NUR Graduate Student Conference Papers

Permanent URI for this collection

https://hdl.handle.net/10057/539

Browse

Recent Submissions

Now showing 1 - 5 of 32
  • Thumbnail Image
    Item
    Obesity classification from facial images using deep learning
    (Wichita State University, 2021-04-02) Siddiqui, Hera; Siddiqui, Hera; Rattani, Ajita; Cure Vellojin, Laila N.; Woods, Nikki Keene; Lewis, Rhonda K.; Twomey, Janet M.; Smith-Campbell, Betty; Hill, Twyla J.
    INTRODUCTION: Obesity is a serious health problem that is on the rise both in the United States and globally. Obesity is frequently defined using the clinical Body Mass Index (BMI) ratio of height and weight. Overweight individuals have a BMI between 25-30, and those over 30 are classified as obese. Obesity can lead to heart disease, type 2 diabetes, and many other serious health conditions. Self-diagnostic face-based solutions are being investigated for obesity classification and monitoring. PURPOSE: To classify obesity status based on facial images using deep learning-based convolutional neural networks (CNNs). METHODS: The four CNNs (VGG16, ResNet50, DenseNet121, and MobileNetV2) used in this study were pre-trained on three public datasets (ImageNet, VGGFace, and VGGFace2). Using the above CNNs, we extracted deep features from the FIW-BMI and VisualBMI datasets annotated with BMI information. The deep features from 8298 images in the FIW-BMI dataset along with BMI values were then used to train a Support Vector Classification (SVC) classifier. The trained SVC model was tested on 4206 different images from the VisualBMI dataset for the validation. RESULTS: CNNs trained on ImageNet dataset obtained an initial accuracy (percentage of correct obese and non-obese classifications) in the range 64% to 72%. Accuracy of 84% to 86% was obtained by using CNNs trained on VGGFace dataset. 86% accuracy was obtained by concatenating features from pre-trained (VGGFace) and fine-tuned (FIW-BMI) model. ResNet-50 trained on VGGFace2 dataset obtained an accuracy of 91% when features from the original image datasets were used and 92% accuracy when features were fused from the original image with the horizontally flipped image. The fused image modifications resulted in a model with Sensitivity, Specificity, and Precision of 0.90, 0.94, and 0.95, respectively. Mean Absolute Error (MAE) of this model in predicting BMI is 3.16 and area under the curve (AUC) is 0.97. CONCLUSION: Obesity can be predicted from facial images using deep learning models with a promising accuracy. SVC models trained on deep features extracted from models pre-trained on VGGFace2 dataset performed better than models pre-trained on ImageNet dataset. ResNet-50 (pre-trained on VGGFace2) obtained the highest accuracy of 92% by combining features from the original image and horizontally flipped image. These models when deployed on smartphones can help individuals in monitoring their obesity status, BMI, and weight changes.
  • Thumbnail Image
    Item
    Heart failure transitional care services: An APRN led clinic
    (Wichita State University, 2021-04-02) Swain-Abraham, Kathryn; McGarvey, Jeremy; Smith-Campbell, Betty; Habtemariam, Maryon
    INTRODUCTION: Heart Failure (HF) is a debilitating disease, that doesn't let the heart muscle pump enough blood. HF effects 6.5 million, with costs over $30 billion annually in the US (Yancy, 2013) The coronavirus (Covid-19) pandemic re-enforced the importance of using telehealth to deliver care. Advanced Practice Registered Nurse (APRN) clinics have been found to reduce hospital admissions and costs (Rice, 2018). PURPOSE: This project compared the impact of an APRN-run and Medical doctor (MD) run clinic on hospital admission rates, Emergency Department (ED) visits and clinic costs at a metropolitan mid-western hospital. METHOD: This Quality Improvement project (a systematic approach to data collection) collected retrospective aggregated data through an electronic report. Data included HF patients age (18 yrs. and older), provider type (APRN or MD), insurance status, zip code, gender, race/ethnicity and diagnostic ICD-10 codes for 2019 and post implementation of telemedicine due to Covid-19 in 2020. RESULTS: Hospital admission and ED rates were not significantly different between the APRN and MD clinic in 2019. The findings were similar following implementation of telemedicine due to COVID19 in 2020. Median total costs (the middle cost) and direct costs (expenses) were found to be significantly higher, and direct contribution margin (how a product contributes to the overall profit of the company) significantly lower for the APRN clinic compared to the MD clinic. This changed in 2020, using telemedicine there was no longer a significant difference in direct contribution margin CONCLUSION: APRN and MD clinic HF patients had similar outcomes in hospital readmission rates and ED visits at one metropolitan hospital. The outcomes could be due to the APRN and MD providing similar patient care. Costs were expected to be lower for the APRN clinic because APRNs provide cost-effective outcomes through less expensive labor input that produce healthcare services with the desired clinical result (Rice, 2018). In this study the APRN clinic costs could be because the overhead costs for a hospital outpatient clinic (HOD) may be more than an outpatient clinic (OPC). This cost was negated with telemedicine. There is a need to assess additional HF patient outcomes other than just hospital admission rates (Yancy, 2013).
  • Thumbnail Image
    Item
    Obstructive sleep apnea: Using the STOP-Bang questionnaire as a tool to identify those at risk in the primary care setting
    (Wichita State University, 2021-04-02) Kramer, Gary P.; Sebes, Jennifer; Harrington, Jamie
