John Zech

About Me

I am an assistant professor of radiology at Columbia.

Prior to this, I completed medical school at Mount Sinai, a preliminary medicine year at California Pacific Medical Center, radiology residency at Columbia, and a musculoskeletal radiology fellowship at NYU. Before medicine, I worked in investment management and completed a master’s degree in Statistics at Columbia.

Research Interest

I am working to understand how we can apply machine learning to create effective clinical decision support in radiology, particularly in musculoskeletal imaging.

Publications

  1. S.M. Santomartino, J.R. Zech, K. Hall, J. Jeudy, V. Parekh, P.H. Yi. Evaluating the performance and bias of natural language processing tools in labeling chest radiograph reports. Radiology. 2024.
  2. J.R. Zech, C.O. Ezuma, S. Patel, C.R. Edwards, R. Posner, E. Hannon, F. Williams, S.V. Lala, Z.Y. Ahmad, M.P. Moy, T.T. Wong. Artificial intelligence improves resident detection of pediatric and young adult upper extremity fractures. Skeletal Radiology. 2024.
  3. J.R. Zech, L. Santos, S. Staffa, D. Zurakowski, K.A. Rosenwasser, A. Tsai, D. Jaramillo. Lower Extremity Growth according to AI Automated Femorotibial Length Measurement on Slot-Scanning Radiographs in Pediatric Patients. Radiology. 2024.
  4. E.F. Alaia, M. Samim, I. Khodarahmi, J.R. Zech, A.R. Spath, M.D.S. Cardoso, S. Gyftopoulos. Utility of MRI for Patients 45 Years Old and Older With Hip or Knee Pain: A Systematic Review. Americal Journal of Roentgenology. 2024.
  5. J. Adleberg, C.L. Benitez, N. Primiano, A. Patel, D. Mogel, R. Kalra, A. Adhia, M. Berns, C. Chin, S. Tanghe, P. Yi, J. Zech, A. Kohli, T. Martin-Carreras, I. Corcuera-Solano, M. Huang, J. Ngeow. Fully Automated Measurement of the Insall-Salvati Ratio with Artificial Intelligence. Journal of Imaging Informatics in Medicine. 2024.
  6. J.R. Zech, D. Jaramillo, J. Altosaar, C.A. Popkin, T.T. Wong. Artificial intelligence to identify fractures on pediatric and young adult upper extremity radiographs. Pediatric Radiology. 2023.
  7. P.H. Yi, H.W. Garner, A. Hirschmann, J.A. Jacobson, P. Omoumi, K. Oh, J.R. Zech, Y.H. Lee. Clinical Applications, Challenges, and Recommendations for Artificial Intelligence in Musculoskeletal and Soft Tissue Ultrasound: AJR Expert Panel Narrative Review. American Journal of Roentgenology. 2023.
  8. J.R. Zech, G. Carotenuto, Z. Igbinoba, C.V. Tran, E. Insley, A. Baccarella, T.T. Wong. Detecting pediatric wrist fractures using deep-learning-based object detection. Pediatric Radiology. 2023.
  9. J.R. Zech. Using BERT Models to Label Radiology Reports. Radiology: Artificial Intelligence. 2022.
  10. J.R. Zech, S.M. Santomartino, P.H. Yi. Artificial Intelligence (AI) for Fracture Diagnosis: An Overview of Current Products and Considerations for Clinical Adoption, From the AJR Special Series on AI Applications. American Journal of Roentgenology. 2022.
  11. J.R. Zech, G. Carotenuto, D. Jaramillo. Inferring pediatric knee skeletal maturity from MRI using deep learning. Skeletal Radiology. 2022.
  12. S.L. Tummalapalli, J.R. Zech, H.J. Cho, C. Goetz. Risk stratification for hydronephrosis in the evaluation of acute kidney injury: a cross-sectional analysis. BMJ Open. 2021.
  13. G. Carotenuto, A. Brewer-Hofmann, J.R. Zech, S. Sajjad, Z.N. Bekheet, D. Jaramillo, T.T. Wong. Identifying Factors Important to Patients for Resuming Elective Imaging During the COVID-19 Pandemic. Journal of the American College of Radiology. 2021.
  14. J.R. Zech, J.Z. Forde, M. L. Littman. Individual predictions matter: Assessing the effect of data ordering in training fine-tuned CNNs for medical imaging. arXiv. 2019.
  15. M.A. Badgeley, J.R. Zech, L. Oakden-Rayner, B.S. Glicksberg, M. Liu, W. Gale, M.V. McConnell, B. Percha, T.M. Snyder, J.T. Dudley. Deep learning predicts hip fracture using confounding patient and healthcare variables. Nature Digital Medicine. 2019. Previously preprinted on arXiv.
  16. D.A Kaji, J.R. Zech, J.S. Kim, S.K. Cho, N.S. Dangayach, A.B. Costa, E.K. Oermann. An attention based deep learning model of clinical events in the intensive care unit. PloS One 14 (2), e0211057. 2019.
  17. B. Marinelli, M. Kang, M. Martini, J.R. Zech, J. Titano, S. Cho, A.B. Costa, E.K. Oermann. Combination of Active Transfer Learning and Natural Language Processing to Improve Liver Volumetry Using Surrogate Metrics with Deep Learning. Radiology: Artificial Intelligence. 2019.
  18. J.R. Zech, M.A. Badgeley, M. Liu, A.B. Costa, J.J. Titano, E.K. Oermann. Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study. PLoS Med 15(11): e1002683. 2018. Previously preprinted on arXiv as ‘Confounding variables can degrade generalization performance of radiological deep learning models.’
  19. M.A. Badgeley, B.S. Glicksberg, M. Liu, M. Shervey, J. Zech, S. Khader, N.A. Knudson, A. Costa, J. Titano, J. Schefflein, A. Su, M.V. McConnell, J. Lehar, E.K. Oermann, T.M. Snyder, J.T. Dudley. CANDI: an R package and Shiny app for annotating radiographs and evaluating Computer-Aided Diagnosis. Bioinformatics. 2018.
  20. J. Zech, J. Forde, J. Titano, D. Kaji, A. Costa, E.K. Oermann. Detecting insertion, substitution, and deletion errors in radiology reports using neural sequence-to-sequence models. Annals of Translational Medicine. 2018.
  21. J. Titano, M. Badgeley, A. Costa, M. Pain, A. Su, N. Swinburne, M. Cai, J. Zech, J. Kim, J. Mocco, J. Bederson, J. Caridi, S. Cho, J. Lehar, E.K. Oermann. Automated surveillance of head CTs for acute neurologic events with deep neural networks. Nature Medicine. 2018.
  22. J. Zech, M. Pain, J. Titano, J. Schefflein, A. Su, M. Badgeley, E. Gordon, J. Bederson, J. Lehar, E.K. Oermann. Natural language based machine learning models for the annotation of clinical radiology reports. Radiology 287 (2), 570-580. 2018.
  23. J. Titano, D. Biederman, J. Zech, R. Korff, M. Ranade, R. Patel, E. Kim, F. Nowakowski, R. Lookstein, A. Fischman. Safety and feasibility of transradial access in patients with elevated INR. Journal of Vascular and Interventional Radiology. 2018.
  24. J. Zech, G. Husk, T. Moore, J.S. Shapiro. Measuring the degree of unmatched patient records in a health information exchange using exact matching. Applied Clinical Informatics 7 (02), 330-340. 2016.
  25. J. Zech, G. Husk, T. Moore, G.J. Kuperman, and J.S. Shapiro. Identifying homelessness using health information exchange data. Journal of the American Medical Informatics Association 22 (3), 682-687. 2015.

