Nicholas Heller

Jianning Li, Zongwei Zhou, Jiancheng Yang et al.

Prior to the deep learning era, shape was commonly used to describe the objects. Nowadays, state-of-the-art (SOTA) algorithms in medical imaging are predominantly diverging from computer vision, where voxel grids, meshes, point clouds, and implicit surface models are used. This is seen from numerous shape-related publications in premier vision conferences as well as the growing popularity of ShapeNet (about 51,300 models) and Princeton ModelNet (127,915 models). For the medical domain, we present a large collection of anatomical shapes (e.g., bones, organs, vessels) and 3D models of surgical instrument, called MedShapeNet, created to facilitate the translation of data-driven vision algorithms to medical applications and to adapt SOTA vision algorithms to medical problems. As a unique feature, we directly model the majority of shapes on the imaging data of real patients. As of today, MedShapeNet includes 23 dataset with more than 100,000 shapes that are paired with annotations (ground truth). Our data is freely accessible via a web interface and a Python application programming interface (API) and can be used for discriminative, reconstructive, and variational benchmarks as well as various applications in virtual, augmented, or mixed reality, and 3D printing. Exemplary, we present use cases in the fields of classification of brain tumors, facial and skull reconstructions, multi-class anatomy completion, education, and 3D printing. In future, we will extend the data and improve the interfaces. The project pages are: https://medshapenet.ikim.nrw/ and https://github.com/Jianningli/medshapenet-feedback

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.

Nicholas Heller, Fabian Isensee, Dasha Trofimova et al.

This paper presents the challenge report for the 2021 Kidney and Kidney Tumor Segmentation Challenge (KiTS21) held in conjunction with the 2021 international conference on Medical Image Computing and Computer Assisted Interventions (MICCAI). KiTS21 is a sequel to its first edition in 2019, and it features a variety of innovations in how the challenge was designed, in addition to a larger dataset. A novel annotation method was used to collect three separate annotations for each region of interest, and these annotations were performed in a fully transparent setting using a web-based annotation tool. Further, the KiTS21 test set was collected from an outside institution, challenging participants to develop methods that generalize well to new populations. Nonetheless, the top-performing teams achieved a significant improvement over the state of the art set in 2019, and this performance is shown to inch ever closer to human-level performance. An in-depth meta-analysis is presented describing which methods were used and how they faired on the leaderboard, as well as the characteristics of which cases generally saw good performance, and which did not. Overall KiTS21 facilitated a significant advancement in the state of the art in kidney tumor segmentation, and provides useful insights that are applicable to the field of semantic segmentation as a whole.

Matthias Eisenmann, Annika Reinke, Vivienn Weru et al.

International benchmarking competitions have become fundamental for the comparative performance assessment of image analysis methods. However, little attention has been given to investigating what can be learnt from these competitions. Do they really generate scientific progress? What are common and successful participation strategies? What makes a solution superior to a competing method? To address this gap in the literature, we performed a multi-center study with all 80 competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021. Statistical analyses performed based on comprehensive descriptions of the submitted algorithms linked to their rank as well as the underlying participation strategies revealed common characteristics of winning solutions. These typically include the use of multi-task learning (63%) and/or multi-stage pipelines (61%), and a focus on augmentation (100%), image preprocessing (97%), data curation (79%), and postprocessing (66%). The "typical" lead of a winning team is a computer scientist with a doctoral degree, five years of experience in biomedical image analysis, and four years of experience in deep learning. Two core general development strategies stood out for highly-ranked teams: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases. According to the organizers, 43% of the winning algorithms exceeded the state of the art but only 11% completely solved the respective domain problem. The insights of our study could help researchers (1) improve algorithm development strategies when approaching new problems, and (2) focus on open research questions revealed by this work.

Tim Rädsch, Annika Reinke, Vivienn Weru et al.

This paper does not describe a novel method. Instead, it studies an essential foundation for reliable benchmarking and ultimately real-world application of AI-based image analysis: generating high-quality reference annotations. Previous research has focused on crowdsourcing as a means of outsourcing annotations. However, little attention has so far been given to annotation companies, specifically regarding their internal quality assurance (QA) processes. Therefore, our aim is to evaluate the influence of QA employed by annotation companies on annotation quality and devise methodologies for maximizing data annotation efficacy. Based on a total of 57,648 instance segmented images obtained from a total of 924 annotators and 34 QA workers from four annotation companies and Amazon Mechanical Turk (MTurk), we derived the following insights: (1) Annotation companies perform better both in terms of quantity and quality compared to the widely used platform MTurk. (2) Annotation companies' internal QA only provides marginal improvements, if any. However, improving labeling instructions instead of investing in QA can substantially boost annotation performance. (3) The benefit of internal QA depends on specific image characteristics. Our work could enable researchers to derive substantially more value from a fixed annotation budget and change the way annotation companies conduct internal QA.

