Touchstone Benchmark: Are We on the Right Way for Evaluating AI Algorithms for Medical Segmentation?

How can we test AI performance? This question seems trivial, but it isn't. Standard benchmarks often have problems such as in-distribution and small-size test sets, oversimplified metrics, unfair comparisons, and short-term outcome pressure. As a consequence, good performance on standard benchmarks does not guarantee success in real-world scenarios. To address these problems, we present Touchstone, a large-scale collaborative segmentation benchmark of 9 types of abdominal organs. This benchmark is based on 5,195 training CT scans from 76 hospitals around the world and 5,903 testing CT scans from 11 additional hospitals. This diverse test set enhances the statistical significance of benchmark results and rigorously evaluates AI algorithms across various out-of-distribution scenarios. We invited 14 inventors of 19 AI algorithms to train their algorithms, while our team, as a third party, independently evaluated these algorithms on three test sets. In addition, we also evaluated pre-existing AI frameworks--which, differing from algorithms, are more flexible and can support different algorithms--including MONAI from NVIDIA, nnU-Net from DKFZ, and numerous other open-source frameworks. We are committed to expanding this benchmark to encourage more innovation of AI algorithms for the medical domain.

Decoupling Semantic Similarity from Spatial Alignment for Neural Networks

What representation do deep neural networks learn? How similar are images to each other for neural networks? Despite the overwhelming success of deep learning methods key questions about their internal workings still remain largely unanswered, due to their internal high dimensionality and complexity. To address this, one approach is to measure the similarity of activation responses to various inputs. Representational Similarity Matrices (RSMs) distill this similarity into scalar values for each input pair. These matrices encapsulate the entire similarity structure of a system, indicating which input leads to similar responses. While the similarity between images is ambiguous, we argue that the spatial location of semantic objects does neither influence human perception nor deep learning classifiers. Thus this should be reflected in the definition of similarity between image responses for computer vision systems. Revisiting the established similarity calculations for RSMs we expose their sensitivity to spatial alignment. In this paper, we propose to solve this through semantic RSMs, which are invariant to spatial permutation. We measure semantic similarity between input responses by formulating it as a set-matching problem. Further, we quantify the superiority of semantic RSMs over spatio-semantic RSMs through image retrieval and by comparing the similarity between representations to the similarity between predicted class probabilities.

Overcoming Common Flaws in the Evaluation of Selective Classification Systems

Selective Classification, wherein models can reject low-confidence predictions, promises reliable translation of machine-learning based classification systems to real-world scenarios such as clinical diagnostics. While current evaluation of these systems typically assumes fixed working points based on pre-defined rejection thresholds, methodological progress requires benchmarking the general performance of systems akin to the $\mathrm{AUROC}$ in standard classification. In this work, we define 5 requirements for multi-threshold metrics in selective classification regarding task alignment, interpretability, and flexibility, and show how current approaches fail to meet them. We propose the Area under the Generalized Risk Coverage curve ($\mathrm{AUGRC}$), which meets all requirements and can be directly interpreted as the average risk of undetected failures. We empirically demonstrate the relevance of $\mathrm{AUGRC}$ on a comprehensive benchmark spanning 6 data sets and 13 confidence scoring functions. We find that the proposed metric substantially changes metric rankings on 5 out of the 6 data sets.

Mitigating False Predictions In Unreasonable Body Regions

Despite considerable strides in developing deep learning models for 3D medical image segmentation, the challenge of effectively generalizing across diverse image distributions persists. While domain generalization is acknowledged as vital for robust application in clinical settings, the challenges stemming from training with a limited Field of View (FOV) remain unaddressed. This limitation leads to false predictions when applied to body regions beyond the FOV of the training data. In response to this problem, we propose a novel loss function that penalizes predictions in implausible body regions, applicable in both single-dataset and multi-dataset training schemes. It is realized with a Body Part Regression model that generates axial slice positional scores. Through comprehensive evaluation using a test set featuring varying FOVs, our approach demonstrates remarkable improvements in generalization capabilities. It effectively mitigates false positive tumor predictions up to 85% and significantly enhances overall segmentation performance.

nnU-Net Revisited: A Call for Rigorous Validation in 3D Medical Image Segmentation

The release of nnU-Net marked a paradigm shift in 3D medical image segmentation, demonstrating that a properly configured U-Net architecture could still achieve state-of-the-art results. Despite this, the pursuit of novel architectures, and the respective claims of superior performance over the U-Net baseline, continued. In this study, we demonstrate that many of these recent claims fail to hold up when scrutinized for common validation shortcomings, such as the use of inadequate baselines, insufficient datasets, and neglected computational resources. By meticulously avoiding these pitfalls, we conduct a thorough and comprehensive benchmarking of current segmentation methods including CNN-based, Transformer-based, and Mamba-based approaches. In contrast to current beliefs, we find that the recipe for state-of-the-art performance is 1) employing CNN-based U-Net models, including ResNet and ConvNeXt variants, 2) using the nnU-Net framework, and 3) scaling models to modern hardware resources. These results indicate an ongoing innovation bias towards novel architectures in the field and underscore the need for more stringent validation standards in the quest for scientific progress.

