Anindo Saha

Matthias Eisenmann, Annika Reinke, Vivienn Weru et al. · utoronto

The number of international benchmarking competitions is steadily increasing in various fields of machine learning (ML) research and practice. So far, however, little is known about the common practice as well as bottlenecks faced by the community in tackling the research questions posed. To shed light on the status quo of algorithm development in the specific field of biomedical imaging analysis, we designed an international survey that was issued to all participants of challenges conducted in conjunction with the IEEE ISBI 2021 and MICCAI 2021 conferences (80 competitions in total). The survey covered participants' expertise and working environments, their chosen strategies, as well as algorithm characteristics. A median of 72% challenge participants took part in the survey. According to our results, knowledge exchange was the primary incentive (70%) for participation, while the reception of prize money played only a minor role (16%). While a median of 80 working hours was spent on method development, a large portion of participants stated that they did not have enough time for method development (32%). 25% perceived the infrastructure to be a bottleneck. Overall, 94% of all solutions were deep learning-based. Of these, 84% were based on standard architectures. 43% of the respondents reported that the data samples (e.g., images) were too large to be processed at once. This was most commonly addressed by patch-based training (69%), downsampling (37%), and solving 3D analysis tasks as a series of 2D tasks. K-fold cross-validation on the training set was performed by only 37% of the participants and only 50% of the participants performed ensembling based on multiple identical models (61%) or heterogeneous models (39%). 48% of the respondents applied postprocessing steps.

Bram de Wilde, Anindo Saha, Maarten de Rooij et al.

Diffusion models for text-to-image generation, known for their efficiency, accessibility, and quality, have gained popularity. While inference with these systems on consumer-grade GPUs is increasingly feasible, training from scratch requires large captioned datasets and significant computational resources. In medical image generation, the limited availability of large, publicly accessible datasets with text reports poses challenges due to legal and ethical concerns. This work shows that adapting pre-trained Stable Diffusion models to medical imaging modalities is achievable by training text embeddings using Textual Inversion. In this study, we experimented with small medical datasets (100 samples each from three modalities) and trained within hours to generate diagnostically accurate images, as judged by an expert radiologist. Experiments with Textual Inversion training and inference parameters reveal the necessity of larger embeddings and more examples in the medical domain. Classification experiments show an increase in diagnostic accuracy (AUC) for detecting prostate cancer on MRI, from 0.78 to 0.80. Further experiments demonstrate embedding flexibility through disease interpolation, combining pathologies, and inpainting for precise disease appearance control. The trained embeddings are compact (less than 1 MB), enabling easy data sharing with reduced privacy concerns.

Annika Reinke, Evangelia Christodoulou, Sthuthi Sadananda et al.

Benchmarking competitions are central to the development of artificial intelligence (AI) in medical imaging, defining performance standards and shaping methodological progress. However, it remains unclear whether these benchmarks provide data that are sufficiently representative, accessible, and reusable to support clinically meaningful AI. In this work, we assess fairness along two complementary dimensions: (1) whether challenge datasets are representative of real-world clinical diversity, and (2) whether they are accessible and legally reusable in line with the FAIR principles. To address this question, we conducted a large-scale systematic study of 241 biomedical image analysis challenges comprising 458 tasks across 19 imaging modalities. Our findings show substantial biases in dataset composition, including geographic location, modality-, and problem type-related biases, indicating that current benchmarks do not adequately reflect real-world clinical diversity. Despite their widespread influence, challenge datasets were frequently constrained by restrictive or ambiguous access conditions, inconsistent or non-compliant licensing practices, and incomplete documentation, limiting reproducibility and long-term reuse. Together, these shortcomings expose foundational fairness limitations in our benchmarking ecosystem and highlight a disconnect between leaderboard success and clinical relevance.

Matin Hosseinzadeh, Anindo Saha, Joeran Bosma et al.

Quality of deep convolutional neural network predictions strongly depends on the size of the training dataset and the quality of the annotations. Creating annotations, especially for 3D medical image segmentation, is time-consuming and requires expert knowledge. We propose a novel semi-supervised learning (SSL) approach that requires only a relatively small number of annotations while being able to use the remaining unlabeled data to improve model performance. Our method uses a pseudo-labeling technique that employs recent deep learning uncertainty estimation models. By using the estimated uncertainty, we were able to rank pseudo-labels and automatically select the best pseudo-annotations generated by the supervised model. We applied this to prostate zonal segmentation in T2-weighted MRI scans. Our proposed model outperformed the semi-supervised model in experiments with the ProstateX dataset and an external test set, by leveraging only a subset of unlabeled data rather than the full collection of 4953 cases, our proposed model demonstrated improved performance. The segmentation dice similarity coefficient in the transition zone and peripheral zone increased from 0.835 and 0.727 to 0.852 and 0.751, respectively, for fully supervised model and the uncertainty-aware semi-supervised learning model (USSL). Our USSL model demonstrates the potential to allow deep learning models to be trained on large datasets without requiring full annotation. Our code is available at https://github.com/DIAGNijmegen/prostateMR-USSL.

