BIG DATA NATURAL LANGUAGE PROCESSING RESEARCH RETROSPECTIVE ELECTRONIC MINING MACHINE LEARNING TEXT MEDICAL STATISTICS RECORDS INTERPRETABLE CLASSIFICATION NARRATIVE ACCURACY CLINICAL DIAGNOSIS
PAS 2024

This year's PAS presenters who used Doctor T all demonstrate that modestly-resourced research teams can generate impressively upscaled results using Doctor T workflows and AI-based clinical NLP technology.

Abstract. Authors
Prevalence of Osteoarthritic infection patients presenting with low inflammatory markers: a multi-center analysis
Session: Emergency Medicine 1 Works in Progress
Friday, May 3, 2024
5:15 PM – 7:15 PM ET
Nickolas Mancini,
Rate of bacteremia among children with skin abscesses that had a blood culture drawn
Session: Emergency Medicine 4: Infections
Saturday, May 4, 2024
3:30 PM – 6:00 PM ET
Elizabeth M. Waltman
Discrepancy between expected skin abscess incision & drainage outcome and amount of pus expressed
Session: Emergency Medicine 4: Infections
Saturday, May 4, 2024
3:30 PM – 6:00 PM ET
Jeffrey T. Neal, Harvard Medical School / Boston Children's Hospital,
Guidelines to standardize care for facial lacerations can improve racial inequities in plastic surgery repairs in the pediatric emergency department
Session: Health Equity/Social Determinants of Health 4
Saturday, May 4, 2024
3:30 PM – 6:00 PM ET
David Mills,
Natural Language Processing-Assisted Evaluation of Empiric Antibiotics for Skin and Soft Tissue Infections
Session: Emergency Medicine 4: Infections
Saturday, May 4, 2024
3:30 PM – 6:00 PM ET
James Rudloff,
Enhancing surveillance of healthcare-associated violence using Natural Language Processing and clinical notes
Session: Quality Improvement/Patient Safety 2
Sunday, May 5, 2024
3:30 PM – 6:00 PM ET
Amir Kimia,
Assessing Medical Large Language Models for Semantic Search in Pediatric Clinical Narratives
Session: Technology 2
Sunday, May 5, 2024
3:30 PM – 6:00 PM ET
Assaf Landschaft,
Use of Artificial Intelligence for lost-to-follow-up surveillance
Session: Quality Improvement/Patient Safety 3
Monday, May 6, 2024
9:30 AM – 11:30 AM ET
Amir Kimia
About Doctor T

Doctor T (DRT – Document Review Tool) is a graphic user interface for natural language processing (NLP)-assisted and NLP driven data review. DRT was developed by pediatric emergency medicine physician, Dr. Amir Kimia MD, and a system architect computer scientist, Assaf Landschaft MSC.

DRT software can be used by those individuals with little to no IT/computer science expertise. The software is user friendly and clinicians can leverage their training, insights, and experience to review the clinical notes provided by the data mining.

There was extensive time and energy put into which NLP modules are incorporated into DRT to make it intuitive, and prone to less errors. As artificial intelligence tools continue to evolve, newer methodologies are ongoing along with updated versions. DRT is not a commercial product or software. It has been used extensively by clinicians and quality improvement professionals at a pediatric hospital and a community hospital. If you are interested or have an idea to help improve patient care, the team that services and generates DRT includes contractors, computer scientists and research assistants that are paid hourly, mainly through our funded collaborative projects.

For those interested in using DRT for unfunded projects and cannot support the cost of installation and technical support/version updates, it is recommended to reach out to see if there is potential for collaboration. Our team is always open to new projects involving collaboration in federal and non-federal research grants. Alternatively, DRT can be installed at your institution for a fee which includes a minimum number of hours for installation, technical support, software updates, and natural language processing basic rules and teaching.

DRT workflow includes checks and balances, along with your medical ethics, to support a high standard of data analysis to help advance patient care. We have a high regard ensuring new users understand the AI and platform. We have a moral responsibility to make sure the data and science being used are sound. For questions please use the below contact information.

For example of some of our work, our recent publications grants and presentations appear on this website.

Contact:

Amir Kimia, MD

Associate Professor of Pediatrics, Harvard Medical School

Faculty Boston Children’s ED and Clinical Informatics Fellowship Program

Email: amir.kimia@childrens.harvard.edu

Phone: 617-355-6624