Dylan Wu Email & Phone Number

United States

United States

Help organize and curate 30+ articles per semester for the Harvard Brain, a premier undergraduate neuroscience journal dedicated to the study of Mind, Brain, and Behavior

Bethesda, Maryland, United States

(1) Development of a Deep Learning Model for Automated Pixel-WiseSegmentation of Reticular Pseudodrusen from Infra-Red Retinal Images- Created a novel DeepLabV3+ segmentation algorithm on infrared en-face imaging of the retina to facilitate accurate identification and quantification of reticular pseudodrusen lesions, a clinically significant risk factor in progression to age-related macular degenerationAdvisor: Tiarnan Keenan, MD, Ph.D(2) Updated Deep Learning Framework Incorporating Reticular Pseudodrusen for Improved Risk Prediction of Macular Degeneration on Color Fundus Photographs from AREDS/AREDS2- Investigating the impact of reticular pseudodrusen (RPD) on improving the prediction of late age-related macular degeneration (AMD)- Constructed pipeline utilizing an InceptionV3-based Convolutional Neural Network and a Vision Transformer Model to automate grading of imaging features, as well as a Cox Proportional Hazards model for survival analysis of demographic, genetic, and imaging features as risk factors for late AMD and its subtypesAdvisors: Tiarnan Keenan, MD, Ph.D; Emily Chew, MD; Zhiyong Lu, Ph.D

Cambridge, Massachusetts, United States

Cambridge, Massachusetts, United States

- Led a team of associates and analysts while directly coordinating with the client in weekly team meetings- Developed a large language model (LLM)-based platform to streamline the process for collecting peer and public comments from clinicians and other healthcare practitioners for clinical practice guideline development- Created a web-application using a React front-end, PostgreSQL back-end, and a Flask (Python) Rest API endpoint- Configured midpoint and final presentation and deliverables

Cambridge, Massachusetts, United States

- Coordinated with the Case-Team Lead (CTL) to help facilitate case team-client relations to understand the goal and mission behind the data analysis project- Conducted industry market research and analyzing brand performance of a leading global beverage company with over 200 locations worldwide- Developed an interactive dashboard using predictive analysis techniques to highlight data visualizations displaying key performance metrics to simulate possible future outcomes

Cambridge, Massachusetts, United States

Projects:- Developed a model to predict the responsiveness of demand to pricing variables based on historic order activity and other related datasets to maximize profitability and growth for hardware retailer with over 4,500 independent retail locations worldwide- Developed machine learning algorithms to highlight anomalies in how payers are reimbursing the services of a leading genetics testing company and highlight these changes to the appropriate leaders in the organization- Configured midpoint and final deliverables for client presentations and discussion

- Developed a Convolutional Neural Network in TensorFlow to automate gray to white matter ratio calculation in the brain computed tomography scans of comatose cardiac arrest patients- Cleaned, labeled, and preprocessed data through Linux and shell scripting; chose a successful transfer learning architecture to construct the neural network; and performed hyperparameter optimization with techniques like data augmentation and learning rate scheduling to resolve overfitting- Achieved a validation accuracy of ~97% and area under curve (AUC) metric of ~99% for the receiver operating characteristic (ROC) curve.Advisor: Ona Wu, PhD

- Directly assisted with patients daily for their eye check-ups, clinical exams, and visual testing- Skills: Fundus photography, Automatic refraction, Visual field test, Optical coherence tomography, Pachymetry

(1) Comparison of the Clinical and Cost-Effectiveness of Multi-Drop Versus Separate Postoperative Drop Use After Cataract Surgery - Presented at the American Society of Cataract and Refractive Surgery 2022; selected as Best Poster of session(2) An Analysis of the Clinical Efficacy of and Patient Satisfaction with Pilocarpine Hydrochloride 1.25% in Patients with Presbyopia - Analyzed improvement in vision and satisfaction with Vuity™ drops in patients with presbyopia - Used Python to clean, visualize, and analyze data like Mixed Linear Models, Chi Square, T-test - Accepted as a poster to the 2024 American Society of Cataract and Refractive Surgery 2024 in Boston, MA(3) A Comparison of Clinical Efficacy and Patient Satisfaction between a Diffractive Extended-Depth-of-Focus and Enhanced Monofocal Lens - Study comparing the effectiveness of Eyhance™ and Vivity® lenses and patient satisfaction - Utilized Python to analyze results clinical outcomes and survey data - Accepted as poster to 2024 American Society of Cataract and Refractive Surgery 2024 in Boston, MA