Joeran S. Bosma, Anindo Saha, Matin Hosseinzadeh et al.

Deep learning-based diagnostic performance increases with more annotated data, but large-scale manual annotations are expensive and labour-intensive. Experts evaluate diagnostic images during clinical routine, and write their findings in reports. Leveraging unlabelled exams paired with clinical reports could overcome the manual labelling bottleneck. We hypothesise that detection models can be trained semi-supervised with automatic annotations generated using model predictions, guided by sparse information from clinical reports. To demonstrate efficacy, we train clinically significant prostate cancer (csPCa) segmentation models, where automatic annotations are guided by the number of clinically significant findings in the radiology reports. We included 7,756 prostate MRI examinations, of which 3,050 were manually annotated. We evaluated prostate cancer detection performance on 300 exams from an external centre with histopathology-confirmed ground truth. Semi-supervised training improved patient-based diagnostic area under the receiver operating characteristic curve from $87.2 \pm 0.8\%$ to $89.4 \pm 1.0\%$ ($P<10^{-4}$) and improved lesion-based sensitivity at one false positive per case from $76.4 \pm 3.8\%$ to $83.6 \pm 2.3\%$ ($P<10^{-4}$). Semi-supervised training was 14$\times$ more annotation-efficient for case-based performance and 6$\times$ more annotation-efficient for lesion-based performance. This improved performance demonstrates the feasibility of our training procedure. Source code is publicly available at github.com/DIAGNijmegen/Report-Guided-Annotation. Best csPCa detection algorithm is available at grand-challenge.org/algorithms/bpmri-cspca-detection-report-guided-annotations/.

Alessa Hering, Sarah de Boer, Anindo Saha et al.

The PI-CAI (Prostate Imaging: Cancer AI) challenge led to expert-level diagnostic algorithms for clinically significant prostate cancer detection. The algorithms receive biparametric MRI scans as input, which consist of T2-weighted and diffusion-weighted scans. These scans can be misaligned due to multiple factors in the scanning process. Image registration can alleviate this issue by predicting the deformation between the sequences. We investigate the effect of image registration on the diagnostic performance of AI-based prostate cancer diagnosis. First, the image registration algorithm, developed in MeVisLab, is analyzed using a dataset with paired lesion annotations. Second, the effect on diagnosis is evaluated by comparing case-level cancer diagnosis performance between using the original dataset, rigidly aligned diffusion-weighted scans, or deformably aligned diffusion-weighted scans. Rigid registration showed no improvement. Deformable registration demonstrated a substantial improvement in lesion overlap (+10% median Dice score) and a positive yet non-significant improvement in diagnostic performance (+0.3% AUROC, p=0.18). Our investigation shows that a substantial improvement in lesion alignment does not directly lead to a significant improvement in diagnostic performance. Qualitative analysis indicated that jointly developing image registration methods and diagnostic AI algorithms could enhance diagnostic accuracy and patient outcomes.

Anindo Saha, Joeran S. Bosma, Jasper J. Twilt et al.

In this intercontinental, confirmatory study, we include a retrospective cohort of 22,481 MRI examinations (21,288 patients; 46 cities in 22 countries) to train and externally validate the PI-CAI-2B model, i.e., an efficient, next-generation iteration of the state-of-the-art AI system that was developed for detecting Gleason grade group $\geq$2 prostate cancer on MRI during the PI-CAI study. Of these examinations, 20,471 cases (19,278 patients; 26 cities in 14 countries) from two EU Horizon projects (ProCAncer-I, COMFORT) and 12 independent centers based in Europe, North America, Asia and Africa, are used for training and internal testing. Additionally, 2010 cases (2010 patients; 20 external cities in 12 countries) from population-based screening (STHLM3-MRI, IP1-PROSTAGRAM trials) and primary diagnostic settings (PRIME trial) based in Europe, North and South Americas, Asia and Australia, are used for external testing. Primary endpoint is the proportion of AI-based assessments in agreement with the standard of care diagnoses (i.e., clinical assessments made by expert uropathologists on histopathology, if available, or at least two expert urogenital radiologists in consensus; with access to patient history and peer consultation) in the detection of Gleason grade group $\geq$2 prostate cancer within the external testing cohorts. Our statistical analysis plan is prespecified with a hypothesis of diagnostic interchangeability to the standard of care at the PI-RADS $\geq$3 (primary diagnosis) or $\geq$4 (screening) cut-off, considering an absolute margin of 0.05 and reader estimates derived from the PI-CAI observer study (62 radiologists reading 400 cases). Secondary measures comprise the area under the receiver operating characteristic curve (AUROC) of the AI system stratified by imaging quality, patient age and patient ethnicity to identify underlying biases (if any).

