Routine clinical assessments for Parkinson's disease are essential instruments in both clinical practice and research, often used to identify disease sub-types and monitor the progression of disease severity. However, each clinic has limited access to information and the quality of these assessments is often degraded by the amount of missing information recorded at the time of each visit. The main objective of this study is to evaluate the performance of Federated Learning (FL) algorithms for imputing missing clinical data, enhancing the quality of decentralized Parkinson's disease assessments while maintaining data privacy. Specifically, we explore the impact of various aggregation strategies on the imputation of clinical data from 1,370 patients in the Parkinson Progression Marker Initiative (PPMI). Notably, the Cyclic Weight Transfer (CWT) algorithm stands out for its lower imputation errors. To validate this study, we conducted a downstream analysis using imputed data to predict symptoms progression. We observed that a FL-based approach yields superior model performance based on imputation errors, when compared to traditional learning strategies. These improvements can achieve 37.7% and 31.5% lower mean imputation errors with low and moderate degree of missing scores in the training data, respectively. In addition, we achieved better classification scores with Random Forest models trained with imputed data from FL-based approaches, compared to traditional statistical methods, with improvements of 0.5% in PR-AUC, 0.6% in ROC-AUC, and 1.3% in F-1 score. These results highlight FL as a robust and secure solution for decentralized clinical data management, offering improved performance while preserving patient privacy.

Recent advancements in vision-language models (VLMs), such as CLIP, have demonstrated substantial success in self-supervised representation learning for vision tasks. However, effectively adapting VLMs to downstream applications remains challenging, as their accuracy often depends on time-intensive and expertise-demanding prompt engineering, while full model fine-tuning is costly. This is particularly true for biomedical images, which, unlike natural images, typically suffer from limited annotated datasets, unintuitive image contrasts, and nuanced visual features. Recent prompt learning techniques, such as Context Optimization (CoOp) intend to tackle these issues, but still fall short in generalizability. Meanwhile, explorations in prompt learning for biomedical image analysis are still highly limited. In this work, we propose BiomedCoOp, a novel prompt learning framework that enables efficient adaptation of BiomedCLIP for accurate and highly generalizable few-shot biomedical image classification. Our approach achieves effective prompt context learning by leveraging semantic consistency with average prompt ensembles from Large Language Models (LLMs) and knowledge distillation with a statistics-based prompt selection strategy. We conducted comprehensive validation of our proposed framework on 11 medical datasets across 9 modalities and 10 organs against existing state-of-the-art methods, demonstrating significant improvements in both accuracy and generalizability. The code is publicly available at https://github.com/HealthX-Lab/BiomedCoOp.

Image registration is a fundamental medical image analysis task, and a wide variety of approaches have been proposed. However, only a few studies have comprehensively compared medical image registration approaches on a wide range of clinically relevant tasks. This limits the development of registration methods, the adoption of research advances into practice, and a fair benchmark across competing approaches. The Learn2Reg challenge addresses these limitations by providing a multi-task medical image registration data set for comprehensive characterisation of deformable registration algorithms. A continuous evaluation will be possible at https://learn2reg.grand-challenge.org. Learn2Reg covers a wide range of anatomies (brain, abdomen, and thorax), modalities (ultrasound, CT, MR), availability of annotations, as well as intra- and inter-patient registration evaluation. We established an easily accessible framework for training and validation of 3D registration methods, which enabled the compilation of results of over 65 individual method submissions from more than 20 unique teams. We used a complementary set of metrics, including robustness, accuracy, plausibility, and runtime, enabling unique insight into the current state-of-the-art of medical image registration. This paper describes datasets, tasks, evaluation methods and results of the challenge, as well as results of further analysis of transferability to new datasets, the importance of label supervision, and resulting bias. While no single approach worked best across all tasks, many methodological aspects could be identified that push the performance of medical image registration to new state-of-the-art performance. Furthermore, we demystified the common belief that conventional registration methods have to be much slower than deep-learning-based methods.

