Insights into the inner workings of transformer models for protein function prediction.

MOTIVATION: We explored how explainable artificial intelligence (XAI) can help to shed light into the inner workings of neural networks for protein function prediction, by extending the widely used XAI method of integrated gradients such that latent representations inside of transformer models, which were finetuned to Gene Ontology term and Enzyme Commission number prediction, can be inspected too. RESULTS: The approach enabled us to identify amino acids in the sequences that the transformers pay particular attention to, and to show that these relevant sequence parts reflect expectations from biology and chemistry, both in the embedding layer and inside of the model, where we identified transformer heads with a statistically significant correspondence of attribution maps with ground truth sequence annotations (e.g. transmembrane regions, active sites) across many proteins. AVAILABILITY AND IMPLEMENTATION: Source code can be accessed at https://github.com/markuswenzel/xai-proteins.

Fall risk stratification of community-living older people. Commentary on the world guidelines for fall prevention and management.

The Task Force on Global Guidelines for Falls in Older Adults has put forward a fall risk stratification tool for community-dwelling older adults. This tool takes the form of a flowchart and is based on expert opinion and evidence. It divides the population into three risk categories and recommends specific preventive interventions or treatments for each category. In this commentary, we share our insights on the design, validation, usability and potential impact of this fall risk stratification tool with the aim of guiding future research.

From modern CNNs to vision transformers: Assessing the performance, robustness, and classification strategies of deep learning models in histopathology.

While machine learning is currently transforming the field of histopathology, the domain lacks a comprehensive evaluation of state-of-the-art models based on essential but complementary quality requirements beyond a mere classification accuracy. In order to fill this gap, we developed a new methodology to extensively evaluate a wide range of classification models, including recent vision transformers, and convolutional neural networks such as: ConvNeXt, ResNet (BiT), Inception, ViT and Swin transformer, with and without supervised or self-supervised pretraining. We thoroughly tested the models on five widely used histopathology datasets containing whole slide images of breast, gastric, and colorectal cancer and developed a novel approach using an image-to-image translation model to assess the robustness of a cancer classification model against stain variations. Further, we extended existing interpretability methods to previously unstudied models and systematically reveal insights of the models' classification strategies that allow for plausibility checks and systematic comparisons. The study resulted in specific model recommendations for practitioners as well as putting forward a general methodology to quantify a model's quality according to complementary requirements that can be transferred to future model architectures.

Comparison of Anatomical Preformed Titanium Implants and Patient-Specific CAD/CAM Implants in the Primary Reconstruction of Isolated Orbital Fractures-A Retrospective Study.

BACKGROUND/AIM: Reconstruction of the fractured orbit remains a challenge. The aim of this study was to compare anatomical preformed titanium orbital implants with patient-specific CAD/CAM implants for precision and intraoperative applicability. MATERIAL AND METHODS: A total of 75 orbital reconstructions from 2012 to 2022 were retrospectively assessed for their precision of implant position and intra- and postoperative revision rates. For this purpose, the implant position after digital orbital reconstruction was checked for deviations by mirroring the healthy orbit at 5 defined points, and the medical records of the patients were checked for revisions. RESULTS: The evaluation of the 45 anatomical preformed orbital implant cases showed significantly higher deviations and an implant inaccuracy of 66.6% than the 30 CAD/CAM cases with only 10% inaccuracy. In particular, the CAD/CAM implants were significantly more precise in medial and posterior positioning. In addition, the intraoperative revision rates of 26.6% vs. 11% after 3D intraoperative imaging and the postoperative revision rates of 13% vs. 0 for the anatomical preformed implants were significantly higher than for patient-specific implants. CONCLUSION: We conclude that patient-specific CAD/CAM orbital implants are highly suitable for primary orbital reconstruction. These seem to be preferable to anatomical preformed implants in terms of precision and revision rates.

Machine Learning for Health: Algorithm Auditing & Quality Control.

