Description: Air 11:Users:Bao:Desktop:VA-logo.png


RESEARCH AIM: Automation in Radiology.





Description: Pooter:Users:baodo:Desktop:1.jpg Description: Pooter:Users:baodo:Desktop:segond study:training_images:segonds:scvmc-export--317445581.jpg


“Expert” level perception of subtle x-ray findings using sparse data

A.I. automatic diagnosis of a subtle x-ray finding (Segond fracture, less than 0.04% of entire area of an image) using contextual interpretation, like a radiologist. Machine was 100% accurate (F-score 1.0) in our limited 20 test samples. Science poster presentation RSNA 2018.



Automatic chondroid bone tumor detection using deep learning

Can machines identify formless objects ie, “fluid” shaped tumors such as chondroid mass ? This is a simple patch based CNN that was slow and not particularly sensitive or specific.


Machine learning decision trees in radiology

Experiment using A.I. to automatically build expert decision trees from radiology observations to partition tumor classes. 100% accuracy in limited testing but broad application (classify 10+ tumor dxs, 15+ observations per dx) was plagued w/ over-fitting due to small sample size at this time.


IVC filter machine learning

Teach computer to recognize IVC filters on x-ray images and automatically track locations, potentially for decision support.



Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.29.02 PM.jpeg

Stanford bone diagnosis machine

Build a system to classify radiographic bone tumors using deep learning and Bayesian networks. Funded by Stanford-Philips Grant, 2018-2019.



Description: Pooter:Users:baodo:Desktop:unnamed.png


Chondroid machine

Adapted Stanford Bone Bayes engine to discriminate benign (enchondroma) VS malignant (chondrosarcoma) chondroid bone tumors. Preliminary performance exceeded >85% of radiologists. Do, BH, Beaulieu CF, Human VS Machine: Distinguishing Enchondroma from Chondrosarcoma with a Bayesian Network, Society of Skeletal Radiology meeting, 3/2018. Rec’d SSR “Man Vs Machine” Award, 2018.





Radiology management system to support clinical operations in the Veteran’s Health Administration. For example, to use machine learning for automatic protocol of exams.


Description: Air 11:Users:Bao:Desktop:Screen Shot 2017-08-14 at 10.36.23 AM.jpg



Deep convolutional neural net to identify bone joints of the human body. Foundation for PACS comparison protocols, automated decision systems such bone tumor diagnosis, information retrieval, QA, etc. Dr. Vossler rec’d a Student Travel Award for Young Investigators at the RSNA 2017 Scientific Meeting.


Description: Air 11:Users:Bao:Desktop:Screen Shot 2017-08-14 at 10.32.40 AM.jpg


Stanford Bone Bayes

A “learning” Bayesian network that models clinical and radiographic inputs to compute diagnosis, differential diagnosis, and probabilities. Current work involves training to learn “world” published literature and incorporate tumor registry statistics of bone features. PubMed. Dr. Langlotz presented at 2016 Scientific Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine (SIIM), Alexandria, VA.


Description: Air 11:Users:Bao:Desktop:Screen Shot 2017-08-14 at 10.48.21 AM.jpg


Co-founded secure cloud PACS that can be customized for clinical care, education, research, media reports. More than 100,000 users have now viewed > 10 million images on clariPACS.




Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.30.55 PM.jpeg



Community of radiologists promoting the exchange of high level musculoskeletal imaging expertise. Searchable gallery of common and rare musculoskeletal disorders for decision support, teaching. Founded by Philip Tirman MD (2003). Served > 1 million images since 2010.



Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.52.00 PM.jpeg


Stanford Atlas of Common MSK Measurements

RadLex-enabled atlas of common MSK measurements. Designed to leverage RadLex ID interoperability for automation systems. RSNA Scientific Meeting, 2010.



Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.36.57 PM.jpeg

ACGME Analytics

To eliminate the labor intensive task of counting Radiology residency workloads at VA centers by mining VistA to generate an ACGME compliant summary of workload by CPTs. The Accreditation Council for Graduate Medical Education (ACGME) requires bi-annual case log statistical reports by Residency programs for all medical specialties. The Residency Review Committee (RRC) for Radiology recognizes 90 CPT codes in 11 categories: chest radiographs, CT body, CT angiography, image guided biopsy and drain, mammography, MRI body, MRI brain, MRI knee, PET/CT, ultrasound pelvis, and MRI spine.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.46.21 PM.jpeg

Hedge Detector

Automatic extraction of uncertainty and recommendation concepts from unstructured radiology reports. Applications include for QA, follow-up, peer review, disease surveillance. Rec’d RSNA Research Trainee Prize for Scientific Paper, RSNA Scientific Meeting, 2009.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.47.52 PM.jpeg

VA Radiology Analytics

Mine and summarize operations metrics for VA Hospitals using VistA radiology data. Common metrics to help identify bottlenecks in radiology clinical operations at Veteran hospitals.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.49.17 PM.jpeg


Stanford MSK MRI Atlas

Free MRI atlas of musculoskeletal radiology.

