2022 PEL SUmmer INTERNSHIP Program
PEL is proud and excited to offer a fully-funded summer internship to students who wish to pursue projects in healthcare innovation. For 10 weeks during the summer between MS1 and MS2, selected students will structure their own internship working with a mentor of their choosing on a project they jointly design. Mentors may be internal to TJUH, or part of the broader healthcare innovation community, and should be identified by the student (though PEL is able to provide contacts and introductions, where appropriate).
Projects should be well-defined, and should have a specific end-goal in mind: such as getting a start-up off the ground, finalizing a research project, creating a new program or prototype, etc. The project should take approximately 30 hours per week over 10 weeks to complete. Prior to submitting an application, students should identify both the mentor and the project, and should receive written verification of the mentors' willingness to host the student’s work. Students should be prepared to complete 3 milestones/deliverables over the summer culminating in a final presentation for the PEL executive board and SKMC leadership in the fall of 2022. The usual stipend is ~$1,500.
If you are interested in applying to the PEL Summer Internship but have not yet identified a mentor, feel free to reach out to the PEL Internship team: kellen.round@students.jefferson.edu, sarah.wenyon@students.jefferson.edu, and sanath.patil@students.jefferson.edu.
All students participating in the PEL Summer Internship Program will be expected to fulfill the following requirements:
Meet the project-specific end-goals laid out in the application
Present on the project and experience at a fall session for PEL executive members and representatives from Jefferson’s Office of Student Affairs
Complete an online program evaluation
Provide evidence of continued feedback from the mentor
Application
At the time of application, students should be first-year medical students at SKMC and in good academic standing. Applications will be evaluated based on:
The clarity of the student’s project proposal and the identification of specific end-goals
The presence of a written agreement from the student’s proposed mentor to participate in the program for 10 full weeks
The students’ prior work or extracurricular experience in self-structured environments
Applicants will be notified if they have been accepted to the program, not accepted, or placed on a wait list. The number of students we accept is entirely determined on the total amount of funding we have to offer - students who are interested in completing their projects and participating in the program in the absence of funding should identify their interest in their application.
Applications for 2022 are closed. Stay tuned for 2023 applications in March 2023!
Past ProjectS
Name - Abhiraj Saxena
Variable ranking of donor and recipient characteristics for transplant survival using machine learning models
Description - Currently, heart transplantation is the gold standard for end-stage heart failure treatment. However, with almost 40% of United States adults being predicted to have heart disease by 2030, there is a serious concern that heart transplantation supply will not be able to meet the demand. More efficient methods of matching donors to recipients are needed in order to save time, money, and lives.
The goal of this project was to investigate further into how to construct a model that would match donors and recipients by looking at the top twenty most important donor and recipient variables most involved in unsuccessful heart transplantation outcomes. If we could find out these variables, then we could consider these factors when matching patients.
I worked with Dr. Vakhtang Tchantchaleishvili on this project to gather the data from the United Network for Organ Sharing (UNOS) and run a minimum redundancy maximum relevance (mMRMe) machine learning model to get a ranking of the top twenty variables. Certain methods such as data imputation were performed in order to predict missing data from the dataset to strengthen the analysis.
The target shuffling method, developed by John Elder, was used to determine the most important characteristics of donors and recipients in predicting survival in heart transplantation operations. Future steps to improve the project include incorporating socioeconomic data, fine-tuning the properties of the model or finding a better one altogether, and addressing the issue of missing data by proposing a standardized list of variables that each heart transplant center reports for patients.
Name - Kathryn Achuck
Decreasing the Healthcare Divide for Children with Neurodiversity
Description - The needs of patients with neurodiversity, who have different ways of processing, and communicating can be unmet, especially in high-stress environments. They have more annual physician visits, Emergency Department (ED) visits, and hospitalizations. Furthermore, these patients experience decreased care quality due to the communication, sensory and behavioral challenges in high-intensity environments.
To enhance the quality of care for neurodiverse patients and bridge this healthcare gap, I worked closely with Dr. Ross and the Center for Neurodiversity to learn more about her patients and their families. Our goals were twofold. First, to improve the needs of her patients on an interpersonal level with their healthcare providers via the creation of a healthcare passport to better communicate needs. Second, to expand and ideologically identify the needs of neurodiverse patients more efficiently by negotiating for better levels of care.
My work has culminated with a healthcare passport prototype that can be used by Dr. Ross’ patients to help address their most salient healthcare needs. We are currently working with Tendo, a health design company, to create an operational version of our prototype. Moreover, through a broader collaboration with the Institute for Exceptional Care think tank and Boston Children’s Hospital, we worked to better identify and code patients with an intellectual disability (IDD) to predict future care resources. By better stratifying patients diagnosed with an IDD, we are beginning the process of negotiating for better funding and levels of care. We are working to publish guidelines for how clinicians can improve the medical coding for their patients. In the future, we plan to use this stratification to generate artificial intelligence that can predict the cost of care so that state and federal services can budget and allocate funds appropriately.