Developers proposing new machine learning for health (ML4H) tools often pledge to match or even surpass the performance of existing tools, yet the reality is usually more complicated. Reliable deployment of ML4H to the real world is challenging as examples from diabetic retinopathy or Covid-19 screening show. We envision an integrated framework of algorithm auditing and quality control that provides a path towards the effective and reliable application of ML systems in healthcare. In this editorial, we give a summary of ongoing work towards that vision and announce a call for participation to the special issue  Machine Learning for Health: Algorithm Auditing & Quality Control in this journal to advance the practice of ML4H auditing.

Evaluation framework to guide implementation of AI systems into healthcare settings.

OBJECTIVES: To date, many artificial intelligence (AI) systems have been developed in healthcare, but adoption has been limited. This may be due to inappropriate or incomplete evaluation and a lack of internationally recognised AI standards on evaluation. To have confidence in the generalisability of AI systems in healthcare and to enable their integration into workflows, there is a need for a practical yet comprehensive instrument to assess the translational aspects of the available AI systems. Currently available evaluation frameworks for AI in healthcare focus on the reporting and regulatory aspects but have little guidance regarding assessment of the translational aspects of the AI systems like the functional, utility and ethical components. METHODS: To address this gap and create a framework that assesses real-world systems, an international team has developed a translationally focused evaluation framework termed 'Translational Evaluation of Healthcare AI (TEHAI)'. A critical review of literature assessed existing evaluation and reporting frameworks and gaps. Next, using health technology evaluation and translational principles, reporting components were identified for consideration. These were independently reviewed for consensus inclusion in a final framework by an international panel of eight expert. RESULTS: TEHAI includes three main components: capability, utility and adoption. The emphasis on translational and ethical features of the model development and deployment distinguishes TEHAI from other evaluation instruments. In specific, the evaluation components can be applied at any stage of the development and deployment of the AI system. DISCUSSION: One major limitation of existing reporting or evaluation frameworks is their narrow focus. TEHAI, because of its strong foundation in translation research models and an emphasis on safety, translational value and generalisability, not only has a theoretical basis but also practical application to assessing real-world systems. CONCLUSION: The translational research theoretic approach used to develop TEHAI should see it having application not just for evaluation of clinical AI in research settings, but more broadly to guide evaluation of working clinical systems.

Wheat Anaphylaxis in Adults Differs from Reactions to Other Types of Food.

BACKGROUND: Wheat is one of the most commonly consumed foods and a known elicitor of anaphylaxis in children and adults. Reactions in adults are often cofactor dependent and characterized by a prolonged time between food intake and the onset of symptoms making the diagnosis of wheat anaphylaxis challenging. OBJECTIVE: To characterize a cohort of patients with the history of wheat anaphylaxis to better understand this atypical phenotype of anaphylaxis. METHODS: Data from the European Anaphylaxis Registry from 2007 to 2019 (n = 10,636) including 250 patients (213 adults and 37 children) with a history of anaphylaxis caused by wheat were analyzed. RESULTS: Wheat was the most common food elicitor of anaphylaxis in adults in the registry in Central Europe. Reactions to wheat in adults were frequently associated with exercise as a cofactor (82.8%) and partially delayed (57.5%). Only 36.9% of patients had atopic comorbidities, which was uncommonly low for adult patients allergic to other kinds of foods (63.2%). Anaphylaxis to wheat presented frequently with cardiovascular symptoms (86.7%) including severe symptoms such as loss of consciousness (41%) and less often with respiratory symptoms (53.6%). The reactions to wheat were more severe than reactions to other foods (odds ratio [OR] = 4.33), venom (OR = 1.58), or drugs (OR = 2.11). CONCLUSIONS: Wheat is a relevant elicitor of anaphylaxis in adults in Central Europe. Wheat anaphylaxis is highly dependent on the presence of cofactors and less frequently associated with atopic diseases compared with other food allergies. More data on mechanisms of wheat-induced anaphylaxis are required to develop preventive measures for this potentially life-threatening disease.

Risk estimation of SARS-CoV-2 transmission from bluetooth low energy measurements.

