Temple University Research Articles

Public health resilience and recovery in Ukraine: Academic partnerships to improve population health

Abstract Background Strengthening the public health workforce is among top priorities for the transformation of Ukraine’s public health system. As part of the USAID Public Health System Recovery and Resilience Activity, an Academic Partnership Collaborative brought together three Ukrainian universities (National University Kyiv-Mohyla Academy, I. Horbachevsky Ternopil National Medical University, Sumy State University), two American universities (Temple University and Drexel University), and the All-Ukrainian Association of Public Health Specialists (UPHA) to enhance the competencies of public health graduates in Ukraine. Objectives The collaborative focuses on strengthening the Ukrainian universities’ public health curricula and the teaching and research competencies of the Ukrainian faculty, making the curricular changes in Ukraine better aligned with US and European academic standards. The university partners are collaborating with the newly established UPHA in building partnerships with peer organizations in the US (American Public Health Association) and Europe (EuPHA) and sharing advancements in public health education with the broader community of public health practitioners in Ukraine to enhance their competencies to improve population health and align with European systems of public health. Results We have established linkages among all partners via an initial meeting in Uzhhorod, Ukraine in April 2024. We anticipate results by November 2024 to include curriculum review recommendations, coaching on prioritized topics for the Ukrainian partners, faculty development plans, and the establishment of continuing professional development programs in collaboration with UPHA. Conclusions Through this partnership, selected Ukrainian universities will have an enhanced and sustainable capacity to train the public health workforce and will become leading examples for other public health programs to simulate and collaborate with beyond the project. Key messages • Ukraine’s resilience and recovery relies upon the strength of the current and future public health workforce; mechanisms for building public health training are essential to a healthy population. • A collaborative process of strategic activities among American and Ukrainian universities and professional associations can be effective in building a resilient public health system in Ukraine.

9.Q. Round table: Strengthening public health law for public health professionals

Abstract Law is a tool that public health experts can use to improve population health. Generally, law offers a range of opportunities: it can encourage healthy and safe behaviours; it can change and shape the physical and social environments; influence the social determinants of health, as well as structure the public health system. Law can also compel or restrict, and legal systems can provide accountability and oversight. ‘Public health law’ stands at the intersection of public health and law, and offers a range of powerful tools for improving health (Burris et al, 2018). While public health law is an important field for public health professionals, there is a lack of knowledge about this field, and a lack of interaction between public health professionals and law professionals. There is also a need to understand the ways law might challenge public health goals. For example, law can be used to promote goals which do not improve public health, or public health concerns can be sidelined, and law can be a tool for challenging specific public health interventions. The focus of this roundtable will be on identifying and building necessary capacities for using law by public health professionals. End-goal is to identify new initiatives to consolidate and to strengthen public health law as an important field for public health professionals. To achieve these aims, we have set three practical steps for the roundtable to embark on: 1) strengthening collaboration; 2) identifying knowledge gaps and opportunities; and 3) starting new initiatives: 1) Strengthening the collaboration and building new partnerships: we bring together experts from law and public health to identify the gaps and opportunities when it comes to implementing law by public health professionals. The roundtable will foster new collaborations and partnerships between law and public health professionals. 2) Identifying gaps and opportunities: there is a need to identify deficiencies in health policy research and how to use scientific legal research methods to overcome these problems. Relatedly, to share experiences and best practices when it comes to teaching public health law. Opportunities include the transdisciplinary model and legal epidemiology; a focus on the commercial determinants; and applying a human rights lens. Panelists - in an interaction with the audience - will identify key opportunities and discuss ways to implement them. 3) Starting new initiatives: panelists will be asked to concretize existing initiatives and discuss with the audience how they can be integrated in the field, based on our new partnerships. New collaborations will be identified. This may vary from building a new network, applying for a joint research grant or for strengthening or complementing existing teaching modules. Altogether, this roundtable aims to strengthen our understanding of existing efforts and of the various ways in which public health expertise can be integrated in better law making. Key messages • Drafting of sound laws that improve health is a joint practice of various professionals, including legal and public health professionals. • We need to strengthen existing capacities and build new ones for using law by public health professionals. Speakers/Panelists Timo Clemens Maastricht University, Maastricht, Netherlands Regien Biesma UMC Utrecht, Utrecht, Netherlands Edward Kirton-Darling University of Bristol, Bristol, UK Scott Burris Temple University, Philadelphia, USA Dominique Mollet Joint Research Council EU, Ispra, Italy

