The study evaluates 15 frozen hematology foundation-model embeddings across four single‑cell acquisition domains, finding that while in‑domain accuracy is near‑saturated (macro‑F1 0.98–0.997), cross‑dataset performance drops dramatically (34–72%) and model rankings shift. Probe‑dependent rank transfer is observed, with 1‑NN retrieval more stable than linear heads, yet neither reliably predicts target robustness. Calibration deteriorates off‑domain (ECE rises from 0.004 to 0.35), and exposure to internal cohorts confounds shift analysis; a training‑free pseudo‑label‑balanced feature normalization (CBR) modestly improves target‑prior robustness and calibration.
whyItMatters":"The findings highlight that frozen hematology foundation models, though accurate in‑domain, may fail under realistic scanner, site, and class‑prior shifts, underscoring the need for comprehensive audits of accuracy, calibration, exposure, and robustness before clinical deployment."
By Jai Kumar Sharma, Peeyush Tapadiya
arXiv:2608.22059v1 Announce Type: cross
Abstract: Pretrained image encoders are central to medical image classification, where expert annotation is costly and task-specific cohorts are often limited....
By Xingtao Lin, Hangqi Ren, Caiwan Sun, You Chen
HERO (Histology Encoder for Robust Representation in Oncology) is a ViT‑G/14 pathology foundation model trained with DINO and iBOT objectives and refined using high‑resolution Gram anchoring on a 500‑million‑tile corpus from about 575,000 clinical whole‑slide images. It demonstrates superior robustness to center, scanner, and stain variation compared to other state‑of‑the‑art foundation models, while maintaining competitive performance on tile‑level classification, segmentation, and gene‑expression prediction. Across 39 slide‑level clinical tasks, HERO ranks first on average and achieves the best average rank across six benchmark frameworks under an equal‑weighted analysis.
By Zhi Li (Caris Life Sciences, Irving, TX, United States), Eghbal Amidi (Caris Life Sciences, Irving, TX, United States), Yating Cheng (Caris Life Sciences, Irving, TX, United States), Tyson Dawson (Caris Life Sciences, Irving, TX, United States), Gorkem Can Ates (Caris Life Sciences, Irving, TX, United States), Shuzhen Kuang (Caris Life Sciences, Irving, TX, United States), Norsang Lama (Caris Life Sciences, Irving, TX, United States), Md Ashequr Rahman (Caris Life Sciences, Irving, TX, United States), Zhiying Lu (Caris Life Sciences, Irving, TX, United States), Elisabeth K. Kong (Caris Life Sciences, Irving, TX, United States), Milan Radovich (Caris Life Sciences, Irving, TX, United States), David Spetzler (Caris Life Sciences, Irving, TX, United States), Matthew Oberley (Caris Life Sciences, Irving, TX, United States), George W. Sledge (Caris Life Sciences, Irving, TX, United States), Ming Chen (Caris Life Sciences, Irving, TX, United States)
Lumen is a pathology vision‑language model that aligns frozen unimodal foundation models (Virchow2 and BioMedBERT) using rank‑4 adapters and projection heads, training only 0.40% of the total parameters on the QUILT‑1M corpus. It achieves the highest mean chance‑corrected balanced accuracy (0.546) across nine zero‑shot patch benchmarks and demonstrates strong performance on lymph‑node metastasis detection, with AUROC scores of 0.964 internally and 0.955 externally. While it ranks third in cross‑modal retrieval, Lumen’s low‑parameter training yields competitive results at both patch and slide levels.
By Kiarash Tajbakhsh, Abdelrahman Faqieh, Michael Jopiti, Javier Garcia-Baroja, Philipp Zens, Branislav Zagrapan, Yuri Tolkach, Martin D. Berger, Aurel Perren, Bastian Dislich, Inti Zlobec, Amjad Khan
arXiv:2608. 10657v1 Announce Type: cross Abstract: Leukemia cell image classification is challenged by real-world domain shifts from acquisition, staining, illumination, and site protocols, causing single-dataset models to generalize poorly in real clinical scenarios.
