Explainable Deep Learning for Personalized Age Prediction With Brain Morphology

A novel explainable machine learning-based healthy ageing scale

BACKGROUND

Ageing is one of the most important challenges in our society. Evaluating how one is ageing is important in many aspects, from giving personalized recommendations to providing insight for long-term care eligibility. Machine learning can be utilized for that purpose, however, user reservations towards "black-box" predictions call for increased transparency and explainability of results. This study aimed to explore the potential of developing a machine learning-based healthy ageing scale that provides explainable results that could be trusted and understood by informal carers.

METHODS

In this study, we used data from 696 older adults collected via personal field interviews as part of independent research. Explanatory factor analysis was used to find candidate healthy ageing aspects. For visualization of key aspects, a web annotation application was developed. Key aspects were selected by gerontologists who later used web annotation applications to evaluate healthy ageing for each older adult on a Likert scale. Logistic Regression, Decision Tree Classifier, Random Forest, KNN, SVM and XGBoost were used for multi-classification machine learning. AUC OvO, AUC OvR, F1, Precision and Recall were used for evaluation. Finally, SHAP was applied to best model predictions to make them explainable.

RESULTS

The experimental results show that human annotations of healthy ageing could be modelled using machine learning where among several algorithms XGBoost showed superior performance. The use of XGBoost resulted in 0.92 macro-averaged AuC OvO and 0.76 macro-averaged F1. SHAP was applied to generate local explanations for predictions and shows how each feature is influencing the prediction.

CONCLUSION

The resulting explainable predictions make a step toward practical scale implementation into decision support systems. The development of such a decision support system that would incorporate an explainable model could reduce user reluctance towards the utilization of AI in healthcare and provide explainable and trusted insights to informal carers or healthcare providers as a basis to shape tangible actions for improving ageing. Furthermore, the cooperation with gerontology specialists throughout the process also indicates expert knowledge as integrated into the model.

Exploring the correlation between DNA methylation and biological age using an interpretable machine learning framework

DNA methylation plays a significant role in regulating transcription and exhibits a systematic change with age. These changes can be used to predict an individual's age. First, to identify methylation sites associated with biological age; second, to construct a biological age prediction model and preliminarily explore the biological significance of methylation-associated genes using machine learning. A biological age prediction model was constructed using human methylation data through data preprocessing, feature selection procedures, statistical analysis, and machine learning techniques. Subsequently, 15 methylation data sets were subjected to in-depth analysis using SHAP, GO enrichment, and KEGG analysis. XGBoost, LightGBM, and CatBoost identified 15 groups of methylation sites associated with biological age. The cg23995914 locus was identified as the most significant contributor to predicting biological age by calculating SHAP values. Furthermore, GO enrichment and KEGG analyses were employed to initially explore the methylated loci's biological significance.

A perspective on brain-age estimation and its clinical promise

Brain-age estimation has gained increased attention in the neuroscientific community owing to its potential use as a biomarker of brain health. The difference between estimated and chronological age based on neuroimaging data enables a unique perspective on brain development and aging, with multiple open questions still remaining in the brain-age research field. This Perspective presents an overview of current advancements in the field and envisions the future evolution of the brain-age framework before its potential deployment in hospital settings.

Personalized identification of autism-related bacteria in the gut microbiome using explainable artificial intelligence

Autism spectrum disorder (ASD) affects social interaction and communication. Emerging evidence links ASD to gut microbiome alterations, suggesting that microbial composition may play a role in the disorder. This study employs explainable artificial intelligence (XAI) to examine the contributions of individual microbial species to ASD. By using local explanation embeddings and unsupervised clustering, the research identifies distinct ASD subgroups, underscoring the disorder's heterogeneity. Specific microbial biomarkers associated with ASD are revealed, and the best classifiers achieved an AU-ROC of 0.965 ± 0.005 and an AU-PRC of 0.967 ± 0.008. The findings support the notion that gut microbiome composition varies significantly among individuals with ASD. This work's broader significance lies in its potential to inform personalized interventions, enhancing precision in ASD management and classification. These insights highlight the importance of individualized microbiome profiles for developing tailored therapeutic strategies for ASD.