CR3DT: Camera-RADAR Fusion for 3D Detection and Tracking

Accurate detection and tracking of surrounding objects is essential to enable self-driving vehicles. While Light Detection and Ranging (LiDAR) sensors have set the benchmark for high performance, the appeal of camera-only solutions lies in their cost-effectiveness. Notably, despite the prevalent use of Radio Detection and Ranging (RADAR) sensors in automotive systems, their potential in 3D detection and tracking has been largely disregarded due to data sparsity and measurement noise. As a recent development, the combination of RADARs and cameras is emerging as a promising solution. This paper presents Camera-RADAR 3D Detection and Tracking (CR3DT), a camera-RADAR fusion model for 3D object detection, and Multi-Object Tracking (MOT). Building upon the foundations of the State-of-the-Art (SotA) camera-only BEVDet architecture, CR3DT demonstrates substantial improvements in both detection and tracking capabilities, by incorporating the spatial and velocity information of the RADAR sensor. Experimental results demonstrate an absolute improvement in detection performance of 5.3% in mean Average Precision (mAP) and a 14.9% increase in Average Multi-Object Tracking Accuracy (AMOTA) on the nuScenes dataset when leveraging both modalities. CR3DT bridges the gap between high-performance and cost-effective perception systems in autonomous driving, by capitalizing on the ubiquitous presence of RADAR in automotive applications.

Shedding Light on the Ageing of Extra Virgin Olive Oil: Probing the Impact of Temperature with Fluorescence Spectroscopy and Machine Learning Techniques

This work systematically investigates the oxidation of extra virgin olive oil (EVOO) under accelerated storage conditions with UV absorption and total fluorescence spectroscopy. With the large amount of data collected, it proposes a method to monitor the oil's quality based on machine learning applied to highly-aggregated data. EVOO is a high-quality vegetable oil that has earned worldwide reputation for its numerous health benefits and excellent taste. Despite its outstanding quality, EVOO degrades over time owing to oxidation, which can affect both its health qualities and flavour. Therefore, it is highly relevant to quantify the effects of oxidation on EVOO and develop methods to assess it that can be easily implemented under field conditions, rather than in specialized laboratories. The following study demonstrates that fluorescence spectroscopy has the capability to monitor the effect of oxidation and assess the quality of EVOO, even when the data are highly aggregated. It shows that complex laboratory equipment is not necessary to exploit fluorescence spectroscopy using the proposed method and that cost-effective solutions, which can be used in-field by non-scientists, could provide an easily-accessible assessment of the quality of EVOO.

Anatomy-informed Data Augmentation for Enhanced Prostate Cancer Detection

Data augmentation (DA) is a key factor in medical image analysis, such as in prostate cancer (PCa) detection on magnetic resonance images. State-of-the-art computer-aided diagnosis systems still rely on simplistic spatial transformations to preserve the pathological label post transformation. However, such augmentations do not substantially increase the organ as well as tumor shape variability in the training set, limiting the model's ability to generalize to unseen cases with more diverse localized soft-tissue deformations. We propose a new anatomy-informed transformation that leverages information from adjacent organs to simulate typical physiological deformations of the prostate and generates unique lesion shapes without altering their label. Due to its lightweight computational requirements, it can be easily integrated into common DA frameworks. We demonstrate the effectiveness of our augmentation on a dataset of 774 biopsy-confirmed examinations, by evaluating a state-of-the-art method for PCa detection with different augmentation settings.

Exploring new ways: Enforcing representational dissimilarity to learn new features and reduce error consistency

Independently trained machine learning models tend to learn similar features. Given an ensemble of independently trained models, this results in correlated predictions and common failure modes. Previous attempts focusing on decorrelation of output predictions or logits yielded mixed results, particularly due to their reduction in model accuracy caused by conflicting optimization objectives. In this paper, we propose the novel idea of utilizing methods of the representational similarity field to promote dissimilarity during training instead of measuring similarity of trained models. To this end, we promote intermediate representations to be dissimilar at different depths between architectures, with the goal of learning robust ensembles with disjoint failure modes. We show that highly dissimilar intermediate representations result in less correlated output predictions and slightly lower error consistency, resulting in higher ensemble accuracy. With this, we shine first light on the connection between intermediate representations and their impact on the output predictions.

Taming Detection Transformers for Medical Object Detection

The accurate detection of suspicious regions in medical images is an error-prone and time-consuming process required by many routinely performed diagnostic procedures. To support clinicians during this difficult task, several automated solutions were proposed relying on complex methods with many hyperparameters. In this study, we investigate the feasibility of DEtection TRansformer (DETR) models for volumetric medical object detection. In contrast to previous works, these models directly predict a set of objects without relying on the design of anchors or manual heuristics such as non-maximum-suppression to detect objects. We show by conducting extensive experiments with three models, namely DETR, Conditional DETR, and DINO DETR on four data sets (CADA, RibFrac, KiTS19, and LIDC) that these set prediction models can perform on par with or even better than currently existing methods. DINO DETR, the best-performing model in our experiments demonstrates this by outperforming a strong anchor-based one-stage detector, Retina U-Net, on three out of four data sets.