Medical Biometry and Statistics — IMBI (@IMBI_fr) header
Medical Biometry and Statistics — IMBI (@IMBI_fr) profile picture

Medical Biometry and Statistics — IMBI

@IMBI_fr

Our core tasks are methodological and applied biomedical research, from artificial intelligence to health services research, teaching and consulting.

 Freiburg im Breisgau, Germany
uniklinik-freiburg.de/imbi.html
Joined December 2018
  • Tweets93
  • Following34
  • Followers176
7Media
Photo by Medical Biometry and Statistics — IMBI (@IMBI_fr)Photo by Medical Biometry and Statistics — IMBI (@IMBI_fr)Photo by Medical Biometry and Statistics — IMBI (@IMBI_fr)Photo by Medical Biometry and Statistics — IMBI (@IMBI_fr)Photo by Medical Biometry and Statistics — IMBI (@IMBI_fr)Photo by Medical Biometry and Statistics — IMBI (@IMBI_fr)
We got the funding for our CRC Small Data!!! This means 31 open doctoral researcher positions for pushing the boundaries of artificial intelligence and modeling, starting on Oct 1st, 2023, thanks to the German Research Foundation. smalldata-initiative.de
Home - SmallData

Home - SmallData

Master Communication with Confidence On October 16, 2026, SMART hosted the course Master Communication with Confidence, led by Dr. Melchior Glatthard, a communication trainer specializing in rhetoric...

smalldata-initiative.de
2
9
22
4,222
Our joint paper in the Journal of Statistical Software has been published here: jstatsoft.org/article/view/v…
1
3
335
Congratulations to successfully funded Transregio CRC 359 "PILOT"! The RDM department of IMBI is involved with an infrastructure project and is looking forward to collaborating with all scientific projects in the CRC: kommunikation.uni-freiburg.d…
1
One more week to apply: #PhD position to work with @AdrianiNikol on a newly funded #DFG Emmy Noether Program. The project entails development of methods for generalizing definitions of populations, treatments, outcomes of interest in #EvidenceSynthesis. uniklinik-freiburg.de/imbi-e…
4
1
3
Our new paper shows that deep learning+ODEs for personalized medicine is feasible with a small number of individuals and time points, such as in rare disease registries. doi.org/10.1002/bimj.2020003…

Deep dynamic modeling with just two time points: Can we still allow for individual trajectories?

Longitudinal biomedical data are often characterized by a sparse time grid and individual-specific development patterns. Specifically, in epidemiological cohort studies and clinical registries we...

onlinelibrary.wiley.com
1
3
Our paper on JuliaConnectoR is online now. Get access to flexible Julia machine learning tools from R. jstatsoft.org/article/view/v…
1
2
Join our workshop on advanced single-cell methods and applications on Monday, March 14th, 16-17:30 (MET/Berlin/Paris time) via Zoom! For registration, visit: tinyurl.com/single-cell-gmds Talks by Vadim N. Gladyshev of @harvardmed and Joshua D. Welch (@LabWelch) of @UMich
3
4
Our paper on interpretable deep learning for single-cell gene expression data is online now. link.springer.com/article/10…

Interpretable generative deep learning: an illustration with single cell gene expression data

Human Genetics - Deep generative models can learn the underlying structure, such as pathways or gene programs, from omics data. We provide an introduction as well as an overview of such techniques,...

link.springer.com
5
4
Our tutorial paper on differentiable programming is online now — with an illustration how these techniques allowed us to quickly build our COVID-19 intensive care bed prediction models for the RKI in the first pandemic wave. tandfonline.com/doi/abs/10.1…
Using Differentiable Programming for Flexible Statistical Modeling

Using Differentiable Programming for Flexible Statistical Modeling

Differentiable programming has recently received much interest as a paradigm that facilitates taking gradients of computer programs. While the corresponding flexible gradient-based optimization app...

tandfonline.com
1
2
This is an instance where one would (nearly) wish that the prognoses from our (COVID-19) models (presentation of Lothar Wieler at 27:16) would not be right: youtube.com/watch?v=lH-gpVGJ…
Our new paper demonstrates considerable benefits when incorporating stratum information into neural networks with time-to-event data. Biostatistics+deep learning=success. academic.oup.com/bib/advance…

Stratified neural networks in a time-to-event setting

Abstract. Deep neural networks are frequently employed to predict survival conditional on omics-type biomarkers, e.g., by employing the partial likelihood

academic.oup.com
1
Harald Binder now is President of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS) @gmdsEV. Looking forward to joint initiatives of the data science disciplines in biomedicine.
5
10
Join our workshop on single cell data at @gmds_tmf_2021 on September 28th (jointly organized with Georg Fuellen). gmds-tmf-2021.de/events/tack…
3
5
We contributed our modeling expertise for a Cochrane review on non-medical COVID-19 prevention measures in long term care facilities. Available now. cochrane.org/CD015085/PUBHLT…
New paper from a collaboration showing how to combine machine learning ideas with causal inference for microbiome data. doi.org/10.1002/bimj.2019003…
1
1
The CRC 1453 has released a new website: sfb1453.uni-freiburg.de. We are participating in projects S1 and INF.
1
Our new paper provides a solution for generating synthetic data by deep learning for privacy-preserving distributed computing. bmcmedresmethodol.biomedcent…
Deep generative models in DataSHIELD

Deep generative models in DataSHIELD

BMC Medical Research Methodology - The best way to calculate statistics from medical data is to use the data of individual patients. In some settings, this data is difficult to obtain due to...

link.springer.com
1
3
Our new paper shows that machine learning/deep learning can help to automate updates of systematic reviews, but careful tuning is needed. onlinelibrary.wiley.com/doi/…
1
Maren Hackenberg (from our machine learning group) won the Bernd Streitberg Award of the German Region of the International Biometric Society with her work on neural networks and differential equations. Congratulations!
5
Daniela Zöller now is the new head of our Medical Data Science group. Looking forward to solutions for federated computing&more. uniklinik-freiburg.de/imbi/m…
2
11

No more items