David Dreizin, MD

Education:

Cornell University, B.S. Molecular Biology, (Magna cum Laude) (1997-2001)

Weill Cornell Medical College, MD (2003-2007)

Training/Residency:

Albert Einstein College of Medicine/Montefiore Medical Center (Internal medicine, Preliminary Year, 2007-2008)

University of Miami/Jackson Memorial Hospital (Residency, 2008-2012). Served as chief resident

Johns Hopkins Cross-sectional Imaging (Fellowship, 2012-2013)

Diplomate, American Board of Radiology, 2012

Dr. Dreizin is an Associate Professor in the section of Trauma and Emergency Radiology, a clinician-scientist, R01 funded PI, and director of TRAIL (the Trauma Radiology Artificial Intelligence Laboratory), a center within the University of Maryland Department of Diagnostic Radiology and Nuclear Medicine.

Dr. Dreizin's previous scientific work: Dr. Dreizin is internationally recognized for leading multi-disciplinary projects and mentoring students and post-docs on detection, segmentation, and quantitative visualization of challenging traumatic pathology on trauma whole body CT, including pooled hemorrhage, active bleeding, organ injuries, and pelvic fractures. He has published several notable firsts including the first algorithms world-wide for segmentation and quantification of traumatic pelvic hematoma and hemoperitoneum, and solid organ lacerations. See highlighted publications for selected works.

Dr. Dreizin's clinical and educational work: As a clinical practicioner and educator, Dr Dreizin believes that mastery of trauma radiology doesn't occur within a vacuum or echo chamber and the practitioner in this specialty is a "doctor's doctor" that must bring value to multidisciplinary trauma teams by providing timely evidence-based interpretations and consultations focused on optimizing surgical outcomes and minimizing morbidity and mortality.

As an initially body fellowship-trained radiologist, through inspiring mentors, Dr. Dreizin discovered early in his career that working in the trauma field is exhilarating, intellectually stimulating, and deeply fulfilling because it has an immediate impact on the lives of trauma victims. He enjoys working with like-minded autodidacts with diverse backgrounds who caught the trauma radiology bug at various stages of their radiology careers (once you catch it, it's incurable!)

Trauma radiology is a life-long learning process that requires intimate knowledge of surgical principles, techniques, and practice patterns. CT is the routine workhorse imaging modality in our field. Dr. Dreizin has written numerous reviews for practicing radiologists and trainees in RadioGraphics focusing on the utility of CT in neurotrauma, musculoskeletal trauma, and organ and vascular trauma that emphasize understanding how trauma surgeons leverage CT imaging to aid decision-making, with a clear-eyed understanding of both the strengths and weaknesses of the modality, and its often-complementary role in this process.

Dr. Dreizin's growth as both a scientist and educator wouldn't have been possible without passionate, generous, and selflessness mentors (it really takes a village...). Nothing has been more fulfilling in his scientific and teaching career than giving back by mentoring the next generation of radiologists and scientists and seeing them succeed in their chosen path.   

About Dr. Dreizin's lab: Our work at TRAIL currently focuses on developing, deploying, and scaling novel detection, quantitative visualization, grading, and outcome prediction methods for hemorrhage-related traumatic torso injuries to very large whole body CT datasets with in-the-wild injury prevalence and distributions. Other avenues being explored include mask-conditional data augmentation, and vision language models. TRAIL compute resources currently include an 8x H100 GPU server, and two 4x A6000 GPU PCs. Further, we are collaborating with a number of highly-regarded trauma centers on dataset curation and external validation. At University of Maryland, TRAIL currently directly supports computer science PhD post-doctoral fellows with experience in computer vision.          

Blunt and penetrating torso trauma, Advanced imaging, Computer-assisted detection and diagnosis, Artificial Intelligence, Machine Learning, and Computer Vision

1. Dreizin D, Zhou Y, Chen T, Li G, Yuille AL, McLenithan A, Morrison JJ. Deep learning-based computer-aided detection and quantification of pelvic hematoma volumes in patients with pelvic fractures: potential role in decision support and prognostication. In submission to Journal of Trauma and Acute Care Surgery. J Trauma Acute Care Surg. 2020;88(3):425‐433. doi:10.1097/TA.0000000000002566