Older Tissue Age Derived From Abdominal Computed Tomography Biomarkers of Muscle, Fat, and Bone Is Associated With Chronic Conditions and Higher Mortality

OBJECTIVE

To determine whether body composition derived from medical imaging may be useful for assessing biologic age at the tissue level because people of the same chronologic age may vary with respect to their biologic age.

METHODS

We identified an age- and sex-stratified cohort of 4900 persons with an abdominal computed tomography scan from January 1, 2010, to December 31, 2020, who were 20 to 89 years old and representative of the general population in Southeast Minnesota and West Central Wisconsin. We constructed a model for estimating tissue age that included 6 body composition biomarkers calculated from abdominal computed tomography using a previously validated deep learning model.

RESULTS

Older tissue age associated with intermediate subcutaneous fat area, higher visceral fat area, lower muscle area, lower muscle density, higher bone area, and lower bone density. A tissue age older than chronologic age was associated with chronic conditions that result in reduced physical fitness (including chronic obstructive pulmonary disease, arthritis, cardiovascular disease, and behavioral disorders). Furthermore, a tissue age older than chronologic age was associated with an increased risk of death (hazard ratio, 1.56; 95% CI, 1.33 to 1.84) that was independent of demographic characteristics, county of residency, education, body mass index, and baseline chronic conditions.

CONCLUSION

Imaging-based body composition measures may be useful in understanding the biologic processes underlying accelerated aging.

Air pollution and mortality for cancer of the respiratory system in Italy: an explainable artificial intelligence approach

Respiratory system cancer, encompassing lung, trachea and bronchus cancer, constitute a substantial and evolving public health challenge. Since pollution plays a prominent cause in the development of this disease, identifying which substances are most harmful is fundamental for implementing policies aimed at reducing exposure to these substances. We propose an approach based on explainable artificial intelligence (XAI) based on remote sensing data to identify the factors that most influence the prediction of the standard mortality ratio (SMR) for respiratory system cancer in the Italian provinces using environment and socio-economic data. First of all, we identified 10 clusters of provinces through the study of the SMR variogram. Then, a Random Forest regressor is used for learning a compact representation of data. Finally, we used XAI to identify which features were most important in predicting SMR values. Our machine learning analysis shows that NO, income and O3 are the first three relevant features for the mortality of this type of cancer, and provides a guideline on intervention priorities in reducing risk factors.

Explainable artificial intelligence for microbiome data analysis in colorectal cancer biomarker identification

Background

Colorectal cancer (CRC) is a type of tumor caused by the uncontrolled growth of cells in the mucosa lining the last part of the intestine. Emerging evidence underscores an association between CRC and gut microbiome dysbiosis. The high mortality rate of this cancer has made it necessary to develop new early diagnostic methods. Machine learning (ML) techniques can represent a solution to evaluate the interaction between intestinal microbiota and host physiology. Through explained artificial intelligence (XAI) it is possible to evaluate the individual contributions of microbial taxonomic markers for each subject. Our work also implements the Shapley Method Additive Explanations (SHAP) algorithm to identify for each subject which parameters are important in the context of CRC.

Results

The proposed study aimed to implement an explainable artificial intelligence framework using both gut microbiota data and demographic information from subjects to classify a cohort of control subjects from those with CRC. Our analysis revealed an association between gut microbiota and this disease. We compared three machine learning algorithms, and the Random Forest (RF) algorithm emerged as the best classifier, with a precision of 0.729 ± 0.038 and an area under the Precision-Recall curve of 0.668 ± 0.016. Additionally, SHAP analysis highlighted the most crucial variables in the model's decision-making, facilitating the identification of specific bacteria linked to CRC. Our results confirmed the role of certain bacteria, such as Fusobacterium, Peptostreptococcus, and Parvimonas, whose abundance appears notably associated with the disease, as well as bacteria whose presence is linked to a non-diseased state.