The Power of Story Slams: A Mixed-Method Analysis of Narrative Medicine Connecting, Encouraging and Comforting Healthcare Trainees and Professionals

Background: Narratives convey information and emotion, evoke understanding, empathy, and connection. Healthcare professionals can use narratives to reduce burnout, benefiting them, patients, and colleagues. Story slams are narrative sharing events used for group reflection, providing an opportunity for participants and listeners to improve their well-being by processing complex emotions and identities. Story slams have traditionally been popular with creative writers and nonprofit organizations. Methods: Following a local event hosted by The Moth, the Temple University Lewis Katz School of Medicine (LKSOM) Narrative Medicine program hosted nine story slams over five years with 93 presenters including students and health systems staff. Through a mixed-method analysis, the authors examined the stories for thematic patterns and surveyed presenters to investigate how the experience had impacted markers such as job satisfaction, stress levels, and connection to patients. Eleven presenters were interviewed to provide additional information about the impact they experienced. Results: Patient-centered care, resilience, and the value of learning comprised over half of all themes presented. Following the event, many presenters experienced improved connections with patients, and everyone reported feeling neutral or increased satisfaction with their profession. The interviews conveyed participation, created a lasting impact, fostered a sense of community, and increased appreciation for diversity. Conclusions: Overall, story slams nourish humanity and promote diversity, enthusiasm, and encouragement. This study and others suggest that story slams are inexpensive and effective events that can help reduce burnout and provide academic medical centers with a way to restore and cultivate community among students, physicians, and staff.

The impact of college diversity: Struggles and successes at age 30 By ElizabethAries, Pennsylvania: Temple University Press. 2023. pp. 224. $34.95 (paperback). ISBN: 9781439923184

The impact of college diversity: Struggles and successes at age 30 By ElizabethAries, Pennsylvania: Temple University Press. 2023. pp. 224. $34.95 (paperback). ISBN: 9781439923184

B-108 Delta Check Performance Assessment by Using Real-WorldData at the Temple University Health System​

Abstract Background Delta check is a post-analytic tool focusing on pre-analytical, analytical and post-analytical errors. It can be sensitive in detecting isolated errors affecting a single patient, as opposed to moving patient averages, which is another QC tool for detecting errors that affect a cohort/population of patients. As a computer-based system on selected tests, delta checks are to detect excessive changes from prior value in the patient test result and to detect possible specimen errors such as misidentified or compromised samples before results are reported. It may also play a role in detecting examination(analytical) issues and clinically significant results. If delta checks exceed established limits, the result is flagged. There is a common agreement that measurands with low indices of individuality (<0.6) make good candidates for the identification of cases of sample misidentification than higher indices of individuality (>1.4). CLSI EP33 Use of Delta Checks in the Medical Laboratory guidelines recommends that laboratories assess the validity of current delta check parameters by reviewing retrospective data on flagged results to determine if errors are being identified efficiently. At Temple University Health System, delta check limits were established historically by using published data, however, its performance characteristics have not been verified previously. Verification of clinical performance of predefined delta check limits can be performed either through mathematical calculations or via clinical audits. Clinical audits are achieved by documenting the follow-up actions and outcomes of the samples flagged by a delta check rule(s). It has been suggested that this should be performed using a protocolized follow-up procedure to ensure uniformity in the investigation and management. Corey M. et al. shared an excel template, a mathematical model that can be adopted easily by replacing the example data with retrospectively extracted data and modifying the parameters to suit their laboratory’s requirements. Methods We collected up to 30,000 data points for each 12 measurand in the clinical chemistry lab retrospectively. By using the excel template, clinical sensitivity and clinical specificity has been calculated. Furthermore, we created ROC curves and evaluated them using the (YI) Youden Index and AUC (Area Under the Curve). Results In general, YI values were found to be very low, less than 0.6. Highest YI was used to establish new delta check limits for each analyte; only the current creatinine delta check limit was found as effective, and no change required. 11 out of 12 measurand were required a revision of delta check limits and were proposed to replace with lower delta checks. AUC values were found between 0.56-0.794. Na showed the highest AUC, which correlates its low individuality index (0.5) and its effects on effectiveness of the delta check system. Conclusions It is of vital importance to establish a solid follow-up evaluation for delta check alerts and distinguish between integrity issues and clinically significant changes. Conclusively, this endeavor can develop into extensive information for the application, evaluation, and follow-up of delta checks whilst improving diagnostic accuracy and quality assurance.