(1) Real World Assessment of a Mobile AI Powered Intracranial Hemorrhage Volume Measurement Tool- Validated the accuracy of the Viz.ai ICH segmentation algorithm as a tool for determining hemorrhage volume by comparing its performance to both the ABC/2 formula and semi-autonomous segmentation (SAS) of CT scans- Accepted to 2023 International Stroke Conference (ISC)- Submitted to Stroke for publication(2) Accuracy of Artificial Intelligence-Driven Automated Detection and Volumetric Assessment of Intraventricular Hemorrhage- Assessed the accuracy of the Viz AI segmentation algorithm as a tool for detecting IVH and determining IVH volume by comparing its performance to SAS- Accepted to 2023 American Association of Neurological Surgeons (AANS) Conference(3) Performance of Artificial Intelligence-Powered Evaluation of Midline Shift in Patients with Intracerebral Hemorrhage- Examined the performance of midline shift (MLS) detection and assessment by a convolutional neural network algorithm at a comprehensive stroke center.- Accepted to 2023 American Association of Neurological Surgeons (AANS) ConferenceAdvisor: Christopher Kellner, MD

• Semifinalist at the National Schadt String Competition (2024), Washington International Competition, Johansen International Competition, Stulberg International Competition, and Klein International Competition, 2021-2022. National winner of National MTNA (Music Teachers National Association) Competition, 2021. Silver medal and laureate at the 9th International Tchaikovsky Competition for Young Musicians, Novosibirsk Russia, 2015: youngest American cellist in competition history to be awarded a medal. First and grand prize winners of many international competitions such as Rising Stars Grand Prix International Music Competition, International Grande Music Competition, International Virtuoso Competition, etc.• Performance of works such as Shostakovich Cello Concerto, Elgar Cello Concerto, Tchaikovsky Rococo Variations as soloist with orchestra in venues such as Lincoln Center, Carnegie Hall Berlin, Moscow, Siberia, Taipei, and Korea.

• National YoungArts Finalist, 2020: Selected as one of 20 classical musicians from around the nation for National YoungArts Week, qualifies for US presidential scholar. • Carnegie Hall National Youth Orchestra, 2020: Selected as one of 11 top cellists nationwide to participate in masterclasses and collaborations with top young musicians from across the U.S. • Chamber Music Society of Lincoln Center, 2021-2022: selected as finalist and performance at Allice Tully Hall, 2022. • Worked with distinguished cellists such as Prof. David Geringas, Prof. Laurenece Lesser, and Prof. Myung Wha Chung in international music festivals such as Chigiana Music Academy, Morningside Music Bridge, and Great Mountains Music Festival.• NPR’s program From the Top, 2013-present. Featured on multiple radio broadcasts since 10 years old, invited back to play for several charity events including From the Top’s Gala Event in Boston, Massachusetts. Filmed a music video with trio “Time for Three” covering Taylor Swift song “Shake it Off” which has over 10.5 million views on social media and YouTube. Due to the popularity of this video, Dylan and his cast-mates were invited in 2018 to be featured at the Horatio Alger Foundation’s fund-raiser and in their March 2019 issue, Strings Magazine named the video one of the best classical covers. Most recently, he was featured when Ponticelli’s Over the Rainbow from the Wizard of Oz was showcased on From the Top’s Daily Joy series during COVID-19.

Highly selective program for students who have a strong interest in the sciences and mathematics.Courses:2019-2020:• Classical and Quantum Computing Devices• Computer Programming in Python2020-2021:• Quantum Physics• Organic Chemistry2021-2022:• Explorations in Data Science with R• Introduction to Algorithms

Selected as 1 of 35 students from around the nation to compete for a spot on the U.S.team to compete in the International Earth Science Olympiad; qualified in 2021 forinternational team

2x AIME Qualifier; top 2.5% of AMC 10 and top 5% of AMC 12 scorers in the nation

Studied number theory proofs at Ohio State University

Bachelor Of Arts - Ba, Applied Mathematics W/ Data Science, Cs Secondary

GPA: 4.0/4.0 Activities: Co-principal cellist, Harvard Radcliffe Orchestra (HRO) Director of Strategy, Harvard Data Analytics Group (HDAG).

Master'S Degree, Music

Private lessons with Prof. Lluis Claret Most Recent Accomplishments include: Performance of Beethoven Triple Concerto in Lotte Concert.

Cello Performance

Studied with Dr. Clara Kim Principal Cellist of the Juilliard Precollege Symphony Orchestra Winner, Juilliard Concerto Competition, 2022.

Education record

Valedictorian Recipient of the Superintendent’s Award, AP Scholars Award (completed 19 AP courses and received a 5/5 in 18 of 19 courses).