diagnostic accuracy of coronary related death, which constitute majority of investigations. Measurements using coronary inflammation with fat attenuation index (FAI) Score, clinical risk factors and coronary plaque metrics from CCTA to predict fatal cardiac events (AI-Risk) in living patients has not been used in post-mortem setting. Purpose This study investigates two new techniques to aid the diagnosis of coronary related death in the post-mortem setting: (1) minimally invasive interventional technique to perform PMCT; (2) diagnostic accuracy of FAI-assisted PMCT with histopathology validation. Method Adults with sudden death of unknown cause &amp;lt;48 hours from death were included. Ultrasound guided femoral access to deliver guidewire to the aortic root (Figure 1A), followed by inflation of aortic occlusion balloon and injection of contrast for angiogram. A cardiac radiologist performed the coronary and FAI analysis (1B &amp;C). Two experienced radiologists reported the cause of death without/with aid of FAI score results. The technique was introduced at a second PMCT centre. Results PMCT analysis of 37 participants (27 males, 10 females, age range 42-98) – 15 minimally invasive PMCT, and 22 PMCT from previous study showed FAI score (1D) and AI-Risk (1E) were significantly higher in participants who had coronary related death vs non-coronary related causes. In the receiver operating characteristic (ROC, 1F) analysis, thresholding at 42% for AI-Risk and 34 for Mean FAI score, was able to detect all 18 true positives, 2 false positives (1xpulmonary embolism, and 1xdual pathology of ruptured abdominal aneurysm with severe coronary artery disease; these cases showed high FAI score and AI-Risk), and all true negatives. Both radiologists identified 17/18 coronary related death. There were 5 cases of false positives for coronary related death in Radiologist 1 (3 years’ experience); 8 false positives in Radiologists 2 (10 years’ experience). In non-coronary related deaths, Radiologist 1 identified 14/19 true negative cases, and Radiologist 2 identified 11/19 true negative cases. In all the false positive and negative cases, the radiologists’ diagnostic confidence level was either possible, or unascertain. With the aid of FAI score results, all the true coronary related deaths were identified and no false negative case. This method was introduced at a second centre with 108 PMCTs analysed. There were 40 cases of non-coronary death, and 68 cases of coronary related death on PMCT (imaging diagnosis only). In the non-coronary group, FAI-score assisted results showed 100% concordance. In the coronary group, there was 10% dis-concordance (7/68). Conclusion Minimally invasive PMCT with FAI score-assisted diagnosis has the potential to increase the diagnostic confidence and accuracy of coronary-related cause of death.

Immediate Diagnosis of Breast Carcinoma on Core Needle Biopsy Using Ex Vivo Fluorescence Confocal Microscopy: Feasibility in a One-Stop Breast Clinic Workflow.

In the one-stop breast clinic setting, breast cytology traditionally provides immediate diagnosis of carcinoma. Fluorescence confocal microscopy (FCM) is an emerging optical technique enabling ex vivo analysis of breast biopsies in real-time. This study represents the first proof of concept for integrating FCM imaging into the routine workflow of breast core needle biopsies (CNB) at Gustave Roussy's one-stop breast clinic. Fifty women with breast masses underwent consecutive enrollment. Biopsies were stained with acridine orange and fast green, followed by imaging using the Vivascope 2500M-G4 (FCM). Interpretation was conducted by two pathologists in real time (PT1) or postoperatively (PT2). Concordance with definitive histology, the duration of the FCM protocol, and its impact on conventional histopathology, immunohistochemistry, and FISH analyses were evaluated. In our study of 50 biopsies, a concordant diagnosis of malignancy was performed using FCM on the malignant cases at definitive histology in 93.5% (29/31 cases) and in 90.3% (28/31 cases) according to PT1 and PT2, respectively. When the FCM suspicious cases were added, FCM identified 100% (31/31 cases) and 96.7% (30/31 cases) of the malignant cases according to PT1 and PT2, respectively. A notable false positive case was identified as a complex sclerosing lesion. The median time for sample preparation (including tissue reception) was 5 min, while the median time for imaging acquisition with interpretation was 3 min for PT1, but 1 min required for interpretation alone by PT2. Histopathological alterations were not more prevalent in FCM-imaged biopsies compared to conventionally treated biopsies. The immunophenotyping and molecular assessment of tissue were preserved after FCM protocol. FCM shows promise as a new histological method for the immediate diagnosis of breast carcinoma on core needle biopsies in a one-stop clinic setting, while also preserving tissue specimens for final histology.