Divide & Conquer: Critical Informatics Approaches to Disinformation

ABSTRACTIn this paper, we use a critical informatics approach to investigate institutional disinformation around 2022–23 labor organizing at three higher education institutions: Rutgers University, Temple University, and the University of California. Our contribution to the study of disinformation is the application of critical informatics perspectives that attend to structural power dynamics of disinformation within an institutional context. Understanding the political economic dynamics of disinformation and how these dynamics function can help more solidly contextualize and clarify how and why disinformation exists across different information systems, so that solutions to this social and institutional problem of disinformation can be more appropriately addressed and understood. The study describes disinformation tactics employed by institutional leaders during higher education labor organizing including: non‐performative commitments to “community”, legal threats, misleading victories, and elite capture.

Newborn Hearing Screening Success… Patient Care Failure.

To assess the efficacy of a newborn hearing screening program in an urban hospital 24 years after a longitudinal study identified failings in its program and recommended solutions. A hospital-specific, Pennsylvania database was queried to determine rates of in-hospital testing, referral, and diagnosis for Temple University Hospital (TUH), and to compare these with reporting sites statewide. At TUH, interviews with newborn hearing screeners and audiologists were conducted to identify barriers to care. In the year 2022, 2,006/2,069 newborns (97%) were successfully screened prior to hospital discharge at TUH. (63 babies were never tested). There were 73 who did not pass automated auditory brainstem response testing (ABR) in at least one ear. All of these 73 infants returned for repeat automated ABR testing and 50/73 passed in both ears. 23/2006 (1%) were referred to diagnostic testing. 13/23 were successfully scheduled. 11/23 passed in both ears. 2/13 were found to have hearing losses. 10 infants (43%) were lost to follow-up and were never diagnosed. Examining lost-to-follow-up rates statewide, we identified a relationship between successful follow-up and local median family income. Legislation on the state level and improvements in audiometric technology have led to highly successful in-hospital screening with relatively low false-positive rates. Despite this, inadequate outpatient follow-up testing and inequities in the American health care system continue to delay audiometric diagnosis and habilitation of congenital hearing loss. Lack of transportation, childcare for siblings, and newborn insurance status impede access to diagnostic testing. Level 3 evidence-retrospective review comparing cases and controls. Laryngoscope, 2024.

Political Black Girl Magic: The Elections and Governance of Black Female Mayors. Edited by Sharon D. Wright Austin. Philadelphia: Temple University Press, 2023. 332p. $115.50 cloth, $39.95 paper.

Political Black Girl Magic: The Elections and Governance of Black Female Mayors. Edited by Sharon D. Wright Austin. Philadelphia: Temple University Press, 2023. 332p. $115.50 cloth, $39.95 paper.

Comparison of Machine Learning Models to a Novel Score in the Identification of Patients at Low Risk for Diabetic Retinopathy

PurposeTo develop an easily applicable predictor of patients at low risk for diabetic retinopathy (DR). DesignAn experimental study on the development and validation of machine learning models and a novel retinopathy risk score to detect patients at low risk for diabetic retinopathy. SubjectsAll individuals aged 18 years or older who participated in the telemedicine retinal screening initiative through Temple University Health Systems from October 1, 2016 through December 31, 2020. The subjects must have documented evidence of their diabetes mellitus diagnosis as well as a documented hemoglobin A1C recorded in their chart within 6 months of the retinal screening photograph. MethodsThe charts of 1930 subjects (1590 evaluable) undergoing telemedicine screening for DR were reviewed, and 30 demographic and clinical parameters were collected. DR is a dichotomous variable where low risk is defined as no or mild retinopathy using the International Clinical Diabetic Retinopathy severity score. Five machine learning models (MLM) were trained to predict patients at low risk for DR using 1050 subjects and further underwent 10-fold cross validation to maximize its performance indicated by the area under the receiver operator characteristic curve (AUC). Additionally, a novel retinopathy risk score (RRS) is defined as the product of hemoglobin A1c closest to screening and years with diabetes mellitus. RRS was also applied to generate a predictive model. Main Outcome MeasuresThe performance of the trained machine learning models and the RRS model was compared using DeLong’s test. The models were further validated using a separate unseen test set of 540 subjects.The performance of the validation models were compared using DeLong’s test and chi-square tests. ResultsUsing the test set, the AUC for the RRS was not statistically different from four out of five MLM. The error rate for predicting low-risk patients using the RRS was significantly lower than the naive rate (0.097 vs 0.19; p<0.0001), and it was comparable to the error rates of the machine learning models. ConclusionThis novel retinopathy risk score is a potentially useful and easily deployable predictor of patients at low risk for DR.