Digital contact tracing approaches based on Bluetooth low energy (BLE) have the potential to efficiently contain and delay outbreaks of infectious diseases such as the ongoing SARS-CoV-2 pandemic. In this work we propose a machine learning based approach to reliably detect subjects that have spent enough time in close proximity to be at risk of being infected. Our study is an important proof of concept that will aid the battery of epidemiological policies aiming to slow down the rapid spread of COVID-19.

USMPep: universal sequence models for major histocompatibility complex binding affinity prediction.

BACKGROUND: Immunotherapy is a promising route towards personalized cancer treatment. A key algorithmic challenge in this process is to decide if a given peptide (neoepitope) binds with the major histocompatibility complex (MHC). This is an active area of research and there are many MHC binding prediction algorithms that can predict the MHC binding affinity for a given peptide to a high degree of accuracy. However, most of the state-of-the-art approaches make use of complicated training and model selection procedures, are restricted to peptides of a certain length and/or rely on heuristics. RESULTS: We put forward USMPep, a simple recurrent neural network that reaches state-of-the-art approaches on MHC class I binding prediction with a single, generic architecture and even a single set of hyperparameters both on IEDB benchmark datasets and on the very recent HPV dataset. Moreover, the algorithm is competitive for a single model trained from scratch, while ensembling multiple regressors and language model pretraining can still slightly improve the performance. The direct application of the approach to MHC class II binding prediction shows a solid performance despite of limited training data. CONCLUSIONS: We demonstrate that competitive performance in MHC binding affinity prediction can be reached with a standard architecture and training procedure without relying on any heuristics.

UDSMProt: universal deep sequence models for protein classification.

MOTIVATION: Inferring the properties of a protein from its amino acid sequence is one of the key problems in bioinformatics. Most state-of-the-art approaches for protein classification are tailored to single classification tasks and rely on handcrafted features, such as position-specific-scoring matrices from expensive database searches. We argue that this level of performance can be reached or even be surpassed by learning a task-agnostic representation once, using self-supervised language modeling, and transferring it to specific tasks by a simple fine-tuning step. RESULTS: We put forward a universal deep sequence model that is pre-trained on unlabeled protein sequences from Swiss-Prot and fine-tuned on protein classification tasks. We apply it to three prototypical tasks, namely enzyme class prediction, gene ontology prediction and remote homology and fold detection. The proposed method performs on par with state-of-the-art algorithms that were tailored to these specific tasks or, for two out of three tasks, even outperforms them. These results stress the possibility of inferring protein properties from the sequence alone and, on more general grounds, the prospects of modern natural language processing methods in omics. Moreover, we illustrate the prospects for explainable machine learning methods in this field by selected case studies. AVAILABILITY AND IMPLEMENTATION: Source code is available under https://github.com/nstrodt/UDSMProt. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Integrating neurophysiologic relevance feedback in intent modeling for information retrieval.

The use of implicit relevance feedback from neurophysiology could deliver effortless information retrieval. However, both computing neurophysiologic responses and retrieving documents are characterized by uncertainty because of noisy signals and incomplete or inconsistent representations of the data. We present the first-of-its-kind, fully integrated information retrieval system that makes use of online implicit relevance feedback generated from brain activity as measured through electroencephalography (EEG), and eye movements. The findings of the evaluation experiment (N = 16) show that we are able to compute online neurophysiology-based relevance feedback with performance significantly better than chance in complex data domains and realistic search tasks. We contribute by demonstrating how to integrate in interactive intent modeling this inherently noisy implicit relevance feedback combined with scarce explicit feedback. Although experimental measures of task performance did not allow us to demonstrate how the classification outcomes translated into search task performance, the experiment proved that our approach is able to generate relevance feedback from brain signals and eye movements in a realistic scenario, thus providing promising implications for future work in neuroadaptive information retrieval (IR).

Automatic classification of dopamine transporter SPECT: deep convolutional neural networks can be trained to be robust with respect to variable image characteristics.