High mortality rates among individuals misdiagnosed with tuberculosis: a matched retrospective cohort study of individuals diagnosed with tuberculosis in Brazil.

We conducted a matched retrospective cohort study comparing mortality among individuals receiving a false-positive tuberculosis diagnosis (n=3701) to individuals correctly diagnosed with TB (n=8595) in Brazil from 2007-2016. Over an average 5.4-year follow-up period, we estimated a mortality rate ratio of 1.95 (95% confidence interval: 1.80, 2.11) for individuals incorrectly diagnosed with TB compared to controls. The leading causes of death among the misdiagnosed were malignant neoplasms (40.9%) and respiratory system disorders (15.9%), conditions with symptoms similar to tuberculosis. Our findings highlight the need for improved follow-up care after identification of false-positive cases to increase survival for this high-risk population.

Case Ascertainment in Pediatric Heart Failure Using International Classification of Disease Clinical Modification (ICD-CM) Codes.

Most epidemiological studies in pediatric heart failure (HF) use administrative database sources, defining patient cohorts by presence of a single HF ICD code. However, the ability of ICD codes to identify true HF patients is unknown in pediatrics. Here we describe the accuracy of HF ICD-10-CM code search algorithms, in identifying pediatric patients with HF from electronic data sources. Based on the adult HF literature, search algorithms were designed to incorporate HF ICD codes, imaging, and medications. Sensitivity, specificity, positive and negative predictive value and accuracy of the algorithms were tested among children in an advanced HF clinic ("Clinic cohort"). Top-performing algorithms were then tested in a large-scale regional electronic data warehouse (EDW), 01/2017 to 01/2020, generating the "EDW Cohort". False positive cases were identified and characterized by chart review. Within the Clinic Cohort, 78/378 patients (21%) had gold standard HF diagnoses. A search algorithm with one HF ICD coded visit was more sensitive but less specific than >1 HF ICD coded visit, (sensitivity 94% and specificity 89% versus 69% and 97%, respectively). Correspondingly, >1 ICD coded visit had a higher PPV than one ICD coded visit; 84% vs. 69%. Accuracy was similar (90% vs 91%). Presence of 1 HF ICD code combined with HF medication had high sensitivity, specificity, PPV, NPV and accuracy, all higher than the single ICD code algorithm. The "1 HF coded visit + any medication" algorithm resulted in highest accuracy (93%). Top-performing algorithms were tested in the EDW: the algorithm with > 1 HF ICD coded visit, and the algorithm with one HF ICD coded visit combined with HF medication. In the EDW Cohort, 133/248 (53.6%) patients had gold standard HF diagnoses though 115/248 (46.3%) were false positive cases; 41% of those had pulmonary over-circulation from congenital heart disease. Excluding children < 30 days old and those with a history of an isolated VSD repair, complete AVSD repair, or PDA closure further reduced the proportion of false positives to 50/248 (20%). A search algorithm using a single HF ICD code can have acceptable sensitivity, specificity, PPV, NPV and accuracy in identifying children with HF from within electronic medical records. Similar to adult HF literature, specificity improves by including HF medication. When applied to large data sources, however, the search algorithms result in a high proportion of patients with pulmonary overcirculation related to congenital heart disease. To narrow the population to those with myocardial dysfunction, case identification may require use of ICD codes with linked of administrative, surgical, and/or imaging databases.

Defining sensitivity and specificity for quality control of organized cancer screening using Prefectural Cancer Registry data in Japan