In brain tumor surgery, the quality and safety of the procedure can be impacted by intra-operative tissue deformation, called brain shift. Brain shift can move the surgical targets and other vital structures such as blood vessels, thus invalidating the pre-surgical plan. Intra-operative ultrasound (iUS) is a convenient and cost-effective imaging tool to track brain shift and tumor resection. Accurate image registration techniques that update pre-surgical MRI based on iUS are crucial but challenging. The MICCAI Challenge 2018 for Correction of Brain shift with Intra-Operative UltraSound (CuRIOUS2018) provided a public platform to benchmark MRI-iUS registration algorithms on newly released clinical datasets. In this work, we present the data, setup, evaluation, and results of CuRIOUS 2018, which received 6 fully automated algorithms from leading academic and industrial research groups. All algorithms were first trained with the public RESECT database, and then ranked based on a test dataset of 10 additional cases with identical data curation and annotation protocols as the RESECT database. The article compares the results of all participating teams and discusses the insights gained from the challenge, as well as future work.

Image guidance has become the standard of care for patient positioning in radiotherapy, where image registration is often a critical step to help manage patient motion. However, in practice, verification of registration quality is often adversely affected by difficulty in manual inspection of 3-D images and time constraint, thus affecting the therapeutic outcome. Therefore, we proposed to employ both bootstrapping and the supervised learning methods of linear discriminant analysis and random forest to help robustly assess registration quality in ultrasound-guided radiotherapy. We validated both approaches using phantom and real clinical ultrasound images, and showed that both performed well for the task. While learning-based techniques offer better accuracy and shorter evaluation time, bootstrapping requires no prior training and has a higher sensitivity.

Most strain imaging techniques follow a pipeline strategy: in the first step, tissue displacement is estimated from radio-frequency (RF) frames, and in the second step, a spatial derivative operation is applied. There are two main issues that arise from this framework. First, the gradient operation amplifies noise, and therefore, smoothing techniques have to be adopted. Second, strain estimation does not exploit the original RF data. It rather relies solely on the noisy displacement field. In this paper, a novel technique is proposed that utilizes both the displacement field and the RF frames to accurately obtain the strain estimates. The normalized cross correlation (NCC) metric between two corresponding windows around the samples of the pre- and post-compressed images is employed to generate a dissimilarity measurement. The derivative of NCC with respect to the strain is analytically derived using the chain rule. This allows an efficient minimization of the dissimilarity metric with respect to the strain using the gradient descent optimization technique. The effectiveness of the proposed method is investigated through simulation data, phantom experiments, and in vivo patient data. The experimental results show that exploiting the information in RF data significantly improves the strain estimates.

Susceptibility-weighted imaging (SWI) venography can produce detailed venous contrast and complement arterial dominated MR angiography (MRA) techniques. However, these dense reversed-contrast SWI venograms pose new segmentation challenges. We present an automatic method for whole-brain venous blood segmentation in SWI using Conditional Random Fields (CRF). The CRF model combines different first and second order potentials. First-order association potentials are modeled as the composite of an appearance potential, a Hessian-based shape potential and a non-linear location potential. Second-order interaction potentials are modeled using an auto-logistic (smoothing) potential and a data-dependent (edge) potential. Minimal post-processing is used for excluding voxels outside the brain parenchyma and visualizing the surface vessels. The CRF model is trained and validated using 30 SWI venograms acquired within a population of deep brain stimulation (DBS) patients (age range [Formula: see text] years). Results demonstrate robust and consistent segmentation in deep and sub-cortical regions (median kappa = 0.84 and 0.82), as well as in challenging mid-sagittal and surface regions (median kappa = 0.81 and 0.83) regions. Overall, this CRF model produces high-quality segmentation of SWI venous vasculature that finds applications in DBS for minimizing hemorrhagic risks and other surgical and non-surgical applications.