2. Dreizin D, Zhou Y, Fu, S, Wang Y, Li G, Champ K, Siegel E, Chen T, Yuille A. A multiscale deep learning method for quantitative visualization of traumatic hemoperitoneum: assessment of feasibility and comparison with subjective categorical estimation for outcome prediction and decision Radiol Artif Intell. 2020 Nov 11;2(6):e190220. doi: 10.1148/ryai.2020190220. eCollection 2020 Nov.PMID: 33330848

3. Dreizin D, Chen T, Liang Y, Zhou Y, Paes F, Wang Y, Yuille A, Roth P, Champ K, Li G, McLenithan A, Morrison JJ. Added value of deep learning-based liver parenchymal CT volumetry for predicting major arterial injury after blunt hepatic trauma: a decision tree analysis. Abdom Radiol (NY). 2021 Jun;46(6):2556-2566.

4. Dreizin D, Goldmann F, LeBedis C, Boscak A, Dattwyler M, Bodanapally U, Li G, Anderson S, Maier A, Unberath M. An automated deep learning method for Tile AO/OTA pelvic fracture severity grading from trauma whole-body CT. J Digit Imaging. 2021 Feb;34(1):53-65. doi: 10.1007/s10278-020-00399-x. Epub 2021 Jan 21

5. Zapaishchykova A, Dreizin D, Li Z, Wu JY, Roohi SF, Unberath M. An Interpretable Approach to Automated Severity Scoring in Pelvic Trauma. Med Image Comput Comput Assist Interv. 2021 Sep-Oct;12903:424-433. doi: 10.1007/978-3-030-87199-4_40. Epub 2021 Sep 21. PMID: 37483538; PMCID: PMC10362989.

6. Dreizin D, Nixon B, Hu J, Albert B, Yan C, Yang G, Chen H, Liang Y, Kim N, Jeudy J, Li G, Smith EB, Unberath M. A pilot study of deep learning-based CT volumetry for traumatic hemothorax. Emerg Radiol. 2022 Dec;29(6):995-1002. doi: 10.1007/s10140-022-02087-5. Epub 2022 Aug 16. PMID: 35971025; PMCID: PMC9649862.

7. Chen, H., Unberath M, Dreizin D. Toward automated interpretable AAST grading for blunt splenic injury. Emerg Radiol. 2023 Feb;30(1):41-50. doi: 10.1007/s10140-022-02099-1. Epub 2022 Nov 12. PMID: 36371579; PMCID: PMC10314366.

8. Dreizin D, Staziaki PV, Khatri GD, Beckmann NM, Feng Z, Liang Y, Delproposto ZS, Klug M, Spann JS, Sarkar N, Fu Y. Artificial intelligence CAD tools in trauma imaging: a scoping review from the American Society of Emergency Radiology (ASER) AI/ML Expert Panel. Emerg Radiol. 2023 Jun;30(3):251-265, 2023 Mar 14. doi: 10.1007/s10140-023-02120-1. Epub ahead of print. PMID: 36917287.

9. Agrawal A, Khatri GD, Khurana B, Sodickson AD, Liang Y, Dreizin D. A survey of ASER members on artificial intelligence in emergency radiology: trends, perceptions, and expectations. Emerg Radiol. 2023 Jun;30(3): 267-277, 2023 Mar 13. doi: 10.1007/s10140-023-02121-0. Epub ahead of print. PMID: 36913061.

10. Dreizin D, Nam AJ, Diaconu S, Bernstein M, Bodanapally UK, Munera F. MDCT of         Midfacial Fractures: Classification Systems, Principles of Reduction, and Common Complications. (cover article) Radiographics. 2018 Jan-Feb;38(1):248-274.

11. Dreizin D, Nam AJ, Tirada N, Levin MD, Stein DM, Bodanapally UK, Mirvis SE, Munera F. Multidetector CT of Mandibular Fractures, Reductions, and Complications: A Clinically Relevant Primer for the Radiologist. Radiographics. 2016 Sep-Oct;36(5):1539-1564.

12. Dreizin D, LeBedis CA, Nascone JW. Imaging Acetabular Fractures. Radiol Clin North Am. 2019 Jul;57(4):823-841. doi: 10.1016/j.rcl.2019.02.004. Epub 2019 Apr 1. PMID: 31076035.

13. Dreizin D, Sakai O, Champ K, Gandhi D, Aarabi B, Nam A, Morales R, Eisenman D. CT of skull base fractures: classification systems, complications, and management. 2021 May-Jun;41(3):762-782. doi: 10.1148/rg.2021200189. Epub 2021 Apr 2. PMID: 33797996.