Discussion

These findings emphasizes the potential of leveraging gut microbiota data within an explainable AI framework for CRC classification. The significant association observed aligns with existing knowledge. The precision exhibited by the RF algorithm reinforces its suitability for such classification tasks. The SHAP analysis not only enhanced interpretability but identified specific bacteria crucial in CRC determination. This approach opens avenues for targeted interventions based on microbial signatures. Further exploration is warranted to deepen our understanding of the intricate interplay between microbiota and health, providing insights for refined diagnostic and therapeutic strategies.

eXplainable Artificial Intelligence (XAI) in aging clock models

XAI is a rapidly progressing field of machine learning, aiming to unravel the predictions of complex models. XAI is especially required in sensitive applications, e.g. in health care, when diagnosis, recommendations and treatment choices might rely on the decisions made by artificial intelligence systems. AI approaches have become widely used in aging research as well, in particular, in developing biological clock models and identifying biomarkers of aging and age-related diseases. However, the potential of XAI here awaits to be fully appreciated. We discuss the application of XAI for developing the "aging clocks" and present a comprehensive analysis of the literature categorized by the focus on particular physiological systems.

Model-agnostic explainable artificial intelligence tools for severity prediction and symptom analysis on Indian COVID-19 data

Introduction

The COVID-19 pandemic had a global impact and created an unprecedented emergency in healthcare and other related frontline sectors. Various Artificial-Intelligence-based models were developed to effectively manage medical resources and identify patients at high risk. However, many of these AI models were limited in their practical high-risk applicability due to their "black-box" nature, i.e., lack of interpretability of the model. To tackle this problem, Explainable Artificial Intelligence (XAI) was introduced, aiming to explore the "black box" behavior of machine learning models and offer definitive and interpretable evidence. XAI provides interpretable analysis in a human-compliant way, thus boosting our confidence in the successful implementation of AI systems in the wild.

Methods

In this regard, this study explores the use of model-agnostic XAI models, such as SHapley Additive exPlanations values (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME), for COVID-19 symptom analysis in Indian patients toward a COVID severity prediction task. Various machine learning models such as Decision Tree Classifier, XGBoost Classifier, and Neural Network Classifier are leveraged to develop Machine Learning models.

Results and discussion

The proposed XAI tools are found to augment the high performance of AI systems with human interpretable evidence and reasoning, as shown through the interpretation of various explainability plots. Our comparative analysis illustrates the significance of XAI tools and their impact within a healthcare context. The study suggests that SHAP and LIME analysis are promising methods for incorporating explainability in model development and can lead to better and more trustworthy ML models in the future.

A scoping review of interpretability and explainability concerning artificial intelligence methods in medical imaging

PURPOSE

To review eXplainable Artificial Intelligence/(XAI) methods available for medical imaging/(MI).

METHOD

A scoping review was conducted following the Joanna Briggs Institute's methodology. The search was performed on Pubmed, Embase, Cinhal, Web of Science, BioRxiv, MedRxiv, and Google Scholar. Studies published in French and English after 2017 were included. Keyword combinations and descriptors related to explainability, and MI modalities were employed. Two independent reviewers screened abstracts, titles and full text, resolving differences through discussion.