Name - Mia Belovsky
Evaluating Driver Fatigue Measures with a Device Proposal
Description - Driver fatigue is a leading cause of traffic accidents in the United States and is responsible for over 328,000 accidents, 109,000 injuries and 6,400 preventable deaths annually. The physiologic effects of sleep deprivation on the body mimics that of alcohol impairment, as being awake for more than 20 hours is equivalent to being legally drunk according to the NSC. This makes drivers three times more likely to get into an accident while behind the wheel. Further research on this topic is imperative in order to minimize the preventable deaths and injuries caused by driver fatigue.
To address this pressing public health issue, I worked with Dr. Miller, who is the Executive Director of the Jefferson Center for Injury Research and Prevention to research driver fatigue and develop a device to most accurately and efficiently measure it. The project began with a literature review of the current measures and devices being used to combat driver fatigue as well as their accuracy, effectiveness and feasibility. A thorough investigation of solutions on the market as well as an exploration of the best metrics used to sense driver fatigue was performed.
To take action against driver fatigue, we incorporated our knowledge of digital health, specifically wearable devices, and medical knowledge to reinvent a device that can monitor fatigue behind the wheel and reduce the number of fatigue related accidents. Using a previously patented design in conjunction with additional biosensors, I developed a seatbelt-based device that detects fatigue by utilizing facial and eye monitoring, heart rate monitoring, a strain gauge system to detect head nod and a built-in alarm system. Our team is now looking to collaborate with designers and engineers at Jefferson’s Kanbar College of Design, Engineering and Commerce to create a more succinct development plan and prototype. The end goal of this project would be to file a patent and bring said device to market after completing testing on patients.
Name - Kayla Holston
Strategic Design Development for Malawian Public Health Center Labor & Delivery Suites
Description - Overcrowded public tertiary care centers in Malawi often require mothers to deliver children in the same room as multiple other mothers. This has created concerns related to maternal privacy (particularly confidentiality, dignity, and nosocomial infection), paternal involvement in birthing process, and staff workflow efficiency at Queen Elizabeth Central Hospital (QECH). The Director of Obstetrics & Gynecology at the University of Malawi College of Medicine (COM), Dr. Luis Gadama, has requested a novel labor and delivery suite design for QECH in Blantyre, Malawi that addresses the aforementioned concerns from a strategic design perspective. This project aims to understand and quantify (where possible) the current state of the aforementioned concerns, develop a sustainable labor and delivery suite architectural design informed by our results, and then reassess the state of factors listed above after design intervention.
Name - Shreyas Chandragiri
3D Printing in Cardiology
Description - The heart is an incredibly complex organ and learning its precise anatomy is a critical component of training for clinical cardiologists. Yet, due to its complexity, it takes many years for both clinical and interventional cardiologists to truly master this subject. A universal limitation in cardiology training and pre-procedural planning is the use of imaging to learn anatomy. While surgeons cut directly into the body and get extensive interaction with the physical anatomy, cardiologists must rely on imaging more heavily. The translation of 2-dimensional medical imaging to its actual 3D structure presents a challenging hurdle. To alleviate this gap, I proposed the incorporation of 3D printing to better facilitate the visualization of cardiac anatomy from imaging. I designed several teaching models commissioned by the cardiology department designed to aid cardiology department in teaching their trainees. These include, 3D printed coronary angiogram teaching models to facilitate teaching of coronary angiogram views, a 3D printed echo model, and a 3D printed model of the left and right outflow tracts at the cardiac annuli for EP fellowship training. The first three projects will serve as the initial proof of concept to display to the cardiology department the value that 3D printing can add for trainees. Each of these devices is currently under research using a cohort of trainees to assess their utility. We hope and suspect that the results of these studies will indicate that the use of 3D models can greatly enhance skill acquisition cardiology training and will become a standard practice in education. Additionally, to prove that these 3D printed models can be used in a wide variety of clinical settings, a low-cost medical grade 3D printer was built, and utilized in the production of the teaching models.
Name - David B. Ney
Description - 99% of all Alzheimer’s disease drug trials have failed. It is increasingly likely that none of the current trials will prove to be effective and it is conceivable that no pharmaceutical treatment will prevent or delay the effects of dementia in my lifetime. However, patients and families living with dementia need support for this devastating illness. First, I wrote a reflective essay for peer-reviewed and competition-based submission about communicating with patients with dementia. This essay included personal accounts and focus on themes of advocacy, creativity, and empathy. It will include a description of TimeSlips for the purpose of empowering caregivers. Second, I recorded and produced a podcast series about receiving a dementia diagnosis and what a patient/family faces. In particular, this series focused on the innovative organization TimeSlips. Third, I directed and filmed a short video about TimeSlips’s unique approach to communication with patients with dementia. The purpose of the podcast and video was twofold: 1) increase the visibility of a novel approach to dementia care in Philadelphia and 2) recruit members for Jefferson’s own TimeSlips chapter, which is launched in September of 2019 at Old City Presbyterian Apartment. Terrence Casey at the Penn Memory Center will directly oversee the project with additional help from Megan Voeller, Jason Karlawish, Louis Massiah, and Anne Basting. The impact of the summer internship will extend beyond the summer in the form of the TimeSlips chapter at Jefferson University and with the possibility of publications for reviews on “dementia-friendly” care as well as peer-reviewed editorials about the launch of TimeSlips at a medical school.