Objective To decrease cancer mortality by implementing cancer screening programs, rigorous quality control measures that utilize standardized indicators are imperative. In Japan, although each municipality performing cancer screening programs implements quality control for their programs using the checklist authorized by the Ministry of Health, Labour and Welfare, compliance with all the items listed is not possible because calculating sensitivity and specificity using cancer registry data is difficult under these circumstances. This report elucidates the methodology for calculating indicators, including sensitivity and specificity, by delineating the parameters of false-negative cases within population-based cancer screening programs in Japan. Furthermore, the inherent challenges associated with ensuring the quality control of cancer screening procedures are expounded upon in this report.Method Data from the Prefectural Cancer Registry of Japan and cancer screening records compiled by municipalities were used to differentiate true-positive, true-negative, false-positive, and false-negative cases based on the combination of screening test outcomes and subsequent cancer incidence.Results A false-negative case was defined as an examinee who received a cancer diagnosis within one year after undergoing the screening test, notwithstanding the negative judgment of the cancer screening decision. The duration for judgment of true-positive, true-negative, and false-negative cases was also extended to one year. Cancer identification after cancer screening was ascertained using data from the Prefectural Cancer Registry, ensuring uniform categorization of the four cases. Subsequently, sensitivity and specificity values were calculated for municipalities conducting cancer screening programs.Conclusion Sensitivity and specificity are indispensable metrics for the inherent quality control of cancer screening because these parameters directly assess the efficacy of screening tests. The anticipated increase in the number of municipalities engaged in comprehensive quality control of cancer screening in Japan is poised to enhance the efficiency of cancer control policies. This augmentation will be accomplished through the meticulous utilization of the sensitivity and specificity values elucidated in the present report. The forthcoming challenges involve the proliferation of medical institutes reporting their adherence to the checklist stipulated by the National Cancer Center of Japan and the widespread dissemination of fundamental knowledge pertaining to cancer screening.

B-320 Understanding CA 72-4: Exploring False Positives in Clinical Context

Abstract Background CA 72-4, initially identified in 1981 as an antigen reactive to murine antibodies, is a glycoprotein expressed on the cell surface of certain carcinoma cells. Its production is elevated in various cancers, including ovarian, pancreatic, colorectal, and gastric cancer. However, CA 72-4 serum concentration may rise in non-cancerous conditions, leading to potential misdiagnosis and a low positive predictive value (PPV), particularly in asymptomatic patients. Our aim in this study was to describe CA 72-4 false positive values and assess the pathological conditions and pharmacological treatments associated with this. We also intended to compare other tumour markers status in false and true positive results of CA 72-4. Methods A retrospective analysis form pseudo-anonymized database was conducted using data from 1290 laboratory analysis from different patients taking a blood test at our laboratory hospital from 2017 to 2019. CA 72-4 and other tumour markers were analysed by ECLIA (Roche Diagnostics Gmbh. Mannheim, Germany). Results Among 1290 CA 72-4 laboratory tests in different patients, 204 were considered positive because they were above the reference interval (&amp;lt;7 U/mL). Within this subgroup, 44 of the 204 were clearly classified as false positives, as patients were not diagnosed with any form of neoplastic process. 45% of cases were between 7-14 U/mL (group 1), 30% were between 14-28 U/mL (group 2), and finally 25% of patients were above 28 U/mL ranging from 35.1 to 240.4 U/mL (group 3). Interestingly, CA19-9, CA 125 and CEA were measured in 37 of the false positive cases and it was found that 59.5% had at least another tumour marker above the reference range. In 103 true positive cases it was found that 88% of patients had at least one tumour marker above the normal range. Patients exhibiting false positive results for CA 72-4 demonstrated a wide spectrum of pathological conditions, with pulmonary disease, inflammation, infection, and heart and kidney disease emerging as the most prevalent. Group 1 patients most common underlying pathologies were pulmonary disease in 25%, heart disease in 25% and either viral or bacterial infection in 60% of patients. Group 2 patients most common pathologies were heart disease in 42%, pulmonary disease in 33% and infection in 25% of patients. In group 3 patients, renal disease was present in 45% of patients, heart disease was found in 45% of patients and inflammation/sepsis was found in 36% of patients. When looking into pharmacological treatment, 86.3% of patients were taking at least one of previously described treatments increasing CA 72-4 in serum; proton pump inhibitors, corticosteroids, antiplatelet drugs and colchicine. Drug treatment proportion was very similar across the different groups. Conclusions CA72-4 false positive cases are sometimes in conjunction with other false positive tumour markers but not in the same proportion as true positives. In addition, most false positive patients can be explained by pharmacological treatments. This research may help in the understanding and identification of potential factors influencing tumor marker levels and thus contribute to the interpretation of results.

Medial posterior tibial slope measurements are overestimated on long radiographs and 3D CT compared to measurements on short lateral radiographs.