Comparative study of cemento-ossifying fibroma and focal cemento-osseous dysplasia in the jaw – 15 year retrospective inquiry and literature review

BackgroundCemento-ossifying fibroma (COF) and focal cemento-osseous dysplasia (FCOD) are benign neoplasms occurring in the mandible and maxilla; histologically, they both show normal bone replacement by a fibrous stroma containing variable amounts of calcified fragments resembling bone and cementum. FCOD is relatively common and requires observation; COF is rare and its growth over time can lead to deformity in the head and neck region. Surgical removal of COF at an early stage is critical to prevent recurrence and deformity. Therefore, distinguishing COF from FCOD is important for proper further management. Methods and MaterialsTo determine the parameters differentiating COF from FCOD, we retrospectively analyzed 15 COF cases and 27 FCOD cases reviewed between 2005 and 2020 at Temple University Hospital. Demographic, radiographic and histopathologic characteristics were compared by using Student’s t test for continuous variables and Chi-square test for categorical variables. ResultsIn the non-histopathologic spectrum, COF occurred at a younger age (average age-31yo versus FCOD 43 yo). FCOD showed female predilection compared with COF. No significant difference was observed regarding location and radiographic density. However, COF was significantly associated with radiographic cortical expansion. In the histopathologic spectrum, FCOD was more frequently associated with significant intralesional hemorrhage, increased vascularity, osteoblastic rimming of trabeculae, and ginger roots –like bone trabeculae. DiscussionWe demonstrated 7 clinical, radiographic, and pathologic parameters that can help differentiate COF from FCOD. Comprehensive assessment of the distinguishing characteristics is valuable for accurate diagnosis and treatment of these two entities, aiding in minimizing unnecessary procedures for FCOD and averting facial deformities caused by COF.

Brain age revisited: Investigating the state vs. trait hypotheses of EEG-derived brain-age dynamics with deep learning

Abstract The brain’s biological age has been considered as a promising candidate for a neurologically significant biomarker. However, recent results based on longitudinal magnetic resonance imaging (MRI) data have raised questions on its interpretation. A central question is whether an increased biological age of the brain is indicative of brain pathology and if changes in brain age correlate with diagnosed pathology (state hypothesis). Alternatively, could the discrepancy in brain age be a stable characteristic unique to each individual (trait hypothesis)? To address this question, we present a comprehensive study on brain aging based on clinical Electroencephalography (EEG), which is complementary to previous MRI-based investigations. We apply a state-of-the-art temporal convolutional network (TCN) to the task of age regression. We train on recordings of the Temple University Hospital EEG Corpus (TUEG) explicitly labeled as non-pathological and evaluate on recordings of subjects with non-pathological as well as pathological recordings, both with examinations at a single point in time TUH Abnormal EEG Corpus (TUAB) and repeated examinations over time. Therefore, we created four novel subsets of TUEG that include subjects with multiple recordings: repeated non-pathological (RNP): all labeled non-pathological; repeated pathological (RP): all labeled pathological; transition non-patholoigical pathological (TNPP): at least one recording labeled non-pathological followed by at least one recording labeled pathological; and transition pathological non-pathological (TPNP): similar to TNPP but with opposing transition (first pathological and then non-pathological). The results show that our TCN reaches state-of-the-art performance in age decoding on non-pathological subjects of TUAB with a mean absolute error of 6.6 years and an R2 score of 0.73. Our extensive analyses demonstrate that the model underestimates the age of non-pathological and pathological subjects, the latter significantly (-1 and -5 years, paired t-test, p = 0.18 and p = 6.6e−3). Furthermore, there exist significant differences in average brain age gap between non-pathological and pathological subjects both with single examinations (TUAB) and repeated examinations (RNP vs. RP) (-4 and -7.48 years, permutation test, p = 1.63e−2 and p = 1e−5). We find mixed results regarding the significance of pathology classification based on the brain age gap biomarker. While it is indicative of pathological EEG in datasets TUAB and RNP versus RP (61.12% and 60.80% BACC, permutation test, p = 1.32e−3 and p = 1e−5), it is not indicative in TNPP and TPNP (44.74% and 47.79% BACC, permutation test, p = 0.086 and p = 0.483). Additionally, all of these classification scores are clearly inferior to the ones obtained from direct EEG pathology classification at 86% BACC and higher. Furthermore, we could not find evidence that a change of EEG pathology status within subjects relates to a significant change in brain age gap in datasets TNPP and TPNP (0.46 and 1.35 years, permutation test, p = 0.825 and p = 0.43; and Wilcoxon-Mann-Whitney and Brunner-Munzel test, p = 0.13). Our findings, thus, support the trait rather than the state hypothesis for brain age estimates derived from EEG. In summary, our findings indicate that the neural underpinnings of brain age changes are likely more multifaceted than previously thought, and that taking this into account will benefit the interpretation of empirically observed brain age dynamics.