Hongwei Bran Li, Fernando Navarro, Ivan Ezhov et al.

Uncertainty in medical image segmentation tasks, especially inter-rater variability, arising from differences in interpretations and annotations by various experts, presents a significant challenge in achieving consistent and reliable image segmentation. This variability not only reflects the inherent complexity and subjective nature of medical image interpretation but also directly impacts the development and evaluation of automated segmentation algorithms. Accurately modeling and quantifying this variability is essential for enhancing the robustness and clinical applicability of these algorithms. We report the set-up and summarize the benchmark results of the Quantification of Uncertainties in Biomedical Image Quantification Challenge (QUBIQ), which was organized in conjunction with International Conferences on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2020 and 2021. The challenge focuses on the uncertainty quantification of medical image segmentation which considers the omnipresence of inter-rater variability in imaging datasets. The large collection of images with multi-rater annotations features various modalities such as MRI and CT; various organs such as the brain, prostate, kidney, and pancreas; and different image dimensions 2D-vs-3D. A total of 24 teams submitted different solutions to the problem, combining various baseline models, Bayesian neural networks, and ensemble model techniques. The obtained results indicate the importance of the ensemble models, as well as the need for further research to develop efficient 3D methods for uncertainty quantification methods in 3D segmentation tasks.

Anindo Saha, Joeran Bosma, Jasper Linmans et al.

We hypothesize that probabilistic voxel-level classification of anatomy and malignancy in prostate MRI, although typically posed as near-identical segmentation tasks via U-Nets, require different loss functions for optimal performance due to inherent differences in their clinical objectives. We investigate distribution, region and boundary-based loss functions for both tasks across 200 patient exams from the publicly-available ProstateX dataset. For evaluation, we conduct a thorough comparative analysis of model predictions and calibration, measured with respect to multi-class volume segmentation of the prostate anatomy (whole-gland, transitional zone, peripheral zone), as well as, patient-level diagnosis and lesion-level detection of clinically significant prostate cancer. Notably, we find that distribution-based loss functions (in particular, focal loss) are well-suited for diagnostic or panoptic segmentation tasks such as lesion detection, primarily due to their implicit property of inducing better calibration. Meanwhile, (with the exception of focal loss) both distribution and region/boundary-based loss functions perform equally well for anatomical or semantic segmentation tasks, such as quantification of organ shape, size and boundaries.

Annika Reinke, Minu D. Tizabi, Carole H. Sudre et al.

While the importance of automatic image analysis is continuously increasing, recent meta-research revealed major flaws with respect to algorithm validation. Performance metrics are particularly key for meaningful, objective, and transparent performance assessment and validation of the used automatic algorithms, but relatively little attention has been given to the practical pitfalls when using specific metrics for a given image analysis task. These are typically related to (1) the disregard of inherent metric properties, such as the behaviour in the presence of class imbalance or small target structures, (2) the disregard of inherent data set properties, such as the non-independence of the test cases, and (3) the disregard of the actual biomedical domain interest that the metrics should reflect. This living dynamically document has the purpose to illustrate important limitations of performance metrics commonly applied in the field of image analysis. In this context, it focuses on biomedical image analysis problems that can be phrased as image-level classification, semantic segmentation, instance segmentation, or object detection task. The current version is based on a Delphi process on metrics conducted by an international consortium of image analysis experts from more than 60 institutions worldwide.

Anindo Saha, Matin Hosseinzadeh, Henkjan Huisman

We present a multi-stage 3D computer-aided detection and diagnosis (CAD) model for automated localization of clinically significant prostate cancer (csPCa) in bi-parametric MR imaging (bpMRI). Deep attention mechanisms drive its detection network, targeting salient structures and highly discriminative feature dimensions across multiple resolutions. Its goal is to accurately identify csPCa lesions from indolent cancer and the wide range of benign pathology that can afflict the prostate gland. Simultaneously, a decoupled residual classifier is used to achieve consistent false positive reduction, without sacrificing high sensitivity or computational efficiency. In order to guide model generalization with domain-specific clinical knowledge, a probabilistic anatomical prior is used to encode the spatial prevalence and zonal distinction of csPCa. Using a large dataset of 1950 prostate bpMRI paired with radiologically-estimated annotations, we hypothesize that such CNN-based models can be trained to detect biopsy-confirmed malignancies in an independent cohort. For 486 institutional testing scans, the 3D CAD system achieves 83.69$\pm$5.22% and 93.19$\pm$2.96% detection sensitivity at 0.50 and 1.46 false positive(s) per patient, respectively, with 0.882$\pm$0.030 AUROC in patient-based diagnosis $-$significantly outperforming four state-of-the-art baseline architectures (U-SEResNet, UNet++, nnU-Net, Attention U-Net) from recent literature. For 296 external biopsy-confirmed testing scans, the ensembled CAD system shares moderate agreement with a consensus of expert radiologists (76.69%; $kappa$ $=$ 0.51$\pm$0.04) and independent pathologists (81.08%; $kappa$ $=$ 0.56$\pm$0.06); demonstrating strong generalization to histologically-confirmed csPCa diagnosis.