Has served > 600,000 pages & > 20 million+ images to users from 100+ countries since 2011. RSNA Scientific Meeting, 2010.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.54.29 PM.jpeg



Radiology report search engine, originally indexed nearly 1,000,000 cases in Stanford radiology. Shared for free to numerous academic institutions, hospitals, and individuals around the world, including deployment at Hospital de Pediatria J. P. Garrahan, a children’s hospital in Buenos Aires, Argentina (Darío Filippo MD), where it enables search of over 250,000 radiology, pathology, and surgery notes. Interesting behaviors in medical imaging search. RSNA Scientific Meeting, 2009.




Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 1.56.46 PM.jpeg



Audience response system w/ MCQ style interactive polling for Stanford radiology lectures.



Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.04.30 PM.jpeg


NLP to structure MSK MRI knee reporting

Auto structured knee MRI reports from free dictation in real time.

RSNA Scientific Meeting, 2010.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.10.04 PM.jpeg


NLP to detect missing concepts in MSK tumor reporting

May augment self learning systems via extraction of ground truth correlation of RIS/PATHOLOGY. RSNA Scientific Meeting, 2012.

Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.11.57 PM.jpeg

NLP to automate quality assurance using RIS reports - Fractures

Fully automatic extraction of accuracy statistics from unstructured radiology report data using NLP to compare knee x-ray reads against the follow-up MRI/CT reports as ground truth. QA, education, teaching, peer review. RSNA Scientific Meeting, 2011.



Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.16.23 PM.jpeg


NLP to automate quality assurance using RIS reports - Osteoporosis

Automatically extract accuracy metrics for osteoporosis from radiology reports in the RIS, using NLP to compare against the follow-up DEXA reports, respectively, as ground truth. RSNA Scientific Meeting, 2011.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.14.25 PM.jpeg


NLP to derive central line usage from unstructured ICU x-ray reports

Using NLP to build an ICU patient consensus and central line indwelling estimate from unstructured chest x-ray reports.



Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 2.17.49 PM.jpeg


Feedback NLP of fractures in unstructured reports of ED studies

Real-time retrieval of fracture classifications and clinical pearls extracted from free dictation for decision support. Rec’d RSNA Research Trainee Prize for Scientific Paper, RSNA Scientific Meeting, 2007.


Description: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 3.31.25 PM.jpegDescription: Pooter:Users:baodo:Desktop:Screen Shot 2017-11-08 at 3.31.19 PM.jpeg


Characterizing radiology search patterns using web analytics

System to analyze how radiologists review studies using heat maps of mouse movement and magnification patterns, deployed completely via the web. Note how experienced, attending radiologists (LEFT) require fewer mouse movements and magnification compared to trainee radiology residents (RIGHT). Deployed in online simulator of call cases for residents. RSNA Scientific Meeting, 2008.







Available on request.


1. Bone joint radiographs. Full resolution PNG of major appendicular joints; 793 exams, 13 class labels, pre-processed using contrast limited adapted histogram equalization. About: RSNA 2017.


2. Musculoskeletal MRI anatomy. Comprehensive collection of MRI images, in all 3 planes, of major joints of the body (hip, knee, ankle, etc). Expert labeled annotations of RadLex structure, RID, & spatial coordinate. Link.


3. Musculoskeletal disease collection. > 1,000 cases of various MSK diseases, of all Radiology modalities, indexed by joint, history, diagnosis (+/- RID), discussion, and references. Link.





Stanford Radiology Residency Informatics Day 2018: A.I., November 8, 2018

SLIDES: Build your own AI to predict tumors !

FILES: source code | header data | training data | testing data

IDE: Python interpreter




Non-technical introduction to deep learning: Deep Learning in a Nutshell.


Chartrand G. et al. Deep Learning: A Primer for Radiologists. Radiographics. 2017 Nov-Dec;37(7):2113-2131. doi: 10.1148/rg.2017170077.