PURPOSE: This study investigated the potential of deep convolutional neural networks (CNN) for automatic classification of FP-CIT SPECT in multi-site or multi-camera settings with variable image characteristics. METHODS: The study included FP-CIT SPECT of 645 subjects from the Parkinson's Progression Marker Initiative (PPMI), 207 healthy controls, and 438 Parkinson's disease patients. SPECT images were smoothed with an isotropic 18-mm Gaussian kernel resulting in 3 different PPMI settings: (i) original (unsmoothed), (ii) smoothed, and (iii) mixed setting comprising all original and all smoothed images. A deep CNN with 2,872,642 parameters was trained, validated, and tested separately for each setting using 10 random splits with 60/20/20% allocation to training/validation/test sample. The putaminal specific binding ratio (SBR) was computed using a standard anatomical ROI predefined in MNI space (AAL atlas) or using the hottest voxels (HV) analysis. Both SBR measures were trained (ROC analysis, Youden criterion) using the same random splits as for the CNN. CNN and SBR trained in the mixed PPMI setting were also tested in an independent sample from clinical routine patient care (149 with non-neurodegenerative and 149 with neurodegenerative parkinsonian syndrome). RESULTS: Both SBR measures performed worse in the mixed PPMI setting compared to the pure PPMI settings (e.g., AAL-SBR accuracy = 0.900 ± 0.029 in the mixed setting versus 0.957 ± 0.017 and 0.952 ± 0.015 in original and smoothed setting, both p < 0.01). In contrast, the CNN showed similar accuracy in all PPMI settings (0.967 ± 0.018, 0.972 ± 0.014, and 0.955 ± 0.009 in mixed, original, and smoothed setting). Similar results were obtained in the clinical sample. After training in the mixed PPMI setting, only the CNN provided acceptable performance in the clinical sample. CONCLUSIONS: These findings provide proof of concept that a deep CNN can be trained to be robust with respect to variable site-, camera-, or scan-specific image characteristics without a large loss of diagnostic accuracy compared with mono-site/mono-camera settings. We hypothesize that a single CNN can be used to support the interpretation of FP-CIT SPECT at many different sites using different acquisition hardware and/or reconstruction software with only minor harmonization of acquisition and reconstruction protocols.

Implicit relevance feedback from electroencephalography and eye tracking in image search.

OBJECTIVE: Methods from brain-computer interfacing (BCI) open a direct access to the mental processes of computer users, which offers particular benefits in comparison to standard methods for inferring user-related information. The signals can be recorded unobtrusively in the background, which circumvents the time-consuming and distracting need for the users to give explicit feedback to questions concerning the individual interest. The obtained implicit information makes it possible to create dynamic user interest profiles in real-time, that can be taken into account by novel types of adaptive, personalised software. In the present study, the potential of implicit relevance feedback from electroencephalography (EEG) and eye tracking was explored with a demonstrator application that simulated an image search engine. APPROACH: The participants of the study queried for ambiguous search terms, having in mind one of the two possible interpretations of the respective term. Subsequently, they viewed different images arranged in a grid that were related to the query. The ambiguity of the underspecified search term was resolved with implicit information present in the recorded signals. For this purpose, feature vectors were extracted from the signals and used by multivariate classifiers that estimated the intended interpretation of the ambiguous query. MAIN RESULT: The intended interpretation was inferred correctly from a combination of EEG and eye tracking signals in 86% of the cases on average. Information provided by the two measurement modalities turned out to be complementary. SIGNIFICANCE: It was demonstrated that BCI methods can extract implicit user-related information in a setting of human-computer interaction. Novelties of the study are the implicit online feedback from EEG and eye tracking, the approximation to a realistic use case in a simulation, and the presentation of a large set of photographies that had to be interpreted with respect to the content.

The Berlin Brain-Computer Interface: Progress Beyond Communication and Control.

The combined effect of fundamental results about neurocognitive processes and advancements in decoding mental states from ongoing brain signals has brought forth a whole range of potential neurotechnological applications. In this article, we review our developments in this area and put them into perspective. These examples cover a wide range of maturity levels with respect to their applicability. While we assume we are still a long way away from integrating Brain-Computer Interface (BCI) technology in general interaction with computers, or from implementing neurotechnological measures in safety-critical workplaces, results have already now been obtained involving a BCI as research tool. In this article, we discuss the reasons why, in some of the prospective application domains, considerable effort is still required to make the systems ready to deal with the full complexity of the real world.