Jun Ma, Feifei Li, Sumin Kim et al.

Promptable segmentation foundation models have emerged as a transformative approach to addressing the diverse needs in medical images, but most existing models require expensive computing, posing a big barrier to their adoption in clinical practice. In this work, we organized the first international competition dedicated to promptable medical image segmentation, featuring a large-scale dataset spanning nine common imaging modalities from over 20 different institutions. The top teams developed lightweight segmentation foundation models and implemented an efficient inference pipeline that substantially reduced computational requirements while maintaining state-of-the-art segmentation accuracy. Moreover, the post-challenge phase advanced the algorithms through the design of performance booster and reproducibility tasks, resulting in improved algorithms and validated reproducibility of the winning solution. Furthermore, the best-performing algorithms have been incorporated into the open-source software with a user-friendly interface to facilitate clinical adoption. The data and code are publicly available to foster the further development of medical image segmentation foundation models and pave the way for impactful real-world applications.

Nicholas Heller, Derek Anderson, Matt Baker et al.

Air travel is one of the fastest growing modes of transportation, however, the effects of aircraft noise on populations surrounding airports is hindering its growth. In an effort to study and ultimately mitigate the impact that this noise has, many airports continuously monitor the aircraft noise in their surrounding communities. Noise monitoring and analysis is complicated by the fact that aircraft are not the only source of noise. In this work, we show that a Convolutional Neural Network is well-suited for the task of identifying noise events which are not caused by aircraft. Our system achieves an accuracy of 0.970 when trained on 900 manually labeled noise events. Our training data and a TensorFlow implementation of our model are available at https://github.com/neheller/aircraftnoise.

Rikhil Seshadri, Jayant Siva, Angelica Bartholomew et al.

Kidney cancer is a global health concern, and accurate assessment of patient frailty is crucial for optimizing surgical outcomes. This paper introduces AI Age Discrepancy, a novel metric derived from machine learning analysis of preoperative abdominal CT scans, as a potential indicator of frailty and postoperative risk in kidney cancer patients. This retrospective study of 599 patients from the 2023 Kidney Tumor Segmentation (KiTS) challenge dataset found that a higher AI Age Discrepancy is significantly associated with longer hospital stays and lower overall survival rates, independent of established factors. This suggests that AI Age Discrepancy may provide valuable insights into patient frailty and could thus inform clinical decision-making in kidney cancer treatment.

Jason D Kelly, Nicholas Heller, Ashley Petersen et al.

Purpose: Previous research has shown that obtaining non-expert crowd evaluations of surgical performances concords with the gold standard of expert surgeon review, and that faster playback speed increases ratings for videos of higher-skilled surgeons in laparoscopic simulation. The aim of this research is to extend this investigation to real surgeries that use non-expert crowd evaluations. We address two questions (1) whether crowds award more favorable ratings to videos shown at increased playback speeds, and (2) if crowd evaluations of the first minute of a surgical procedure differ from crowd evaluations of the entire performance. Methods: A set of 56 videos of practicing surgeons were used to evaluate the technical skill of the surgeons at each video playback speed used for the first minute of the previously rated performance, using the GEARS assessment criteria. Results: Crowds on average did rate videos higher as playback speed was increased. This effect was observed for both proficient and expert surgeons. Each increase in the playback speed by 0.8x was associated with, on average, a 0.16-point increase in the GEARS score for expert surgeons and a 0.27-point increase in GEARS score for proficient surgeons, with both groups being perceived as obtaining relatively equal skill at the fastest playback speed. It was also found that 22 out of the 56 surgeons were perceived to be significantly different in skill when just viewing the first minute of performance. Conclusion: The observed increase in skill ratings with video playback speed replicates findings for laparoscopy in [2], and extends to real robotic surgeries. Furthermore, the large differences in skill labels when comparing the first minute of surgery to the entire 15 minute video warrants further investigation into how much perceived skill ratings vary in time (sub-task level) vs. summative metrics (task level).

