Explainable Artificial Intelligence (XAI)

DeepDiff-SHAP: Interpretable deep learning for subgroup-specific causal inference using conditional SHAP

DeepDiff-SHAP identifies subgroup-specific causal relationships in biomedical data using deep learning and SHAP, revealing clinically meaningful changes in feature interactions across patient subgroups.

Trustworthy detection of exencephaly in high-throughput micro-CT embryo screens with focal-loss transformers

The paper develops a focal-loss transformer model to detect exencephaly in imbalanced micro-CT embryo data, achieving high accuracy and interpretable saliency maps.

Identifying intervention strategies from machine learning models with COALA: a counterfactual optimization framework

COALA identifies optimal counterfactual interventions for individuals using gradient boosted trees on NHANES data, achieving 85.4% accuracy in predicting feature changes.

XAI-based Data Visualization in Multimodal Medical Data

This paper examines advanced XAI techniques for multimodal medical data, focusing on perturbation-based, concept-based, and example-based explanations to improve model interpretability in healthcare.

BenchXAI: Comprehensive Benchmarking of Post-hoc Explainable AI Methods on Multi-Modal Biomedical Data

BenchXAI benchmarks fifteen post-hoc XAI methods on multi-modal biomedical data, finding some methods perform well across tasks while others struggle.

Adaptive Example Selection: Prototype-Based Explainability for Interpretable Mitosis Detection

This paper introduces Adaptive Example Selection (AES), an XAI framework for mitosis detection that enhances model interpretability by comparing test cases with annotated prototypes.

Biophysical validation of explainable AI for functional brain imaging: bridging cellular mechanisms and network dynamics

This paper validates an explainable AI framework to identify brain regions affected by excitation/inhibition imbalance using simulations and empirical data from autism.

Sensitivity based model agnostic scalable explanations of deep learning

This paper introduces SensX, an XAI framework that improves explanation accuracy and speed for deep learning models, validated on vision transformers and single-cell RNA-seq data.

Explainable machine learning for health disparities: type 2 diabetes in the All of Us research program

This paper uses explainable machine learning to identify and quantify T2D risk factors contributing to health disparities among different racial and ethnic groups in the All of Us research program.

Ensemble Post-hoc Explainable AI for Multilead ECG: Identifying Disease-Relevant Features in Single-Lead Interpretations

The paper evaluates XAI methods for a 1D-ResNet ECG classifier, finding some methods consistently highlight clinically relevant features while others do not.

Quantifying Interpretation Reproducibility in Vision Transformer Models with TAVAC

The paper introduces TAVAC, a metric to quantify overfitting and interpretation reproducibility in Vision Transformer models using attention consistency across training and validation data.

Interpretable and explainable predictive machine learning models for data-driven protein engineering

This paper explores integrating explainable AI techniques into machine learning models for predicting protein properties, demonstrating their effectiveness in identifying DNA binding proteins and optimizing Green Fluorescent Protein brightness.

Quantitative Attributions with Counterfactuals

This paper introduces QuAC, a method for explaining deep neural network decisions on images by generating counterfactuals and attributing class differences, and evaluates it on synthetic and biological datasets.

Comparison of Explainable AI Models for MRI-based Alzheimer's Disease Classification

This study compares explainable AI models for Alzheimer's disease classification using MRI scans, evaluating interpretability methods and generalization across different populations.

Chimeric U-Net - Modifying the standard U-Net towards Explainability

The paper introduces Chimeric U-Net, an invertible decoder U-Net architecture that enhances explainability in semantic segmentation without compromising performance.

The utility of explainable A.I. for MRI analysis: Relating model predictions to neuroimaging features of the aging brain

This study uses explainable AI to link brain-age prediction models with neuroimaging features of aging, revealing correlations between voxel relevance and structural brain markers.

Exploiting XAI maps to improve MS lesion segmentation and detection in MRI

This paper improves MS lesion segmentation and detection in MRI by using saliency maps to refine prediction scores, achieving higher F1 and PPV scores.

Visual interpretability of image-based classification models by generative latent space disentanglement applied to in vitro fertilization

This paper introduces DISCOVER, a generative model that disentangles latent spaces to provide interpretable explanations for image-based classification models in in vitro fertilization.

Feature Importance in Predicting Clinical Outcome: Statistics vs. Explainable Artificial Intelligence

The paper compares XAI and statistical methods to identify important features predicting neoadjuvant chemotherapy response in breast cancer, finding complementary insights.

Identifying Reproducibly Important EEG Markers of Schizophrenia with an Explainable Multi-Model Deep Learning Approach

This study identifies reproducibly important EEG markers of schizophrenia using an explainable multi-model deep learning approach and finds effects on beta, theta frequency bands, left hemisphere, and interhemispheric interactions.

Interpreting cis-regulatory mechanisms from genomic deep neural networks using surrogate models

This paper introduces SQUID, a framework for interpreting genomic deep neural networks using surrogate models, showing improved motif consistency and variant-effect predictions.

Motif Interactions Affect Post-Hoc Interpretability of Genomic Convolutional Neural Networks

The paper examines how motif interactions affect post-hoc interpretability of genomic CNNs and shows that interpretability methods can miss motifs when interactions are present.

Community-aware explanations in knowledge graphs with XP-GNN

This paper introduces XP-GNN, a method for explaining Graph Neural Networks in biological knowledge graphs by leveraging community structures, showing improved relevance in drug discovery use cases.

Quick and effective approximation of in silico saturation mutagenesis experiments with first-order Taylor expansion

The paper introduces a method to approximate in silico saturation mutagenesis using first-order Taylor expansion, reducing computational cost while maintaining robustness.

Reliable interpretability of biology-inspired deep neural networks

This study evaluates interpretability in biology-inspired deep learning models, identifying biases and proposing methods to improve robustness in P-NET trained on patient mutation data.

Discovering genetic biomarkers for targeted cancer therapeutics with eXplainable AI

This study uses XAI to discover genetic biomarkers for cancer therapeutics, identifying IL27RA as a potential target in high-grade serous ovarian cancer.

A Clear, Legible, Explainable, Transparent, and Elucidative (CLETE) Binary Classification Platform for Tabular Data

The paper introduces a CLETE binary classification platform for tabular data that offers explainable, transparent, and modifiable machine learning models with various feature selection, sampling, and evaluation methods.

Enabling interpretable machine learning for biological data with reliability scores

This paper introduces reliability scores to improve interpretability in biological machine learning by assessing classifier reliability in challenging scenarios.

Correcting gradient-based interpretations of deep neural networks for genomics

The paper corrects gradient-based interpretations of deep neural networks in genomics to reduce noise from off-simplex gradients, improving attribution method interpretability.

Interpretable machine learning with tree-based Shapley additive explanations: application to metabolomics datasets for binary classification

The paper compares SHAP and VIP for interpreting ML models in metabolomics, finding SHAP provides deeper explanations than VIP.