Is Neural Activity Detected by ERP-Based Brain-Computer Interfaces Task Specific?

OBJECTIVE: Brain-computer interfaces (BCIs) that are based on event-related potentials (ERPs) can estimate to which stimulus a user pays particular attention. In typical BCIs, the user silently counts the selected stimulus (which is repeatedly presented among other stimuli) in order to focus the attention. The stimulus of interest is then inferred from the electroencephalogram (EEG). Detecting attention allocation implicitly could be also beneficial for human-computer interaction (HCI), because it would allow software to adapt to the user's interest. However, a counting task would be inappropriate for the envisaged implicit application in HCI. Therefore, the question was addressed if the detectable neural activity is specific for silent counting, or if it can be evoked also by other tasks that direct the attention to certain stimuli. APPROACH: Thirteen people performed a silent counting, an arithmetic and a memory task. The tasks required the subjects to pay particular attention to target stimuli of a random color. The stimulus presentation was the same in all three tasks, which allowed a direct comparison of the experimental conditions. RESULTS: Classifiers that were trained to detect the targets in one task, according to patterns present in the EEG signal, could detect targets in all other tasks (irrespective of some task-related differences in the EEG). SIGNIFICANCE: The neural activity detected by the classifiers is not strictly task specific but can be generalized over tasks and is presumably a result of the attention allocation or of the augmented workload. The results may hold promise for the transfer of classification algorithms from BCI research to implicit relevance detection in HCI.

Classification of Eye Fixation Related Potentials for Variable Stimulus Saliency.

OBJECTIVE: Electroencephalography (EEG) and eye tracking can possibly provide information about which items displayed on the screen are relevant for a person. Exploiting this implicit information promises to enhance various software applications. The specific problem addressed by the present study is that items shown in real applications are typically diverse. Accordingly, the saliency of information, which allows to discriminate between relevant and irrelevant items, varies. As a consequence, recognition can happen in foveal or in peripheral vision, i.e., either before or after the saccade to the item. Accordingly, neural processes related to recognition are expected to occur with a variable latency with respect to the eye movements. The aim was to investigate if relevance estimation based on EEG and eye tracking data is possible despite of the aforementioned variability. APPROACH: Sixteen subjects performed a search task where the target saliency was varied while the EEG was recorded and the unrestrained eye movements were tracked. Based on the acquired data, it was estimated which of the items displayed were targets and which were distractors in the search task. RESULTS: Target prediction was possible also when the stimulus saliencies were mixed. Information contained in EEG and eye tracking data was found to be complementary and neural signals were captured despite of the unrestricted eye movements. The classification algorithm was able to cope with the experimentally induced variable timing of neural activity related to target recognition. SIGNIFICANCE: It was demonstrated how EEG and eye tracking data can provide implicit information about the relevance of items on the screen for potential use in online applications.

EEG-based usability assessment of 3D shutter glasses.

OBJECTIVE: Neurotechnology can contribute to the usability assessment of products by providing objective measures of neural workload and can uncover usability impediments that are not consciously perceived by test persons. In this study, the neural processing effort imposed on the viewer of 3D television by shutter glasses was quantified as a function of shutter frequency. In particular, we sought to determine the critical shutter frequency at which the 'neural flicker' vanishes, such that visual fatigue due to this additional neural effort can be prevented by increasing the frequency of the system. APPROACH: Twenty-three participants viewed an image through 3D shutter glasses, while multichannel electroencephalogram (EEG) was recorded. In total ten shutter frequencies were employed, selected individually for each participant to cover the range below, at and above the threshold of flicker perception. The source of the neural flicker correlate was extracted using independent component analysis and the flicker impact on the visual cortex was quantified by decoding the state of the shutter from the EEG. MAIN RESULT: Effects of the shutter glasses were traced in the EEG up to around 67 Hz-about 20 Hz over the flicker perception threshold-and vanished at the subsequent frequency level of 77 Hz. SIGNIFICANCE: The impact of the shutter glasses on the visual cortex can be detected by neurotechnology even when a flicker is not reported by the participants. Potential impact. Increasing the shutter frequency from the usual 50 Hz or 60 Hz to 77 Hz reduces the risk of visual fatigue and thus improves shutter-glass-based 3D usability.