Code

childfx: Interactively explore pediatric upper extremity fracture detection model described in article published in Pediatric Radiology and validated in article published in Skeletal Radiology. Model training code available on github.

reproduce-chexnet: recreates the ChexNet model of Rajpurkar et al., and includes heatmaps to evaluate which regions of an image influenced predictions. Can be run entirely in your web browser using binder.

rad-report-annotator: allows you to annotate a large corpus of radiology reports using a small labeled subset using the methods described in our paper which appeared in Radiology.

Media

How Can Doctors Be Sure A Self-Taught Computer Is Making The Right Diagnosis? All Things Considered : NPR. 2019 April 1.

Digital assistants aid disease diagnosis. Nature: Outlook. 2019 September 25.

John Zech, Third Year Radiology Resident. Columbia Radiology. 2021 July 21.

John Zech, MD, MA Receives 2022 RSNA Resident/Fellow Research Grant. Columbia Radiology. 2022 June 7.

Your Donations In Action: John Zech, MD, MA. RSNA News. 2024 June 17.

The Promise and Perils of AI Medical Care. Bloomberg News. 2018 August 15.

AI brings scoliosis monitoring on x-rays into modern era. Aunt Minnie. May 1, 2024.

AI model can help diagnose pediatric buckle fractures. Aunt Minnie. December 5, 2021.