Spectral changes in electroencephalography linked to neuroactive medications: A computational pipeline for data mining and analysis

Background and objectivesThe increasing amount of open-access medical data provides new opportunities to gain clinically relevant information without recruiting new patients. We developed an open-source computational pipeline, that utilizes the publicly available electroencephalographic (EEG) data of the Temple University Hospital to identify EEG profiles associated with the usage of neuroactive medications. It facilitates access to the data and ensures consistency in data processing and analysis, thus reducing the risk of errors and creating comparable and reproducible results. Using this pipeline, we analyze the influence of common neuroactive medications on brain activity. MethodsThe pipeline is constructed using easily controlled modules. The user defines the medications of interest and comparison groups. The data is downloaded and preprocessed, spectral features are extracted, and statistical group comparison with visualization through a topographic EEG map is performed. The pipeline is adjustable to answer a variety of research questions. Here, the effects of carbamazepine and risperidone were statistically compared with control data and with other medications from the same classes (anticonvulsants and antipsychotics). ResultsThe comparison between carbamazepine and the control group showed an increase in absolute and relative power for delta and theta, and a decrease in relative power for alpha, beta, and gamma. Compared to antiseizure medications, carbamazepine showed an increase in alpha and theta for absolute powers, and for relative powers an increase in alpha and theta, and a decrease in gamma and delta. Risperidone compared with the control group showed a decrease in absolute and relative power for alpha and beta and an increase in theta for relative power. Compared to antipsychotic medications, risperidone showed a decrease in delta for absolute powers. These results show good agreement with state-of-the-art research. The database allows to create large groups for many different medications. Additionally, it provides a collection of records labeled as “normal” after expert assessment, which is convenient for the creation of control groups. ConclusionsThe pipeline allows fast testing of different hypotheses regarding links between medications and EEG spectrum through ecological usage of readily available data. It can be utilized to make informed decisions about the design of new clinical studies.

AI in ECG: Validating an ambulatory semiology labeller and predictor

ObjectivesEarly prediction of epileptic seizures can help reduce morbidity and mortality. In this work, we explore using electrocardiographic (ECG) signal as input to a seizure prediction system and note that the performance can be improved by using selected signal processing techniques. MethodsWe used frequency domain analysis with a deep neural network backend for all our experiments in this work. We further analysed the effect of the proposed system for different seizure semiologies and prediction horizons. We explored refining the signal using signal processing to enhance the system's performance. ResultsOur final system using the Temple University Hospital’s Seizure (TUHSZ) corpus gave an overall prediction accuracy of 84.02 %, sensitivity of 87.59 %, specificity of 81.9 %, and an area under the receiver operating characteristic curve (AUROC) of 0.9112. Notably, these results surpassed the state-of-the-art outcomes reported using the TUHSZ database; all findings are statistically significant. We also validated our study using the Siena scalp EEG database. Using the frequency domain data, our baseline system gave a performance of 75.17 %, 79.17 %, 70.04 % and 0.82 for prediction accuracy, sensitivity, specificity and AUROC, respectively. After selecting the optimal frequency band of 0.8–15 Hz, we obtained a performance of 80.49 %, 89.51 %, 75.23 % and 0.89 for prediction accuracy, sensitivity, specificity and AUROC, respectively which is an improvement of 5.32 %, 10.34 %, 5.19 % and 0.08 for prediction accuracy, sensitivity, specificity and AUROC, respectively. ConclusionsThe seizure information in ECG is concentrated in a narrow frequency band. Identifying and selecting that band can help improve the performance of seizure detection and prediction. SignificanceEEG is susceptible to artefacts and is not preferred in a low-cost ambulatory device. ECG can be used in wearable devices (like chest bands) and is feasible for developing a low-cost ambulatory device for seizure prediction. Early seizure prediction can provide patients and clinicians with the required alert to take necessary precautions and prevent a fatality, significantly improving the patient’s quality of life.