Randomized controlled trial of noninvasive positive pressure ventilation (NPPV) versus servoventilation in patients with CPAP-induced central sleep apnea (complex sleep apnea).

STUDY OBJECTIVES: To compare the treatment effect of noninvasive positive pressure ventilation (NPPV) and anticyclic servoventilation in patients with continuous positive airway pressure (CPAP)-induced central sleep apnea (complex sleep apnea). DESIGN: Randomized controlled trial. SETTING: Sleep center. PATIENTS: Thirty patients who developed complex sleep apnea syndrome (CompSAS) during CPAP treatment. INTERVENTIONS: NPPV or servoventilation. MEASUREMENTS AND RESULTS: Patients were randomized to NPPV or servo-ventilation. Full polysomnography (PSG) was performed after 6 weeks. On CPAP prior to randomization, patients in the NPPV and servoventilator arm had comparable apnea-hypopnea indices (AHI, 28.6 ± 6.5 versus 27.7 ± 9.7 events/h (mean ± standard deviation [SD])), apnea indices (AI,19 ± 5.6 versus 21.1 ± 8.6 events/h), central apnea indices (CAI, 16.7 ± 5.4 versus 18.2 ± 7.1 events/h), oxygen desaturation indices (ODI,17.5 ± 13.1 versus 24.3 ± 11.9 events/h). During initial titration NPPV and servoventilation significantly improved the AHI (9.1 ± 4.3 versus 9 ± 6.4 events/h), AI (2 ± 3.1 versus 3.5 ± 4.5 events/h) CAI (2 ± 3.1 versus 2.5 ± 3.9 events/h) and ODI (10.1 ± 4.5 versus 8.9 ± 8.4 events/h) when compared to CPAP treatment (all P < 0.05). After 6 weeks we observed the following differences: AHI (16.5 ± 8 versus 7.4 ± 4.2 events/h, P = 0.027), AI (10.4 ± 5.9 versus 1.7 ± 1.9 events/h, P = 0.001), CAI (10.2 ± 5.1 versus 1.5 ± 1.7 events/h, P < 0.0001)) and ODI (21.1 ± 9.2 versus 4.8 ± 3.4 events/h, P < 0.0001) for NPPV and servoventilation, respectively. Other sleep parameters were unaffected by any form of treatment. CONCLUSIONS: After 6 weeks, servoventilation treated respiratory events more effectively than NPPV in patients with complex sleep apnea syndrome.

Visuomotor functional network topology predicts upcoming tasks.

It is a vital ability of humans to flexibly adapt their behavior to different environmental situations. Constantly, the rules for our sensory-to-motor mappings need to be adapted to the current task demands. For example, the same sensory input might require two different motor responses depending on the actual situation. How does the brain prepare for such different responses? It has been suggested that the functional connections within cortex are biased according to the present rule to guide the flow of information in accordance with the required sensory-to-motor mapping. Here, we investigated with fMRI whether task settings might indeed change the functional connectivity structure in a large-scale brain network. Subjects performed a visuomotor response task that required an interaction between visual and motor cortex: either within each hemisphere or across the two hemispheres of the brain depending on the task condition. A multivariate analysis on the functional connectivity graph of a cortical visuomotor network revealed that the functional integration, i.e., the connectivity structure, is altered according to the task condition already during a preparatory period before the visual cue and the actual movement. Our results show that the topology of connection weights within a single network changes according to and thus predicts the upcoming task. This suggests that the human brain prepares to respond in different conditions by altering its large scale functional connectivity structure even before an action is required.