Erickson B. et al. Machine Learning for Medical Imaging. Radiographics. 2017 Mar-Apr;37(2):505-515. doi: 10.1148/rg.2017160130. Epub 2017 Feb 17.


Neural Networks: All YOU Need to Know


8 Neural Network Architectures Machine Learning Researchers Need to Learn.



Introduction to CNNs


Introduction to FCNs and Practical deep CNN hands on


Object detection review


Limited data thoughts Part 2


Metrics for evaluation


When Not to Use Deep Learning


Bonus: Deep learning basketball – link 1, link 2


Description: Pooter:Users:baodo:Desktop:Screen Shot 2018-04-25 at 12.33.23 PM.jpeg


Free machine learning textbooks





Cai T et al. Natural Language Processing Technologies in Radiology Research and Clinical Applications. Radiographics. 2016 Jan-Feb;36(1):176-91. doi: 10.1148/rg.2016150080.


Classification using logistic regression link


Is Regression Analysis Really Machine Learning ?





1. 100 days of ML coding “curriculum” by Avik Jain ( with awesome infographics ).


2. Hypothesis driven machine learning project in Radiology in collaboration with a clinical advisor.





Bao Do, MD

Clinical Associate Professor (Affiliated), Dept. of Radiology, Stanford University

Affiliated Faculty, Integrative Biomedical Imaging Informatics Section (IBIIS), Stanford University

Member, Radiological Society North America

Email: baodo @





2011, Musculoskeletal Fellowship, Stanford University

2010, Diagnostic Radiology Residency, Stanford University

2008, Diagnostic Radiology Residency, University of Iowa

2005, University of Illinois College of Medicine





Precision Diagnosis of Bone Tumors with Advanced Semantic Modeling and Radiomics Analysis

Co-PI: Christopher Beaulieu, MD PhD

Stanford-Philips Grant, 2018-2019


HONORS AND AWARDS:                


2005   Research Trainee Prize for Scientific Paper, RSNA

2007   Research Trainee Prize for Scientific Paper, RSNA

2009   Research Trainee Prize for Scientific Paper, RSNA

2010   Roentgen Resident Research Award, RSNA, Stanford University

2011   Fellow Teaching Award, Stanford Hospital and Clinics

2012   Clinical Educator of the Year, Stanford Hospital and Clinics





Yeluri, V., Arlagada, V.K., Do, B., Beaulieu, CF. Systems and methods for natural language processing to provide smart links in radiology reports. U.S. patent # US 2014/0006926 A1





Horvath, R., Scharfe, C., Hoeltzenbein, M., Do BH, Schröder, C., Warzok, R., Vogelgesang, S., Lochmüller, H., Müller-Höcker, J., Gerbitz, KD, Oefner, PJ, and Jaksch, M. Childhood Onset Mitochondrial Myopathy and Lactic Acidosis Caused by a Stop Mutation in the Mitochondrial Cytochrome C Oxidase III Gene. J of Med Gen, Nov;39(11):812-6.


Horvath, R., Lochmuller, H., Scharfe, C., Do BH, Oefner, PJ., Muller-Hocker, J., Schoser, BG, Pongratz, D., Auer, DP, Jaksch, M.  A tRNA (Ala) Mutation Causing Mitochondrial Myopathy Clinically Resembling Myotonic Dystrophy. J Med Genetics. 2003 Oct;40(10):752-757.


Do B, Lossos, IS, Thorstenson, Y., Oefner, PJ, and Levy, R. Analysis of FAS (CD95) Gene Mutations in Higher Grade Transformation of Follicle Cell Lymphoma. Leukemia Lymphoma 2003 Aug;44(8):1317-23.


Sivakumaran TA, Shen P, Wall DP, Do BH, Kucheria K, Oefner PJ. Conservation of the RB1 Gene in Human and Primates. Hum Mutat. 2005 Apr;25(4):396-409.


Kivisild T, Shen P, Wall DP, Do B, Sung R, Davis KK, Passarino G, Underhill PA, Scharfe C, Torroni A, Scozzari R, Modiano D, Coppa A, de Knjiff P, Feldman MW, Cavalli-Sforza LL, Oefner PJ. The Role of Selection in the Evolution of Human Mitochondrial Genomes. Genetics. 2006 Jan;172(1):373-87.


