Erik B. Dam

Raghavendra Selvan, Nikhil Bhagwat, Lasse F. Wolff Anthony et al.

The increasing energy consumption and carbon footprint of deep learning (DL) due to growing compute requirements has become a cause of concern. In this work, we focus on the carbon footprint of developing DL models for medical image analysis (MIA), where volumetric images of high spatial resolution are handled. In this study, we present and compare the features of four tools from literature to quantify the carbon footprint of DL. Using one of these tools we estimate the carbon footprint of medical image segmentation pipelines. We choose nnU-net as the proxy for a medical image segmentation pipeline and experiment on three common datasets. With our work we hope to inform on the increasing energy costs incurred by MIA. We discuss simple strategies to cut-down the environmental impact that can make model selection and training processes more efficient.

Sumit Pandey, Kuan-Fu Chen, Erik B. Dam

This paper introduces a comprehensive approach for segmenting regions of interest (ROI) in diverse medical imaging datasets, encompassing ultrasound, CT scans, and X-ray images. The proposed method harnesses the capabilities of the YOLOv8 model for approximate boundary box detection across modalities, alongside the Segment Anything Model (SAM) and High Quality (HQ) SAM for fully automatic and precise segmentation. To generate boundary boxes, the YOLOv8 model was trained using a limited set of 100 images and masks from each modality. The results obtained from our approach are extensively computed and analyzed, demonstrating its effectiveness and potential in medical image analysis. Various evaluation metrics, including precision, recall, F1 score, and Dice Score, were employed to quantify the accuracy of the segmentation results. A comparative analysis was conducted to assess the individual and combined performance of the YOLOv8, YOLOv8+SAM, and YOLOv8+HQ-SAM models. The results indicate that the SAM model performs better than the other two models, exhibiting higher segmentation accuracy and overall performance. While HQ-SAM offers potential advantages, its incremental gains over the standard SAM model may not justify the additional computational cost. The YOLOv8+SAM model shows promise for enhancing medical image segmentation and its clinical implications.

Sandeep Singh Sengara, Christopher Meulengrachtb, Mikael Ploug Boesenb et al.

UNet has become the gold standard method for segmenting 2D medical images that any new method must be validated against. However, in recent years, several variations of the seminal UNet have been proposed with promising results. However, there is no clear consensus on the generalisability of these architectures, and UNet currently remains the methodological gold standard. The purpose of this study was to evaluate some of the most promising UNet-inspired architectures for 3D segmentation. For the segmentation of 3D scans, UNet-inspired methods are also dominant, but there is a larger variety across applications. By evaluating the architectures in a different dimensionality, embedded in a different method, and for a different task, we aimed to evaluate if any of these UNet-alternatives are promising as a new gold standard that generalizes even better than UNet. Specifically, we investigated the architectures as the central 2D segmentation core in the Multi-Planar Unet 3D segmentation method that previously demonstrated excellent generalization in the MICCAI Segmentation Decathlon. Generalisability can be demonstrated if a promising UNet-variant consistently outperforms UNet in this setting. For this purpose, we evaluated four architectures for cartilage segmentation from three different cohorts with knee MRIs.

Raghavendra Selvan, Julian Schön, Erik B Dam

The accelerated development of machine learning methods, primarily deep learning, are causal to the recent breakthroughs in medical image analysis and computer aided intervention. The resource consumption of deep learning models in terms of amount of training data, compute and energy costs are known to be massive. These large resource costs can be barriers in deploying these models in clinics, globally. To address this, there are cogent efforts within the machine learning community to introduce notions of resource efficiency. For instance, using quantisation to alleviate memory consumption. While most of these methods are shown to reduce the resource utilisation, they could come at a cost in performance. In this work, we probe into the trade-off between resource consumption and performance, specifically, when dealing with models that are used in critical settings such as in clinics.

Sophia N. Wilson, Sebastian Mair, Mophat Okinyi et al.

Large-scale data has fuelled the success of frontier artificial intelligence (AI) models over the past decade. This expansion has relied on sustained efforts by large technology corporations to aggregate and curate internet-scale datasets. In this work, we examine the environmental, social, and economic costs of large-scale data in AI through a sustainability lens. We argue that the field is shifting from building models from data to actively creating data for building models. We characterise this transition as hyper-datafication, which marks a critical juncture for the future of frontier AI and its societal impacts. To quantify and contextualise data-related costs, we analyse approximately 550,000 datasets from the Hugging Face Hub, focusing on dataset growth, storage-related energy consumption and carbon footprint, and societal representation using language data. We complement this analysis with qualitative responses from data workers in Kenya to examine the labour involved, including direct employment by big tech corporations and exposure to graphic content. We further draw on external data sources to substantiate our findings by illustrating the global disparity in data centre infrastructure. Our analyses reveal that hyper-datafication does not merely increase resource consumption but systematically redistributes environmental burdens, labour risks, and representational harms toward the Global South, precarious data workers, and under-represented cultures. Thus, we propose Data PROOFS recommendations spanning provenance, resource awareness, ownership, openness, frugality, and standards to mitigate these costs. Our work aims to make visible the often-overlooked costs of data that underpin frontier AI and to stimulate broader debate within the research community and beyond.

