Anne-Christin Hauschild

Research profile

Research summary

Anne-Christin Hauschild is a researcher at Justus-Liebig-Universität Gießen, Germany, working across medical informatics, medical image analysis, and computational pathology, with federated machine learning as the connecting methodological thread. Her OpenAlex-linked profile records an h-index of 26, an i10-index of 41, 41 works over the last three years, and 26 distinct co-authors. Annual output has climbed steeply, from 8 works in 2020 to 13 in 2022, 16 in 2024, and 16 again in 2026, supporting a maximal activity score. Her topical footprint is unusually broad: machine learning in healthcare (25 works), advanced chemical sensor technologies (24), metabolomics and mass spectrometry (17), artificial intelligence in healthcare (17), and bioinformatics and genomic networks (16), tracing an arc from early sensor and metabolomics research to clinical AI. Her field distribution now splits evenly across federated machine learning in oncology, medical image analysis with federated learning, and computational pathology. Three 2026 arXiv papers anchor the Evolve catalog: the FAIrPaCT federated pancreatic-cancer framework update (first author), FedDRAW's annealed federated aggregation for chest radiographs (last author), and FFM-CP's vision-language model fusion for few-shot pathology (last author). Across these, Evolve's reproducibility mean is 43, with significance testing and missing baselines flagged as gaps.

Research keywords

Highlights

• FedDRAW: annealed federated aggregation for heterogeneous chest radiographs

  As last author, she published a 2026 study introducing dual reputation annealing weighting that shifts federated aggregation from client-size priors toward similarity weighting and finally uniform averaging, achieving the best average rank across 12 non-IID chest radiograph scenarios, though Evolve notes the FedAvg edge is not pairwise-significant.

• FFM-CP: fusing vision-language foundation models for few-shot pathology

  Last-authored 2026 work that aligns pathology vision-language backbones via Orthogonal Procrustes alignment and a graph neural network for few-shot histopathology, winning 50 of 54 comparisons; Evolve flags missing significance tests and a naive-ensemble baseline.

• FAIrPaCT: operational lessons from a federated pancreatic cancer framework

  First-authored 2026 strategy update on the FAIrPaCT federated AI framework for optimizing pancreatic cancer treatment, covering privacy-preserving machine learning, oBDS-based data harmonization, and multi-site operations; Evolve values the operational lessons but notes no federated model performance results yet.

Timeline

1. 2009Earliest publication year in the OpenAlex counts-by-year series, with 1 work recorded, marking the start of the visible record.
2. 2014An early peak of 10 works in a single year according to the counts-by-year series.
3. 2022Annual output reached 13 works, continuing a climb from 8 in 2020 that signals the shift into medical AI.
4. 202416 works recorded in a single year, the highest annual count in the series.
5. 2026Three first/last-authored arXiv papers enter the Evolve catalog: FAIrPaCT (first author), FedDRAW and FFM-CP (last author), while the series again records 16 works.

Papers in the Evolve catalog

• 2026-09-23

  FFM-CP: Cross-Backbone Fusion of Vision-Language Foundation Models for Few-Shot Computational Pathology

• 2026-09-21first author

  A Federated Artificial Intelligence Framework for Optimizing Pancreatic Cancer Treatment - Strategy Update

• 2026-09-04

  FedDRAW: Federated Dual Reputation Annealing Weighting for Heterogeneous Multi-Institutional Chest Radiograph Classification

Frequent co-authors