    INTRODUCTION: Obstructive Sleep Apnea (OSA) is a common sleep disorder in which an individual experiences periods of either apnea (stopped breathing) or hypopnea (long shallow breathing) during sleep (Chakravorty, T., Konar, & Chakravorty, I., 2019). The American Sleep Apnea Association (2020) estimates that 22 million Americans suffer from obstructive sleep apnea, with 80% of cases undiagnosed. OSA is a treatable condition and, if managed properly, can help reduce complications of untreated OSA. OSA contributes towards psychosocial conditions such as anxiety, mood disorders, cognitive impairment, and life-long physical conditions such as hypertension, cardiovascular disease, diabetes and cancer. The STOP-Bang screening questionnaire is a proven screening tool that can accurately PURPOSE: To identify those at risk for obstructive sleep apnea using the STOP-Bang questionnaire in a primary care setting. METHODS: A nine-week retrospective chart review was performed to gather baseline data of a family practice clinic's screening methods for sleep apnea. Then, for a nine-week period, patients who met screening criteria were asked to fill out the STOP-Bang questionnaire. Based on the score from the questionnaires, patients were asked to be referred for additional screening. RESULTS: the retrospective chart review found 83 patients that met screening criteria and out of those 83 patients only three were screened and all three were sent for additional testing. During the project, 88 patients met screening criteria and 17 were screened. Out of those 17 patients, only three were referred for additional testing. CONCLUSION: Using the STOP-Bang screening questionnaire helped identify more people at risk for OSA than the previous screening methods at a family practice clinic. Although, this project did not show that using the screening tool was beneficial in referring at-risk patients on for further testing.
  • Thumbnail Image
    Item
    Hepatitis C screening rates in primary care
    (Wichita State University, 2020-05-01) Russell, Lisa; Sebes, Jennifer; Leighton, Maggie
    INTRODUCTION: Chronic hepatitis C virus affects more than three million Americans. Up to 75% of those patients remain unaware of their infected status, as hepatitis C is often asymptomatic. Hepatitis C-related complications include end stage liver disease, cirrhosis, and hepatocellular carcinoma. The majority of those living with hepatitis C were born between 1945 and 1965, known as the baby boomer cohort. The CDC recommends testing for hepatitis C for those in the baby boomer cohort, and those who currently or have ever injected drugs. This recommendation for screening despite possible absence of patient-reported risk factors poses a barrier in primary care settings, in providers considering screening, and potentially in patients accepting screening. Because this infection typically presents without symptoms, risk assessments and hepatitis C antibody testing is needed in primary care to identify asymptomatic, chronically infected patients. PURPOSE: The purpose of this project was to evaluate hepatitis C screening rates for chronic hepatitis C in a primary care setting, to assess provider screening practices, and to determine provider barriers to linking patients with chronic hepatitis C to care. METHODS: A quantitative research design was used to measure hepatitis C screening rates three years after implantation of an in-house treatment program for hepatitis C, from January 2016 to August 2019, at a Federally Qualified Health Center in an urban Midwest location. Screening rates were pulled from the organization's electronic medical record for patients in the baby boomer cohort and those who have injected drugs. A brief survey was sent to medical providers to assess their hepatitis C screening practices and any barriers they may perceive to screening and linkage to care. RESULTS: One hundred eight-seven patients from 2016-2019 were identified through codes in the medical record indicating age (born between 1945 and 1965) and/or history of injection drug use. The screening rate for both groups was 92%. This screening rate was found to be statistically significant from the national screening rate. Multiple risk factors and barriers to screening were identified by medical providers. CONCLUSION: Screening rates for hepatitis C in this primary care setting were significantly different from the national average. Medical providers surveyed indicated that they were likely to screen for hepatitis C in injection drug users, those with hepatitis B, patients with HIV, people exposed to needle stick injuries, and patients with elevated liver enzymes. Providers found that cost, substance abuse, lack of transportation and difficulties with communication impede providers' abilities to link patients to care. The results of the project bring about new information regarding factors influencing screening and supports the role of advanced practice nurses in treating hepatitis C.
  • Thumbnail Image
    Item
    Patient perceptions of skin lesion assessment and education in primary care setting
    (Wichita State University, 2020-05-01) Hull, Bridgette; Goebel-Roberts, Pamela; Huckstadt, Alicia A.
    INTRODUCTION: As a primary healthcare provider, providing education to patients plays an important role in preventative healthcare. The discussion of skin health should be included with other routine education and preventative screenings, such as mammograms, and held to the same level of significance. With appropriate education, such as how to identify skin lesion changes, patients may be able to identify skin changes and communicate these concerns with their provider. Unfortunately, not having this education as a patient may result in a possible late or missed diagnosis of skin cancer, specifically melanoma, which can have a devastating prognosis. PURPOSE: The purpose of this project was to evaluate if patients felt comfortable with their skin health and if they perceived that their primary care providers assessed their skin and provided education on skin health topics. This included discussing skin protection measures such as the use of sunscreen and a hat, performing a thorough skin examination, and providing education on the importance of recognizing skin lesion changes. METHODS: Any patient who had an appointment for an annual exam, well- child check, or sport physical was eligible to participate in this study. At the completion of the appointment, a survey was provided that gathered information about perceived skin education and assessment received from their healthcare provider. The survey was completed prior to leaving the exam room. The data from the surveys were reviewed and analyzed using the statistical package for social sciences, descriptive statistics. RESULTS: The majority of the participants felt comfortable identifying skin changes, assessing their skin, and their current understanding of skin health. However, more than half of the participants did not perceive that their provider assessed their skin or provided education on skin health. CONCLUSION: Skin health was missed by primary care providers for the majority of patients in this study. Assessing patients' skin and providing them with the tools and knowledge to identify changing skin lesions and preventative skin protective measures, can ultimately increase skin health and decrease the likelihood of skin cancer, specifically melanoma.