Check out David’s project here!
Name - Travis Clarke
Description - Physicians have used computed tomography (CT) scans as a first line diagnostic tool. By using a combination of X-ray and computational mathematics, CT images create rendered voxels that can show problems with a patient obviating the need for other more expensive imaging modalities. Accordingly, physicians have been using CT scans with increasing frequency in recent years. Emergent CT scans must be read quickly, especially in stroke cases. Additionally, though radiologists are highly trained and adept, difficult cases such as cranial fracture can be missed and misdiagnosed. To alleviate these problems, researchers have applied deep learning using convolutional neural networks (CNNs) with marked success to medical imaging and diagnosis. However, current solutions are limited in the number of potential classifications that the algorithm can perform. A better algorithm using an accurate and rich dataset is needed to improve the current deep learning models for classification of CT scans. By using 3D CNNs, I can accurately diagnose head CT scans to improve the radiology workflow. By classifying head CT scans, I have collected a dataset of cases from the Jefferson Hospital System PACs system. By using natural language queries, I have collected over 50,000 cases with appropriate radiological diagnoses. After using preprocessing to crop, level and normalize the images, trained CNNs that have previously shown promise in 3D scans. These networks include but are not limited to, VoxNet, 3D ResNet, and 3D Inception-based CNNs. By training a 3D CNN on my dataset, I was be able to classify head CT scans accurately. Using this model, I created an effective workflow for the radiology department, to help prioritize urgent cases. Patient outcomes and hospital efficiency will improve as physicians are better able to diagnose and triage patients by using the aid of machine learning.
Name - Alexander M. Olson
Description - My PEL project aims to break new ground in the medical education space through creation of accessible and interactive animated content. So many of the processes that are central to the mechanisms and pathophysiology of pre-clinical medicine are in desperate need of detailed animation, an overdue departure from the static and overly crowded diagrams so often included in lecture slides. In partnership with Dr. Steven Herrine, I am in the process of creating a pilot module focused on GI/Hepatology that will be released to the JeffMD class of 2022 during their GI curriculum block. Using the Adobe CC tool suite and Powerpoint/Keynote, I am creating a user-directed, stepwise tutorial on cirrhosis and portal hypertension, integrating concepts and systems from other organ systems where appropriate. The module will conclude with a self-assessment quiz to reinforce critical takeaways. I will present a draft of this project at the AASLD conference in San Francisco this coming November.
How was your PEL Internship experience? What did you enjoy about it?
My PEL Internship experience was fantastic, and continues to be fun. Despite having widely different summer schedules and plans, there was an atmosphere of support and community amongst the interns and PEL board members. We encouraged one another, gave candid feedback and learned from one another's challenges through our widely different projects. Being a part of a small team of students attempting to create something novel and exciting has been an invaluable learning experience. I look forward to seeing our respective work come to fruition.
Name - Benjamin Chipkin
Description - Research suggests that doctors who know how to cook and make healthy lifestyle choices have healthier patients and are healthier themselves. JeffCHEPh (Jefferson Cooking Healthily and Equitably for Philadelphia) offers medical students the opportunity to learn about nutrition, community health, physician wellness, and interprofessional collaboration through immersive culinary experiences. Medical students will develop skills to improve the root of many of our most pressing healthcare issues: access to and promotion of healthy eating. Learning to shop for and prepare meals within Philadelphia's unique food landscape, JeffCHEPh participants will build an understanding of the food challenges facing our community and practical steps to ameliorate them. Culinary medicine, the melding of culinary arts and medical science, provides an evidence-based model to realize community wellbeing. JeffCHEPh will bring this growing field of culinary medicine to the Jefferson community to ensure that those we serve and those we work alongside continue to live healthy and sustainable lives.
How I enjoyed the PEL Internship experience:
We all begin medical school with passions that extend beyond the classroom, which are easy to neglect along the way. Arriving to Jefferson with my MS in nutrition and a lifelong love of cooking, the first year of medical school left me hungry for more exposure to clinical nutrition. The PEL Internship helped me reconnect with my passion for food and provided a space for me to innovate in the field of nutrition education. Organizationally, PEL supports interns with community connections and resources that enable meaningful projects to emerge from the summer. The Internship offers a unique opportunity to put creative thought into action and develop leadership skills to become an impactful physician.