By Carlos Zamora, Hiram Zuniga, Ulises Orozco-Rosas, Kenia Picos
Converting free-text radiology reports into structured labels supports cohort building, quality assurance, and monitoring of clinical imaging models, but the strongest label extractors are hosted prop...
The study evaluates whether a fine‑tuned open‑weight model (Gemma‑3‑12B) can match the performance of GPT‑4o in extracting multi‑label intracranial hemorrhage acuity from non‑contrast head‑CT reports. Using a 2×2 design that varied adaptation strategy (classification head vs. instruction fine‑tuning) and training‑data source (distilled real GPT‑4o labels vs. synthetic GPT‑4o‑generated reports), the distilled instruction‑tuned model achieved macro‑F1 scores comparable to GPT‑4o and surpassed the untuned base model. The key finding is that the source of training data—distilled real reports—was more important than the fine‑tuning method, and that the entire fine‑tuning and inference process fits on a single 24 GB consumer GPU.
By Aawez Mansuri, Kush Mehta, Mohammadreza Chavoshi, Jahanzaib Malik, Theodorus Dapamede, Frank Li, Rohan Isaac, Beatrice Brown-Mulry, Chiratidzo Rudado Sanyika, YoungSeok Jeon, Judy W. Gichoya, Ali Emami, Hari Trivedi
arXiv:2607. 24519v2 Announce Type: replace Abstract: Pretrained EEG foundation models are proposed for clinical decoding, but whether reported gains transfer across populations or survive negative controls is unclear.
By Marzieh Zare
An interpretable multi‑instance learning classifier based on a decision tree was developed to predict NPM1 and FLT3‑ITD mutations in acute myeloid leukemia using routine flow cytometry data. In cross‑validation on 197 patients, the model achieved AUROCs of 0.96 for NPM1 and 0.86 for FLT3‑ITD, outperforming a clinical baseline and matching deep learning methods. On an independent cohort of 161 patients, it maintained high performance with AUROCs of 0.90 and 0.82, and positive predictive values of 0.87 and 0.68, while cell‑level interpretation recovered known immunophenotypic signatures.
By Jonathan Legrand (IMB, MONC), Aguirre Mimoun (CHU Bordeaux), Baudouin Denis de Senneville (IMB, MONC), Audrey Bidet (CHU Bordeaux), Pierre-Yves Dumas (CHU Bordeaux, Inserm U1312 - BRIC), Christ\`ele Etchegaray (MONC, IMB)
The study evaluates a deep‑learning model for pediatric pneumonia detection across chest X‑ray datasets from three countries, assessing discrimination, calibration, operating‑point transport, shortcut signals, and limited‑label recovery. Using a frozen DenseNet121 ensemble trained on Guangzhou data, the model achieved high internal AUROC (0.976) but performance dropped when applied zero‑shot to Bangladesh (AUROC 0.798) and Vietnam (AUROC 0.742). Limited‑label adaptation with Platt recalibration restored sensitivity but introduced significant specificity variability, highlighting the need to evaluate multiple performance dimensions in cross‑dataset transport studies.
By Nazim-E-Alam
arXiv:2607. 00385v2 Announce Type: replace-cross Abstract: Automated malaria diagnosis from blood smear microscopy is a critical global health AI challenge; expert scarcity remains the primary diagnostic bottleneck.
By Kaysarul Anas Apurba, Md Hasibul Hasan, Mohammed Ali, Tanzilur Rahman
The study evaluates a deep‑learning model for pediatric pneumonia detection across three countries, testing not only discrimination but also probability calibration, fixed operating‑point transport, shortcut signals, and limited‑label recoverability. Using a DenseNet121 ensemble trained on Guangzhou data, the model achieved high internal AUROC (0.976) but performance dropped to 0.798 and 0.742 on Bangladeshi and Vietnamese datasets, respectively. Limited‑label adaptation with Platt recalibration restored sensitivity but introduced significant specificity variability, highlighting the need for comprehensive cross‑dataset evaluation.