McWilliams School of Biomedical Informatics
banner
mcwlliamssbmi.bsky.social
McWilliams School of Biomedical Informatics
@mcwlliamssbmi.bsky.social
Educating leaders in health informatics since 1997.
Located in the Texas Medical Center at Houston, TX.
Pancreatic Cancer Prediction Using LLM Embeddings with Classifiers #AMIA25 #UTHealth #MSBMI

*ੈ✩‧༺☆༻*ੈ✩‧₊˚
🎉 Congratulations to Bingyu Mao 🎉
Winning 2nd Place
KDDM Innovation Award at AMIA 2025.
*ੈ✩‧˚༺☆༻*ੈ✩‧˚

Click to View Link to Poster:
workshopamia2025.github.io/AMIA-KDDM-20...
November 20, 2025 at 4:47 PM
Congratulations: Bingyu Mao

We are pleased to inform you that our former PhD student Bingyu Mao has won second place for the KDDM Innovation Award at AMIA 2025. Here is more details.

www.linkedin.com/posts/activi...
Sign Up | LinkedIn
500 million+ members | Manage your professional identity. Build and engage with your professional network. Access knowledge, insights and opportunities.
www.linkedin.com
November 20, 2025 at 4:27 PM
Well #AMIA2025 it was fun. Next year, you're in our neighborhood. We will show you a good time. Everything is bigger in Texas! @AMIAinformatics
November 19, 2025 at 9:55 PM
Continued work aims to expand these methods to new conditions, improve model precision, and address the regulatory steps needed to bring this technology into future care #MSBMI #UTHealth #TMC #AI #Machinelearning
November 19, 2025 at 6:42 PM
Zimolzak also highlighted the challenges of building diagnostic AI. Preparing structured data and ensuring quality inputs are essential steps for making these systems useful in clinical settings. #MSBMI #AI #Machinelearning #UTHealth #TMC
November 19, 2025 at 6:33 PM
Two-stage algorithms, combining rule-based triggers with machine learning, reached predictive values above 90 percent. This approach shows promise for identifying diagnostic errors earlier and improving patient safety. #MSBMI #AI #Machinelearning #UTHealth #TMC
November 19, 2025 at 6:26 PM
Manual review showed that these triggers identified missed opportunities in diagnosis nearly half the time. Machine learning models built on these findings further improved detection accuracy. #MSBMI #AI #Machinelearning #TMC #UTHealth
November 19, 2025 at 6:19 PM
Using Veterans Affairs data, this research developed electronic triggers to flag cases at risk for missed diagnoses, like patients discharged with stroke risk factors or abdominal pain who were later hospitalized. #MSBMI #TMC #UTHealth #AI #Machinelearning
November 19, 2025 at 6:09 PM
Reporting from the MSBMI research seminar!

We’re joined by Andrew Zimolzak, MD, MMSc, who is presenting: "Machine Learning to Enhance Electronic Detection of Diagnostic Errors" #MSBMI #Machinelearning #BiomedicalInformatics #Healthcare #UTHealth #TMC
November 19, 2025 at 6:02 PM
MSBMI faculty member Dr. Hui Li is speaking at the #AMIA2025 session 81.
November 18, 2025 at 8:04 PM
Temporal information is often fragmented across multiple screens or tables, complicating the tracking of diagnoses, medications, procedures, and test results over time for individual patients. #AMIA2025 #S81
November 18, 2025 at 7:52 PM
EHRs provide a comprehensive and up-to-date digital representation of a patient’s health data, encompassing a various elements such as demographics, vital signs, diagnoses, and lab results. #AMIA2025 #S81
November 18, 2025 at 7:51 PM
Heart disease is a major cause of death: the world’s biggest killer, 16% of deaths. 75% mortality for patients without intervention. #AMIA2025 #S81
November 18, 2025 at 7:38 PM
Rare disease facts: >7k different rare diseases, 400 million people affected, 95% of rare diseases have no approved treatment, 80% of rare diseases are of genetic origin, 75% of people affected are children, & >5 years of getting an accurate diagnosis. #AMIA2025 #S81
November 18, 2025 at 7:28 PM
Reposted by McWilliams School of Biomedical Informatics
Most pressing challenges facing informatics work at #AMIA2025
November 18, 2025 at 7:15 PM
Study shows that 3 out of 4 Americans do not trust AI in a healthcare setting. #AMIA2025 #S81
November 18, 2025 at 7:01 PM
Coming to you live from the #AMIA2025 #S81 ChronoLogic: Interpretable Modeling Across Patient Timelines.
November 18, 2025 at 7:00 PM
The MSBMI PhD program is 93 semester credit hours completed within 8 years. Flexible format, full or part-time, in-person or remote. About 2 years of courses with a written and oral qualifying exam. #AMIA2025 #S55
November 18, 2025 at 1:35 PM
Number of PhD students and candidates at MSBMI are 46! Average time to graduation is 5 years. Average salary for 1st year PhD grads is $105k and for 10th year, $258K. #AMIA2025 #S55
November 18, 2025 at 1:34 PM
PhD programs at Penn exist within “graduate groups”, not departments. Offers ability for students to take courses and experience mentorship across all 12 schools of the university. #AMIA2025 #S55
November 18, 2025 at 1:23 PM
Active learning as a pedagogical approach to the whole program: reverse classroom, cognitive engagement training throughout. #AMIA2025 #S55
November 18, 2025 at 1:22 PM
Early research and clinical rotations lead to program-long co-mentorship with AI expert + Clicinican working with the student on real-life health challenges in situ. #AMIA2025 #S55
November 18, 2025 at 1:21 PM
Industry demands for AI/ML skills and immediate impact draws students away form traditional doctoral training. #AMIA2025 #S55
November 18, 2025 at 1:19 PM