Central Giant Cell Granuloma of the Mandible and Maxilla: A Clinicopathological Study of 21 Cases.

Background Central giant cell granuloma (CGCG) presents as a locally invasive, intraosseous lesion characterized by the accumulation of multinucleated giant cells amidst a matrix of hemorrhage and reactive fibrous tissue that infiltrates bone trabeculae. This idiopathic non-neoplastic proliferative lesion primarily affects the mandible, typically presenting as either unilocular or multilocular radiolucencies on X-rays. Although trauma or intraosseous hemorrhages are potential triggers, the precise histogenesis and etiology remain unclear. CGCG predominantly occurs in children and young adults, with a slight female predilection. Methods and materials A retrospective analysis of 21 cases of CGCG diagnosed at the Oral Pathology/Pathology department of Temple University Hospital between 2015 and 2022 was conducted. Each case was evaluated based on various parameters, including age, gender, presenting symptoms, radiographic findings, clinical differential diagnosis, and histological confirmation. The primary radiographic technique employed for diagnosis was X-ray imaging of the mandible and maxilla. The histological examination involved cutting paraffin-embedded tissue into 5-micrometer-thick sections, which were then stained using routine hematoxylin and eosin (H&E) stain. Notably, no specialized histochemical or immunohistochemical stains were utilized in the evaluation process. Results In our study, we reviewed 21 cases; 9 were male, 11 were female, and one had no available gender data.The age range was 15-76 years, with a mean of 50 years. The mandible was the most commonly affected location (17 cases; 81%)while the maxilla was less commonly involved (4 cases; 19%). Many CGCG lesions were asymptomatic (13 cases; 62%); eight cases (38%) were symptomatic, with painand fullness of the affected dental region being the main manifestations. In a few cases, conditions such as brown tumor (severe hyperparathyroidism) and odontogenic neoplasms, such as ameloblastoma, were suspected clinically and radiographically. The diagnosis of CGCG with associated acute and chronic inflammation was confirmed in all the cases. Histological evaluation of routinely stained slides was the main diagnostic tool utilized. No special stains or molecular studies were required to establish the final diagnosis. Conclusions Our investigation has determined that CGCG exhibits a non-neoplastic nature, displaying a spectrum of behaviors ranging from non-aggressive to aggressive tendencies. While CGCG is predominantly observed in the mandible, rare instances of involvement in the maxilla have also been documented. Importantly, no confirmed association with neoplastic lesions was identified during our analysis. The clinical course of CGCG tends to be indolent, with some cases presenting in association with impacted teeth. It's noteworthy that CGCG can present features mimicking neoplastic conditions, such as ameloblastoma, or localized lesions linked to systemic disorderssuch as hyperparathyroidism (brown tumor).

Residual and bidirectional LSTM for epileptic seizure detection.

Electroencephalogram (EEG) plays a pivotal role in the detection and analysis of epileptic seizures, which affects over 70 million people in the world. Nonetheless, the visual interpretation of EEG signals for epilepsy detection is laborious and time-consuming. To tackle this open challenge, we introduce a straightforward yet efficient hybrid deep learning approach, named ResBiLSTM, for detecting epileptic seizures using EEG signals. Firstly, a one-dimensional residual neural network (ResNet) is tailored to adeptly extract the local spatial features of EEG signals. Subsequently, the acquired features are input into a bidirectional long short-term memory (BiLSTM) layer to model temporal dependencies. These output features are further processed through two fully connected layers to achieve the final epileptic seizure detection. The performance of ResBiLSTM is assessed on the epileptic seizure datasets provided by the University of Bonn and Temple University Hospital (TUH). The ResBiLSTM model achieves epileptic seizure detection accuracy rates of 98.88-100% in binary and ternary classifications on the Bonn dataset. Experimental outcomes for seizure recognition across seven epilepsy seizure types on the TUH seizure corpus (TUSZ) dataset indicate that the ResBiLSTM model attains a classification accuracy of 95.03% and a weighted F1 score of 95.03% with 10-fold cross-validation. These findings illustrate that ResBiLSTM outperforms several recent deep learning state-of-the-art approaches.