Do B, Mari C, Biswal S, Kalinyack J, Gambhir SS. Diagnosis of Aseptic Deep Venous Thrombosis of the Upper Extremity in a Cancer Patient Using Fluorine-18 Fluorodeoxyglucose Positron Emission Tomography/CTomography (FDG PET/CT): A case report. Ann Nucl Med. 2006 Feb;20(2):151-5.


Wu, A. S., Do, B. H., Kim, J., & Rubin, D. L. (2009). Evaluation of Negation and Uncertainty Detection and its Impact on Precision and Recall in Search. Journal of Digital Imaging, 24(2), 234–242.


Do, B. H., Wu, A., Biswal, S., Kamaya, A., & Rubin, D. L. (2010). Informatics in Radiology: RADTF: A Semantic Search–enabled, Natural Language Processor–generated Radiology Teaching File. RadioGraphics, 30(7), 2039–2048.


Do, B. H., Mari, C., Tseng, J. R., Quon, A., Rosenberg, J., & Biswal, S. (2011). Pattern of 18F-FDG Uptake in the Spinal Cord in Patients With Non-Central Nervous System Malignancy. Spine, 36(21), E1395–E1401.


Do, B. H., Wu, A. S., Maley, J., & Biswal, S. (2012). Automatic Retrieval of Bone Fracture Knowledge Using Natural Language Processing. Journal of Digital Imaging, 26(4), 709–713.


Chu, C. R., Williams, A. A., West, R. V., Qian, Y., Fu, F. H., Do, B. H., & Bruno, S. (2014). Quantitative Magnetic Resonance Imaging UTE-T2* Mapping of Cartilage and Meniscus Healing After Anatomic Anterior Cruciate Ligament Reconstruction. The American Journal of Sports Medicine, 42(8), 1847–1856.


Boas, F. E., Kamaya, A., Do, B., Desser, T. S., Beaulieu, C. F., Vasanawala, S. S., et al. (2014). Classification of Hypervascular Liver Lesions Based on Hepatic Artery and Portal Vein Blood Supply Coefficients Calculated from Triphasic CT Scans. Journal of Digital Imaging, 28(2), 213–223.


Boas, F. E., Do, B., Louie, J. D., Kothary, N., Hwang, G. L., Kuo, W. T., et al. (2015). Optimal Imaging Surveillance Schedules after Liver-Directed Therapy for Hepatocellular Carcinoma. Journal of Vascular and Interventional Radiology, 26(1), 69–73.


Gutierrez, L. B., Do, B. H., Gold, G. E., Hargreaves, B. A., Koch, K. M., Worters, P. W., & Stevens, K. J. (2015). MR Imaging Near Metallic Implants Using MAVRIC SL. Academic Radiology, 22(3), 370–379.


De-Arteaga, M., Eggel, I., Do, B., Rubin, D., Kahn, C. E., Jr., & Müller, H. (2015). Comparing image search behaviour in the ARRS GoldMiner search engine and a clinical PACS/RIS. Journal of Biomedical Informatics, 56, 57–64.


Do, BH, Langlotz, C, Beaulieu, CF. Bone Tumor Diagnosis Using a Naēve Bayesian Model of Demographic and Radiographic Features. Journal of Digitial Imaging. 2017 Jul 27. doi: 10.1007/s10278-017-0001-7.


Chu CR, Sheth S, Erhart-Hledik JC, Do B, Titchenal MR, Andriacchi TP. Mechanically stimulated biomarkers signal cartilage changes over 5 years consistent with disease progression in medial knee osteoarthritis patients. J Orthop Res. 2017 Sep 1. doi: 10.1002/jor.23720.


Zhou, X., Cipriano, P., Kim, B., Dhatt, H., Rosenberg, J., Mittra, E., Do, B., Graves, E., Biswal, S. Detection of nociceptive-related metabolic activity in the spinal cord of low back pain patients using (18)F-FDG PET/CT. Scandinavian journal of pain. 2017; 15: 53–57


Banerjee, I., Kurtz, C., Edward Devorah, A., Do, B., Rubin, D. L., Beaulieu, C. F. Relevance Feedback for Enhancing Content Based Image Retrieval and Automatic Prediction of Semantic Image Features: Application to Bone Tumor Radiographs. Journal of biomedical informatics. 2018.


Williams, A. A., Titchenal, M. R., Do, B. H., Guha, A., Chu, C. R.. MRI UTE-T2* Shows High Incidence of Cartilage Subsurface Matrix Changes 2 Years After ACL Reconstruction. Journal of Orthopaedic Oesearch. 2018.