Nicholas Heller, Fabian Isensee, Klaus H. Maier-Hein et al.

There is a large body of literature linking anatomic and geometric characteristics of kidney tumors to perioperative and oncologic outcomes. Semantic segmentation of these tumors and their host kidneys is a promising tool for quantitatively characterizing these lesions, but its adoption is limited due to the manual effort required to produce high-quality 3D segmentations of these structures. Recently, methods based on deep learning have shown excellent results in automatic 3D segmentation, but they require large datasets for training, and there remains little consensus on which methods perform best. The 2019 Kidney and Kidney Tumor Segmentation challenge (KiTS19) was a competition held in conjunction with the 2019 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) which sought to address these issues and stimulate progress on this automatic segmentation problem. A training set of 210 cross sectional CT images with kidney tumors was publicly released with corresponding semantic segmentation masks. 106 teams from five continents used this data to develop automated systems to predict the true segmentation masks on a test set of 90 CT images for which the corresponding ground truth segmentations were kept private. These predictions were scored and ranked according to their average So rensen-Dice coefficient between the kidney and tumor across all 90 cases. The winning team achieved a Dice of 0.974 for kidney and 0.851 for tumor, approaching the inter-annotator performance on kidney (0.983) but falling short on tumor (0.923). This challenge has now entered an "open leaderboard" phase where it serves as a challenging benchmark in 3D semantic segmentation.

Nicholas Heller, Jack Rickman, Christopher Weight et al.

Widely-used public benchmarks are of huge importance to computer vision and machine learning research, especially with the computational resources required to reproduce state of the art results quickly becoming untenable. In medical image computing, the wide variety of image modalities and problem formulations yields a huge task-space for benchmarks to cover, and thus the widespread adoption of standard benchmarks has been slow, and barriers to releasing medical data exacerbate this issue. In this paper, we examine the role that publicly available data has played in MICCAI papers from the past five years. We find that more than half of these papers are based on private data alone, although this proportion seems to be decreasing over time. Additionally, we observed that after controlling for open access publication and the release of code, papers based on public data were cited over 60% more per year than their private-data counterparts. Further, we found that more than 20% of papers using public data did not provide a citation to the dataset or associated manuscript, highlighting the "second-rate" status that data contributions often take compared to theoretical ones. We conclude by making recommendations for MICCAI policies which could help to better incentivise data sharing and move the field toward more efficient and reproducible science.

Nicholas Heller, Niranjan Sathianathen, Arveen Kalapara et al.

The morphometry of a kidney tumor revealed by contrast-enhanced Computed Tomography (CT) imaging is an important factor in clinical decision making surrounding the lesion's diagnosis and treatment. Quantitative study of the relationship between kidney tumor morphology and clinical outcomes is difficult due to data scarcity and the laborious nature of manually quantifying imaging predictors. Automatic semantic segmentation of kidneys and kidney tumors is a promising tool towards automatically quantifying a wide array of morphometric features, but no sizeable annotated dataset is currently available to train models for this task. We present the KiTS19 challenge dataset: A collection of multi-phase CT imaging, segmentation masks, and comprehensive clinical outcomes for 300 patients who underwent nephrectomy for kidney tumors at our center between 2010 and 2018. 210 (70%) of these patients were selected at random as the training set for the 2019 MICCAI KiTS Kidney Tumor Segmentation Challenge and have been released publicly. With the presence of clinical context and surgical outcomes, this data can serve not only for benchmarking semantic segmentation models, but also for developing and studying biomarkers which make use of the imaging and semantic segmentation masks.

Nicholas Heller, Joshua Dean, Nikolaos Papanikolopoulos

Labeled datasets for semantic segmentation are imperfect, especially in medical imaging where borders are often subtle or ill-defined. Little work has been done to analyze the effect that label errors have on the performance of segmentation methodologies. Here we present a large-scale study of model performance in the presence of varying types and degrees of error in training data. We trained U-Net, SegNet, and FCN32 several times for liver segmentation with 10 different modes of ground-truth perturbation. Our results show that for each architecture, performance steadily declines with boundary-localized errors, however, U-Net was significantly more robust to jagged boundary errors than the other architectures. We also found that each architecture was very robust to non-boundary-localized errors, suggesting that boundary-localized errors are fundamentally different and more challenging problem than random label errors in a classification setting.