Public schools, private governance: Education reform and democracy in New Orleans By J. CelesteLay, Philadelphia, PA: Temple University Press. 2022. 230 pp. $32.95 (paperback). EAN 9781439922644

Public schools, private governance: Education reform and democracy in New Orleans By J. CelesteLay, Philadelphia, PA: Temple University Press. 2022. 230 pp. $32.95 (paperback). EAN 9781439922644

Reaching the ceiling? Empirical scaling behaviour for deep EEG pathology classification

Machine learning techniques, particularly deep convolutional neural networks (ConvNets), are increasingly being used to automate clinical EEG analysis, with the potential to reduce the clinical burden and improve patient care. However, further research is required before they can be used in clinical settings, particularly regarding the impact of the number of training samples and model parameters on their testing error. To address this, we present a comprehensive study of the empirical scaling behaviour of ConvNets for EEG pathology classification. We analysed the testing error with increasing the training samples and model size for four different ConvNet architectures. The focus of our experiments is width scaling, and we have increased the number of parameters to up to 1.8 million. Our evaluation was based on two publicly available datasets: the Temple University Hospital (TUH) Abnormal EEG Corpus and the TUH Abnormal Expansion Balanced EEG Corpus, which together contain 10,707 training samples. The results show that the testing error follows a saturating power-law with both model and dataset size. This pattern is consistent across different datasets and ConvNet architectures. Furthermore, empirically observed accuracies saturate at 85%–87%, which may be due to an imperfect inter-rater agreement on the clinical labels. The empirical scaling behaviour of the test performance with dataset and model size has significant implications for deep EEG pathology classification research and practice. Our findings highlight the potential of deep ConvNets for high-performance EEG pathology classification, and the identified scaling relationships provide valuable recommendations for the advancement of automated EEG diagnostics.

FIT as a population-based and financially viable method to screen underserved patients for colorectal cancer.

e13516 Background: Temple University Hospital (TUH) created an initiative to increase colorectal cancer (CRC) screening by administering fecal immunochemical tests (FIT) to the underserved population of Philadelphia. We hypothesize that this will increase the percentage of diagnostic colonoscopies performed compared to baseline in our fixed capacity colonoscopy suite and ultimately generate a higher reimbursement per colonoscopy. This potential revenue increase can help fund the infrastructure to administer a large quantity of FIT tests to the community and create a truly equitable population-based approach to CRC screening. Methods: 27,169 colonoscopies and their associated CPT codes from 2017-2022 performed at TUH were analyzed. 16 individual CPT codes were identified and categorized as either screening or diagnostic CPT codes. Reimbursement for each of the CPT codes was determined by the Medicare reimbursement per CPT code and used to determine the average reimbursement per screening CPT code and average reimbursement per diagnostic CPT code. A Monte Carolo simulation model was created using Palisades @Risk software to compare the reimbursements of FIT test prompted vs. non-FIT test prompted colonoscopies over a range of possible distributions of diagnostic CPT codes. Monte Carlo simulations model scenarios that incorporate uncertainties with the objective of providing a distribution of outcomes from the least likely to most likely outcomes. Results: The historical distribution of screening vs. diagnostic CPT codes at TUH is 31% and 69%, respectively, with an average Medicare reimbursement of $2,668 per colonoscopy. We have distributed 292 FIT tests, and 5 FIT test prompted colonoscopies have been completed with 7 associated CPT codes. The preliminary distribution of screening CPT codes vs diagnostic CPT codes is 29% and 71%, respectively, with an average reimbursement of $2,691. Lastly, we used a Monte Carlo simulation to run 10 separate simulations with 5000 iterations each to show the most likely reimbursement rates over a possible distribution of diagnostic colonoscopies (Table). Conclusions: Our preliminary data suggests FIT testing can offer three main benefits: 1) Expansion of the number of patients receiving CRC screening; 2) triage optimization of the fixed capacity colonoscopy suite by performing more diagnostic colonoscopies than screening colonoscopies to diagnose more colorectal pathologies; and 3) funding for a population based FIT testing campaign by utilizing the Medicare reimbursement delta from screening to diagnostic colonoscopies. [Table: see text]