Ulrik Friis-Jensen, Frederik L. Johansen, Andy S. Anker et al.

Advances in graph machine learning (ML) have been driven by applications in chemistry as graphs have remained the most expressive representations of molecules. While early graph ML methods focused primarily on small organic molecules, recently, the scope of graph ML has expanded to include inorganic materials. Modelling the periodicity and symmetry of inorganic crystalline materials poses unique challenges, which existing graph ML methods are unable to address. Moving to inorganic nanomaterials increases complexity as the scale of number of nodes within each graph can be broad ($10$ to $10^5$). The bulk of existing graph ML focuses on characterising molecules and materials by predicting target properties with graphs as input. However, the most exciting applications of graph ML will be in their generative capabilities, which is currently not at par with other domains such as images or text. We invite the graph ML community to address these open challenges by presenting two new chemically-informed large-scale inorganic (CHILI) nanomaterials datasets: A medium-scale dataset (with overall >6M nodes, >49M edges) of mono-metallic oxide nanomaterials generated from 12 selected crystal types (CHILI-3K) and a large-scale dataset (with overall >183M nodes, >1.2B edges) of nanomaterials generated from experimentally determined crystal structures (CHILI-100K). We define 11 property prediction tasks and 6 structure prediction tasks, which are of special interest for nanomaterial research. We benchmark the performance of a wide array of baseline methods and use these benchmarking results to highlight areas which need future work. To the best of our knowledge, CHILI-3K and CHILI-100K are the first open-source nanomaterial datasets of this scale -- both on the individual graph level and of the dataset as a whole -- and the only nanomaterials datasets with high structural and elemental diversity.

Frederik Lizak Johansen, Ulrik Friis-Jensen, Erik Bjørnager Dam et al.

Novel materials drive progress across applications from energy storage to electronics. Automated characterization of material structures with machine learning methods offers a promising strategy for accelerating this key step in material design. In this work, we introduce an autoregressive language model that performs crystal structure prediction (CSP) from powder diffraction data. The presented model, deCIFer, generates crystal structures in the widely used Crystallographic Information File (CIF) format and can be conditioned on powder X-ray diffraction (PXRD) data. Unlike earlier works that primarily rely on high-level descriptors like composition, deCIFer is also able to use diffraction data to perform CSP. We train deCIFer on nearly 2.3M crystal structures and validate on diverse sets of PXRD patterns for characterizing challenging inorganic crystal systems. Qualitative checks and quantitative assessments using the residual weighted profile show that deCIFer produces structures that more accurately match the target diffraction data. Notably, deCIFer can achieve a 94% match rate on test data. deCIFer bridges experimental diffraction data with computational CSP, lending itself as a powerful tool for crystal structure characterization.

Sumit Pandey, Satyasaran Changdar, Mathias Perslev et al.

Automated segmentation of distinct tumor regions is critical for accurate diagnosis and treatment planning in pediatric brain tumors. This study evaluates the efficacy of the Multi-Planner U-Net (MPUnet) approach in segmenting different tumor subregions across three challenging datasets: Pediatrics Tumor Challenge (PED), Brain Metastasis Challenge (MET), and Sub-Sahara-Africa Adult Glioma (SSA). These datasets represent diverse scenarios and anatomical variations, making them suitable for assessing the robustness and generalization capabilities of the MPUnet model. By utilizing multi-planar information, the MPUnet architecture aims to enhance segmentation accuracy. Our results show varying performance levels across the evaluated challenges, with the tumor core (TC) class demonstrating relatively higher segmentation accuracy. However, variability is observed in the segmentation of other classes, such as the edema and enhancing tumor (ET) regions. These findings emphasize the complexity of brain tumor segmentation and highlight the potential for further refinement of the MPUnet approach and inclusion of MRI more data and preprocessing.

Raghavendra Selvan, Bob Pepin, Christian Igel et al.