Anindo Saha, Matin Hosseinzadeh, Henkjan Huisman

We hypothesize that anatomical priors can be viable mediums to infuse domain-specific clinical knowledge into state-of-the-art convolutional neural networks (CNN) based on the U-Net architecture. We introduce a probabilistic population prior which captures the spatial prevalence and zonal distinction of clinically significant prostate cancer (csPCa), in order to improve its computer-aided detection (CAD) in bi-parametric MR imaging (bpMRI). To evaluate performance, we train 3D adaptations of the U-Net, U-SEResNet, UNet++ and Attention U-Net using 800 institutional training-validation scans, paired with radiologically-estimated annotations and our computed prior. For 200 independent testing bpMRI scans with histologically-confirmed delineations of csPCa, our proposed method of encoding clinical priori demonstrates a strong ability to improve patient-based diagnosis (upto 8.70% increase in AUROC) and lesion-level detection (average increase of 1.08 pAUC between 0.1-10 false positives per patient) across all four architectures.

Anindo Saha, Fakrul I. Tushar, Khrystyna Faryna et al.

Weakly supervised disease classification of CT imaging suffers from poor localization owing to case-level annotations, where even a positive scan can hold hundreds to thousands of negative slices along multiple planes. Furthermore, although deep learning segmentation and classification models extract distinctly unique combinations of anatomical features from the same target class(es), they are typically seen as two independent processes in a computer-aided diagnosis (CAD) pipeline, with little to no feature reuse. In this research, we propose a medical classifier that leverages the semantic structural concepts learned via multi-resolution segmentation feature maps, to guide weakly supervised 3D classification of chest CT volumes. Additionally, a comparative analysis is drawn across two different types of feature aggregation to explore the vast possibilities surrounding feature fusion. Using a dataset of 1593 scans labeled on a case-level basis via rule-based model, we train a dual-stage convolutional neural network (CNN) to perform organ segmentation and binary classification of four representative diseases (emphysema, pneumonia/atelectasis, mass and nodules) in lungs. The baseline model, with separate stages for segmentation and classification, results in AUC of 0.791. Using identical hyperparameters, the connected architecture using static and dynamic feature aggregation improves performance to AUC of 0.832 and 0.851, respectively. This study advances the field in two key ways. First, case-level report data is used to weakly supervise a 3D CT classifier of multiple, simultaneous diseases for an organ. Second, segmentation and classification models are connected with two different feature aggregation strategies to enhance the classification performance.

Anindo Saha, Prem Prasad, Abdullah Thabit

Fully automatic detection of skin lesions in dermatoscopic images can facilitate early diagnosis and repression of malignant melanoma and non-melanoma skin cancer. Although convolutional neural networks are a powerful solution, they are limited by the illumination spectrum of annotated dermatoscopic screening images, where color is an important discriminative feature. In this paper, we propose an adaptive color augmentation technique to amplify data expression and model performance, while regulating color difference and saturation to minimize the risks of using synthetic data. Through deep visualization, we qualitatively identify and verify the semantic structural features learned by the network for discriminating skin lesions against normal skin tissue. The overall system achieves a Dice Ratio of 0.891 with 0.943 sensitivity and 0.932 specificity on the ISIC 2018 Testing Set for segmentation.

Jaime Simarro, Zohaib Salahuddin, Ahmed Gouda et al.

Identification and segmentation of breast masses in mammograms face complex challenges, owing to the highly variable nature of malignant densities with regards to their shape, contours, texture and orientation. Additionally, classifiers typically suffer from high class imbalance in region candidates, where normal tissue regions vastly outnumber malignant masses. This paper proposes a rigorous segmentation method, supported by morphological enhancement using grayscale linear filters. A novel cascaded ensemble of support vector machines (SVM) is used to effectively tackle the class imbalance and provide significant predictions. For True Positive Rate (TPR) of 0.35, 0.69 and 0.82, the system generates only 0.1, 0.5 and 1.0 False Positives/Image (FPI), respectively.