This study assessed the measurements of the medial posterior tibial slope (MPTS) using long radiographs and three-dimensional (3D) computed tomography (CT) scans and compared them to measurements taken on short lateral knee radiographs. The study aimed to identify whether the at-risk slope measurements previously defined on the short radiographs would be similar to long radiographs and 3D CT scans. A retrospective radiological review of 52 cases, who underwent planning for a slope-changing high tibial osteotomy and had short and long lateral radiographs and 3D CT scans of the tibia. Two independent observers measured the MPTS on the three modalities. The MPTS was defined as the angle between a tangent to the medial tibial plateau and the referenced tibia anatomical axis. The MPTS measurements from the short and long radiographs were compared to each other and then were compared to the measurements performed on the CT scan. False positives were defined as those cases with MPTS measurements of >78° on CT scans or long radiographs while having measurements ≤78° on short radiographs. These false positive cases are the ones which would be falsely labelled as having an abnormal slope based on the previously validated short radiograph slope threshold ≥12°. A total of 52 cases were analysed (67.9% males and 32.1% females). The mean age was 27 ± 5.4 years. The mean weight was 71.5 ± 7.7 kg, and the mean height was 1.8 ± 0.1 m. The mean MPTS measured on the short radiographs was 77.3 ± 2.3°; on the long radiographs, it was 75.8 ± 2.0°; and on the CT scan, it was 75.3 ± 2.1°. There was a positive correlation between the measurements taken on both the short and long radiographs (r = 0.9) (p < 0.001). Additionally, there was a positive correlation between CT tibial slope measurements and both short and long radiographs tibial slope measurements (r = 0.86, r = 0.87), respectively (both p < 0.001). False positives were 13 (25%) patients on long radiographs, and 12 (23.1%) patients on CT scans, who had their MPTS measurements ≤78° (equivalent of PTS ≥ 12°) while their measurements were >78° on the short radiographs. Measurements of the MPTS can be overestimated by 1.5-2° on long lateral knee radiographs or 3D CT scans compared to measurements taken on short lateral radiographs. Different thresholds for the abnormal PTS measurements on long radiographs and CT scans, should be defined, considering the overestimated measurements in these modalities. Level IV case series.

Assessing the Performance of Models from the 2022 RSNA Cervical Spine Fracture Detection Competition at a Level I Trauma Center.

Purpose To evaluate the performance of the top models from the RSNA 2022 Cervical Spine Fracture Detection challenge on a clinical test dataset of both noncontrast and contrast-enhanced CT scans acquired at a level I trauma center. Materials and Methods Seven top-performing models in the RSNA 2022 Cervical Spine Fracture Detection challenge were retrospectively evaluated on a clinical test set of 1828 CT scans (from 1829 series: 130 positive for fracture, 1699 negative for fracture; 1308 noncontrast, 521 contrast enhanced) from 1779 patients (mean age, 55.8 years ± 22.1 [SD]; 1154 [64.9%] male patients). Scans were acquired without exclusion criteria over 1 year (January-December 2022) from the emergency department of a neurosurgical and level I trauma center. Model performance was assessed using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. False-positive and false-negative cases were further analyzed by a neuroradiologist. Results Although all seven models showed decreased performance on the clinical test set compared with the challenge dataset, the models maintained high performances. On noncontrast CT scans, the models achieved a mean AUC of 0.89 (range: 0.79-0.92), sensitivity of 67.0% (range: 30.9%-80.0%), and specificity of 92.9% (range: 82.1%-99.0%). On contrast-enhanced CT scans, the models had a mean AUC of 0.88 (range: 0.76-0.94), sensitivity of 81.9% (range: 42.7%-100.0%), and specificity of 72.1% (range: 16.4%-92.8%). The models identified 10 fractures missed by radiologists. False-positive cases were more common in contrast-enhanced scans and observed in patients with degenerative changes on noncontrast scans, while false-negative cases were often associated with degenerative changes and osteopenia. Conclusion The winning models from the 2022 RSNA AI Challenge demonstrated a high performance for cervical spine fracture detection on a clinical test dataset, warranting further evaluation for their use as clinical support tools. Keywords: Feature Detection, Supervised Learning, Convolutional Neural Network (CNN), Genetic Algorithms, CT, Spine, Technology Assessment, Head/Neck Supplemental material is available for this article. © RSNA, 2024 See also commentary by Levi and Politi in this issue.