14. Dreizin D, Smith EB, Champ K, Morrison JJ. Roles of Trauma CT and CTA in Salvaging the Threatened or Mangled Extremity. Radiographics. 2022 Mar-Apr;42(2):E50-E67. doi: 10.1148/rg.210092. PMID: 35230918; PMCID: PMC8906352.

15. Dreizin D, Smith EB. CT of Sacral Fractures: Classification Systems and Management. Radiographics. 2022 Nov-Dec;42(7):1975-1993. doi: 10.1148/rg.220075. Epub 2022 Sep 16. PMID: 36112523.

16. Dreizin D, Edmond T, Zhang T, Sarkar N, Turan O, Nascone J. CT of Periarticular Adult Knee Fractures: Classification and Management Implications. Radiographics. 2024 Sep;44(9):e240014. doi: 10.1148/rg.240014. PMID: 39146203.

For complete list of citations, see:  Dreizin D - Search Results - PubMed

2010  Certificate of Merit for The Role of Whole Body 64 MDCT Angiography for Blunt Polytrauma, RSNA educational exhibit 

2010  Cum Laude Award for Evaluation of Cervical Spine Injuries using Multi-Detector CT, RSNA educational exhibit 

2011  Certificate of Merit for 64 Slice MDCT with Post-processing in the Evaluation of Penetrating Diaphragmatic Injury, ASER Meeting Educational Poster

2012  Roentgen Resident/Fellow Research Award. RSNA Research and Education          Foundation     

2013  Certificate of Merit for 3 Tesla Chemical Shift MR Imag­ing: Technique, Clinical Util­ity and Pitfalls for Imaging the Skeleton, RSNA  

2014 Certificate of Merit Award for MDCT of High Energy Pelvic Ring Disruptions in Blunt Trauma, RSNA

2014  Cum Laude Award/Invited to Radiographics for Imaging of Pregnant Patients: Fetal Dose and Beyond, RSNA educational exhibit & CME informal oral presentation

2014 Certificate of Merit/Invited to RadioGraphics MDCT of Blunt Mandibular Trauma, RSNA

2015  Certificate of Merit Award/Invited to RadioGraphics for Reviews and Updates: Iodinated Contrast, Pressure Injector, and Protocols in Body Imaging, RSNA

2016  Certificate of Merit/Invited to RadioGraphics for MDCT of Midfacial Fractures: Classification Systems, Principles of Reduction, and Common Complications, RSNA

2016  Manuscript Reviewer with Distinction, RadioGraphics

2017-2019,  2021, 2023  Manuscript Reviewer with Special Distinction, RadioGraphics     

2019  Outstanding mentorship recognition award (from AOA/OSR/UMSOM)

2021  Certificate of Merit for CT of Sacral Fractures: Classification Systems and Management. Invited to RadioGraphics.

2022  Certificate of Merit- ASER for Blunt splenic trauma: accuracy of automated active bleed and contained vascular injury detection on CT with Faster R-CNN

2022  Featured in RSNA Journal Highlights for DECT/Cinematic Rendering paper in Radiology

2023  “CT of adult periarticular knee fractures: classification and management” invited for consideration in RadioGraphics        

07/01/16–06/30/19   RSNA Research Scholar Grant

11/01/17–10/30/20 PI, “DECT and Cinematic Rendering for Pelvic Trauma”, Siemens/UMMC

6/1/2019-5/30/2020 PI, University of Maryland Accelerated Translational Incubator Pilot (ATIP) Grant

7/1/2019- 2023 PI, NIH Mentored Clinician Scientist Award K08 EB027141-01A1

9/1/2023-8/31/2027 PI, NIH R01 GM148987-01

9/1/2024-8/31/2025 NIH RO1 GM148987-01 Equipment Supplement

About TRAIL: Our work currently focuses on scaling novel detection, quantitative visualization, grading, and outcome prediction methods for hemorrhage-related traumatic torso injuries to very large whole body CT datasets with in-the-wild injury prevalence and distributions. Other avenues being explored include mask-conditional data augmentation, and vision language models. TRAIL compute resources currently include an 8x H100 GPU server, and two 4x A6000 GPU PCs. Further, we are collaborating with a number of highly-regarded trauma centers on dataset curation and method validation. At University of Maryland, TRAIL currently directly supports post-doctoral fellows with extensive computer vision expertise.