The recent advances in deep learning (DL) have been accelerated by access to large-scale data and compute. These large-scale resources have been used to train progressively larger models which are resource intensive in terms of compute, data, energy, and carbon emissions. These costs are becoming a new type of entry barrier to researchers and practitioners with limited access to resources at such scale, particularly in the Global South. In this work, we take a comprehensive look at the landscape of existing DL models for medical image analysis tasks and demonstrate their usefulness in settings where resources are limited. To account for the resource consumption of DL models, we introduce a novel measure to estimate the performance per resource unit, which we call the PePR score. Using a diverse family of 131 unique DL architectures (spanning 1M to 130M trainable parameters) and three medical image datasets, we capture trends about the performance-resource trade-offs. In applications like medical image analysis, we argue that small-scale, specialized models are better than striving for large-scale models. Furthermore, we show that using existing pretrained models that are fine-tuned on new data can significantly reduce the computational resources and data required compared to training models from scratch. We hope this work will encourage the community to focus on improving AI equity by developing methods and models with smaller resource footprints.

Raghavendra Selvan, Erik B Dam, Søren Alexander Flensborg et al.

Tensor networks are efficient factorisations of high-dimensional tensors into a network of lower-order tensors. They have been most commonly used to model entanglement in quantum many-body systems and more recently are witnessing increased applications in supervised machine learning. In this work, we formulate image segmentation in a supervised setting with tensor networks. The key idea is to first lift the pixels in image patches to exponentially high-dimensional feature spaces and using a linear decision hyper-plane to classify the input pixels into foreground and background classes. The high-dimensional linear model itself is approximated using the matrix product state (MPS) tensor network. The MPS is weight-shared between the non-overlapping image patches resulting in our strided tensor network model. The performance of the proposed model is evaluated on three 2D- and one 3D- biomedical imaging datasets. The performance of the proposed tensor network segmentation model is compared with relevant baseline methods. In the 2D experiments, the tensor network model yields competitive performance compared to the baseline methods while being more resource efficient.

Raghavendra Selvan, Silas Ørting, Erik B Dam

The recently introduced locally orderless tensor network (LoTeNet) for supervised image classification uses matrix product state (MPS) operations on grids of transformed image patches. The resulting patch representations are combined back together into the image space and aggregated hierarchically using multiple MPS blocks per layer to obtain the final decision rules. In this work, we propose a non-patch based modification to LoTeNet that performs one MPS operation per layer, instead of several patch-level operations. The spatial information in the input images to MPS blocks at each layer is squeezed into the feature dimension, similar to LoTeNet, to maximise retained spatial correlation between pixels when images are flattened into 1D vectors. The proposed multi-layered tensor network (MLTN) is capable of learning linear decision boundaries in high dimensional spaces in a multi-layered setting, which results in a reduction in the computation cost compared to LoTeNet without any degradation in performance.

Raghavendra Selvan, Silas Ørting, Erik B Dam

Tensor networks are factorisations of high rank tensors into networks of lower rank tensors and have primarily been used to analyse quantum many-body problems. Tensor networks have seen a recent surge of interest in relation to supervised learning tasks with a focus on image classification. In this work, we improve upon the matrix product state (MPS) tensor networks that can operate on one-dimensional vectors to be useful for working with 2D and 3D medical images. We treat small image regions as orderless, squeeze their spatial information into feature dimensions and then perform MPS operations on these locally orderless regions. These local representations are then aggregated in a hierarchical manner to retain global structure. The proposed locally orderless tensor network (LoTeNet) is compared with relevant methods on three datasets. The architecture of LoTeNet is fixed in all experiments and we show it requires lesser computational resources to attain performance on par or superior to the compared methods.

Raghavendra Selvan, Erik B. Dam, Nicki S. Detlefsen et al.

Pulmonary opacification is the inflammation in the lungs caused by many respiratory ailments, including the novel corona virus disease 2019 (COVID-19). Chest X-rays (CXRs) with such opacifications render regions of lungs imperceptible, making it difficult to perform automated image analysis on them. In this work, we focus on segmenting lungs from such abnormal CXRs as part of a pipeline aimed at automated risk scoring of COVID-19 from CXRs. We treat the high opacity regions as missing data and present a modified CNN-based image segmentation network that utilizes a deep generative model for data imputation. We train this model on normal CXRs with extensive data augmentation and demonstrate the usefulness of this model to extend to cases with extreme abnormalities.

Arjun D. Desai, Francesco Caliva, Claudia Iriondo et al.