Optimizing colorectal cancer screening outreach: A probability-based strategy by Temple University Hospital.

e13502 Background: Temple University Hospital has created an initiative to increase colorectal cancer (CRC) screening to a historically underserved population in Northern Philadelphia through the administration of Fecal Immunochemical Tests (FIT). FIT testing possesses the advantage of being inexpensive and easy to use. Moreover, FIT testing serves as an additional opportunity for underserved patients to be integrated into a health network to receive lifesaving care when utilized by a large health system such as Temple Health. This study demonstrates how healthcare institutions, such as Temple Health, can employ probability-based modeling created from empiric data to navigate uncertainty and establish goals for the number of colonoscopies they aim to perform per a specific quantity of FIT tests distributed within the community. By doing so, they can significantly enhance the overall screening efforts in Northern Philadelphia and provide an essential service to those who need it most. Methods: FIT tests were distributed to the participants at community engagement events. FIT response rates, positivity rates, and progression to colonoscopy rates were determined and fit to a binomial distribution in Palisades @Risk software and underwent a Monte Carlo simulation of 10,000 iterations to determine the different probabilities of response rates, positivity rate, and colonoscopies performed per 1000 FIT tests distributed. Results: A total of 292 tests were distributed with a response rate of 56.5% (165 patients) and a positivity rate of 8.9% (26 patients). Out of 292 patients, 2.4% (7 patients) have had a colonoscopy or are scheduled to have a colonoscopy at Temple University Hospital. Probability curves were created to demonstrate the probability of number of responses, number of positive tests, and number of colonoscopies performed per 1000 FIT tests administered to the community. Lastly, we modeled the number of FIT tests that need to be administered to the community to perform at least 100 colonoscopies with different levels of certainty (Table). Conclusions: Here we show how hospital systems, like Temple Health, can use probability distributions based on empiric data to set tangible goals regarding their community engagement efforts to increase access to CRC screening. Through the utilization of probability distributions, healthcare organizations can gain deeper insights into resource allocation needs and strategically direct their initiatives, ultimately fostering a more inclusive and equitable approach to population-based CRC screening. [Table: see text]

Exploring the potential of pretrained CNNs and time-frequency methods for accurate epileptic EEG classification: a comparative study

Prompt diagnosis of epilepsy relies on accurate classification of automated electroencephalogram (EEG) signals. Several approaches have been developed to characterize epileptic EEG data; however, none of them have exploited time-frequency data to evaluate the effect of tweaking parameters in pretrained frameworks for EEG data classification. This study compares the performance of several pretrained convolutional neural networks (CNNs) namely, AlexNet, GoogLeNet, MobileNetV2, ResNet-18 and SqueezeNet for the localization of epilepsy EEG data using various time-frequency data representation algorithms. Continuous wavelet transform (CWT), empirical Fourier decomposition (EFD), empirical mode decomposition (EMD), empirical wavelet transform (EWT), and variational mode decomposition (VMD) were exploited for the acquisition of 2D scalograms from 1D data. The research evaluates the effect of multiple factors, including noisy versus denoised scalograms, different optimizers, learning rates, single versus dual channels, model size, and computational time consumption. The benchmark Bern-Barcelona EEG dataset is used for testing purpose. Results obtained show that the combination of MobileNetV2, Continuous Wavelet Transform (CWT) and Adam optimizer at a learning rate of 10−4, coupled with dual-data channels, provides the best performance metrics. Specifically, these parameters result in optimal sensitivity, specificity, f1-score, and classification accuracy, with respective values of 96.06%, 96.15%, 96.08%, and 96.10%. To further corroborate the efficacy of opted pretrained models on exploited Signal Decomposition (SD) algorithms, the classifiers are also being simulated on Temple University database at pinnacle modeling composition. A similar pattern in the outcome readily validate the findings of our study and robustness of deep learning models on epilepsy EEG scalograms.The conclusions drawn emphasize the potential of pretrained CNN-based models to create a robust, automated system for diagnosing epileptiform. Furthermore, the study offers insights into the effectiveness of varying time-frequency techniques and classifier parameters for classifying epileptic EEG data.