An expert rule-based approach for identifying infantile-onset Pompe disease patients using retrospective electronic health records

Pompe disease (OMIM #232300), a rare genetic disorder, leads to glycogen buildup in the body due to an enzyme deficiency, particularly harming the heart and muscles. Infantile-onset Pompe disease (IOPD) requires urgent treatment to prevent mortality, but the unavailability of these methods often delays diagnosis. Our study aims to streamline IOPD diagnosis in the UAE using electronic health records (EHRs) for faster, more accurate detection and timely treatment initiation. This study utilized electronic health records from the Abu Dhabi Healthcare Company (SEHA) healthcare network in the UAE to develop an expert rule-based screening approach operationalized through a dashboard. The study encompassed six diagnosed IOPD patients and screened 93,365 subjects. Expert rules were formulated to identify potential high-risk IOPD patients based on their age, particular symptoms, and creatine kinase levels. The proposed approach was evaluated using accuracy, sensitivity, and specificity. The proposed approach accurately identified five true positives, one false negative, and four false positive IOPD cases. The false negative case involved a patient with both Pompe disease and congenital heart disease. The focus on CHD led to the overlooking of Pompe disease, exacerbated by no measurement of creatine kinase. The false positive cases were diagnosed with Mitochondrial DNA depletion syndrome 12-A (SLC25A4 gene), Immunodeficiency-71 (ARPC1B mutation), Niemann–Pick disease type C (NPC1 gene mutation leading to frameshift), and Group B Streptococcus meningitis. The proposed approach of integrating expert rules with a dashboard facilitated efficient data visualization and automated patient screening, which aids in the early detection of Pompe disease. Future studies are encouraged to investigate the application of machine learning methodologies to enhance further the precision and efficiency of identifying patients with IOPD.

The onset of coarctation of the aorta before birth: Mechanistic insights from fetal arch anatomy and haemodynamics

Accurate prenatal diagnosis of coarctation of the aorta (CoA) is challenging due to high false positive rate burden and poorly understood aetiology. Despite associations with abnormal blood flow dynamics, fetal arch anatomy changes and alterations in tissue properties, its underlying mechanisms remain a longstanding subject of debate hindering diagnosis in utero. This study leverages computational fluid dynamics (CFD) simulations and statistical shape modelling to investigate the interplay between fetal arch anatomy and blood flow alterations in CoA. Using cardiac magnetic resonance imaging data from 188 fetuses, including normal controls and suspected CoA cases, a statistical shape model of the fetal arch anatomy was built. From this analysis, digital twin models of false and true positive CoA cases were generated. These models were then used to perform CFD simulations of the three-dimensional fetal arch haemodynamics, considering physiological variations in arch shape and blood flow conditions across the disease spectrum. This analysis revealed that independent changes in the shape of.the arch and the balance of left-to-right ventricular output led to qualitatively similar haemodynamic alterations. Transitioning from a false to a true positive phenotype increased retrograde flow through the aortic isthmus. This resulted in the appearance of an area of low wall shear stress surrounded by high wall shear stress values at the flow split apex on the aortic posterior wall opposite the ductal insertion point.Our results suggest a distinctive haemodynamic signature in CoA characterised by the appearance of retrograde flow through the aortic isthmus and altered wall shear stress at its posterior side. The consistent link between alterations in shape and blood flow in CoA suggests the need for comprehensive anatomical and functional diagnostic approaches in CoA. This study presents an application of the digital twin approach to support the understanding of CoA mechanisms in utero and its potential for improved diagnosis before birth.

Non-differential misclassification of outcome under (near)-perfect specificity: a simulation study.

Mismeasurement of a dichotomous outcome yields an unbiased risk ratio estimate when there are no false positive cases (perfect specificity) and when sensitivity is non-differential with respect to exposure status. In studies where these conditions are expected, quantitative bias analysis may be considered unnecessary. We conducted a simulation study to explore the robustness of this special case to small departures from perfect specificity and stochastic departures from non-differential sensitivity. We observed substantial bias of the risk ratio with specificity values as high at 99.8%. The magnitude of bias increased directly with the true underlying risk ratio and was markedly stronger at lower baseline risk. Stochastic departure from non-differential sensitivity also resulted in substantial bias in most simulated scenarios; downward bias prevailed when sensitivity was higher among unexposed compared with exposed, and upward bias prevailed when sensitivity was higher among exposed compared with unexposed. Our results show that seemingly innocuous departures from perfect specificity (e.g., 0.2%) and from non-differential sensitivity can yield substantial bias of the risk ratio under outcome misclassification. We present a web tool permitting easy exploration of this bias mechanism under user-specifiable study scenarios.