Purpose: To organize a knee MRI segmentation challenge for characterizing the semantic and clinical efficacy of automatic segmentation methods relevant for monitoring osteoarthritis progression. Methods: A dataset partition consisting of 3D knee MRI from 88 subjects at two timepoints with ground-truth articular (femoral, tibial, patellar) cartilage and meniscus segmentations was standardized. Challenge submissions and a majority-vote ensemble were evaluated using Dice score, average symmetric surface distance, volumetric overlap error, and coefficient of variation on a hold-out test set. Similarities in network segmentations were evaluated using pairwise Dice correlations. Articular cartilage thickness was computed per-scan and longitudinally. Correlation between thickness error and segmentation metrics was measured using Pearson's coefficient. Two empirical upper bounds for ensemble performance were computed using combinations of model outputs that consolidated true positives and true negatives. Results: Six teams (T1-T6) submitted entries for the challenge. No significant differences were observed across all segmentation metrics for all tissues (p=1.0) among the four top-performing networks (T2, T3, T4, T6). Dice correlations between network pairs were high (>0.85). Per-scan thickness errors were negligible among T1-T4 (p=0.99) and longitudinal changes showed minimal bias (<0.03mm). Low correlations (<0.41) were observed between segmentation metrics and thickness error. The majority-vote ensemble was comparable to top performing networks (p=1.0). Empirical upper bound performances were similar for both combinations (p=1.0). Conclusion: Diverse networks learned to segment the knee similarly where high segmentation accuracy did not correlate to cartilage thickness accuracy. Voting ensembles did not outperform individual networks but may help regularize individual models.

Raghavendra Selvan, Erik B Dam

With the increasing adoption of machine learning tools like neural networks across several domains, interesting connections and comparisons to concepts from other domains are coming to light. In this work, we focus on the class of Tensor Networks, which has been a work horse for physicists in the last two decades to analyse quantum many-body systems. Building on the recent interest in tensor networks for machine learning, we extend the Matrix Product State tensor networks (which can be interpreted as linear classifiers operating in exponentially high dimensional spaces) to be useful in medical image analysis tasks. We focus on classification problems as a first step where we motivate the use of tensor networks and propose adaptions for 2D images using classical image domain concepts such as local orderlessness of images. With the proposed locally orderless tensor network model (LoTeNet), we show that tensor networks are capable of attaining performance that is comparable to state-of-the-art deep learning methods. We evaluate the model on two publicly available medical imaging datasets and show performance improvements with fewer model hyperparameters and lesser computational resources compared to relevant baseline methods.

Patrick Bilic, Patrick Christ, Hongwei Bran Li et al.

In this work, we report the set-up and results of the Liver Tumor Segmentation Benchmark (LiTS), which was organized in conjunction with the IEEE International Symposium on Biomedical Imaging (ISBI) 2017 and the International Conferences on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2017 and 2018. The image dataset is diverse and contains primary and secondary tumors with varied sizes and appearances with various lesion-to-background levels (hyper-/hypo-dense), created in collaboration with seven hospitals and research institutions. Seventy-five submitted liver and liver tumor segmentation algorithms were trained on a set of 131 computed tomography (CT) volumes and were tested on 70 unseen test images acquired from different patients. We found that not a single algorithm performed best for both liver and liver tumors in the three events. The best liver segmentation algorithm achieved a Dice score of 0.963, whereas, for tumor segmentation, the best algorithms achieved Dices scores of 0.674 (ISBI 2017), 0.702 (MICCAI 2017), and 0.739 (MICCAI 2018). Retrospectively, we performed additional analysis on liver tumor detection and revealed that not all top-performing segmentation algorithms worked well for tumor detection. The best liver tumor detection method achieved a lesion-wise recall of 0.458 (ISBI 2017), 0.515 (MICCAI 2017), and 0.554 (MICCAI 2018), indicating the need for further research. LiTS remains an active benchmark and resource for research, e.g., contributing the liver-related segmentation tasks in \url{http://medicaldecathlon.com/}. In addition, both data and online evaluation are accessible via \url{www.lits-challenge.com}.

Erik B Dam

Scanner drift is a well-known magnetic resonance imaging (MRI) artifact characterized by gradual signal degradation and scan intensity changes over time. In addition, hardware and software updates may imply abrupt changes in signal. The combined effects are particularly challenging for automatic image analysis methods used in longitudinal studies. The implication is increased measurement variation and a risk of bias in the estimations (e.g. in the volume change for a structure). We proposed two quite different approaches for scanner drift normalization and demonstrated the performance for segmentation of knee MRI using the fully automatic KneeIQ framework. The validation included a total of 1975 scans from both high-field and low-field MRI. The results demonstrated that the pre-processing method denoted Atlas Affine Normalization significantly removed scanner drift effects and ensured that the cartilage volume change quantifications became consistent with manual expert scores.