RESULTS

228 studies met the criteria. XAI publications are increasing, targeting MRI (n = 73), radiography (n = 47), CT (n = 46). Lung (n = 82) and brain (n = 74) pathologies, Covid-19 (n = 48), Alzheimer's disease (n = 25), brain tumors (n = 15) are the main pathologies explained. Explanations are presented visually (n = 186), numerically (n = 67), rule-based (n = 11), textually (n = 11), and example-based (n = 6). Commonly explained tasks include classification (n = 89), prediction (n = 47), diagnosis (n = 39), detection (n = 29), segmentation (n = 13), and image quality improvement (n = 6). The most frequently provided explanations were local (78.1 %), 5.7 % were global, and 16.2 % combined both local and global approaches. Post-hoc approaches were predominantly employed. The used terminology varied, sometimes indistinctively using explainable (n = 207), interpretable (n = 187), understandable (n = 112), transparent (n = 61), reliable (n = 31), and intelligible (n = 3).

CONCLUSION

The number of XAI publications in medical imaging is increasing, primarily focusing on applying XAI techniques to MRI, CT, and radiography for classifying and predicting lung and brain pathologies. Visual and numerical output formats are predominantly used. Terminology standardisation remains a challenge, as terms like "explainable" and "interpretable" are sometimes being used indistinctively. Future XAI development should consider user needs and perspectives.

Machine learning approaches in microbiome research: challenges and best practices

Microbiome data predictive analysis within a machine learning (ML) workflow presents numerous domain-specific challenges involving preprocessing, feature selection, predictive modeling, performance estimation, model interpretation, and the extraction of biological information from the results. To assist decision-making, we offer a set of recommendations on algorithm selection, pipeline creation and evaluation, stemming from the COST Action ML4Microbiome. We compared the suggested approaches on a multi-cohort shotgun metagenomics dataset of colorectal cancer patients, focusing on their performance in disease diagnosis and biomarker discovery. It is demonstrated that the use of compositional transformations and filtering methods as part of data preprocessing does not always improve the predictive performance of a model. In contrast, the multivariate feature selection, such as the Statistically Equivalent Signatures algorithm, was effective in reducing the classification error. When validated on a separate test dataset, this algorithm in combination with random forest modeling, provided the most accurate performance estimates. Lastly, we showed how linear modeling by logistic regression coupled with visualization techniques such as Individual Conditional Expectation (ICE) plots can yield interpretable results and offer biological insights. These findings are significant for clinicians and non-experts alike in translational applications.

Explainable machine learning framework to predict personalized physiological aging

Attaining personalized healthy aging requires accurate monitoring of physiological changes and identifying subclinical markers that predict accelerated or delayed aging. Classic biostatistical methods most rely on supervised variables to estimate physiological aging and do not capture the full complexity of inter-parameter interactions. Machine learning (ML) is promising, but its black box nature eludes direct understanding, substantially limiting physician confidence and clinical usage. Using a broad population dataset from the National Health and Nutrition Examination Survey (NHANES) study including routine biological variables and after selection of XGBoost as the most appropriate algorithm, we created an innovative explainable ML framework to determine a Personalized physiological age (PPA). PPA predicted both chronic disease and mortality independently of chronological age. Twenty-six variables were sufficient to predict PPA. Using SHapley Additive exPlanations (SHAP), we implemented a precise quantitative associated metric for each variable explaining physiological (i.e., accelerated or delayed) deviations from age-specific normative data. Among the variables, glycated hemoglobin (HbA1c) displays a major relative weight in the estimation of PPA. Finally, clustering profiles of identical contextualized explanations reveal different aging trajectories opening opportunities to specific clinical follow-up. These data show that PPA is a robust, quantitative and explainable ML-based metric that monitors personalized health status. Our approach also provides a complete framework applicable to different datasets or variables, allowing precision physiological age estimation.

Explainable artificial intelligence incorporated with domain knowledge diagnosing early gastric neoplasms under white light endoscopy

White light endoscopy is the most pivotal tool for detecting early gastric neoplasms. Previous artificial intelligence (AI) systems were primarily unexplainable, affecting their clinical credibility and acceptability. We aimed to develop an explainable AI named ENDOANGEL-ED (explainable diagnosis) to solve this problem. A total of 4482 images and 296 videos with focal lesions from 3279 patients from eight hospitals were used for training, validating, and testing ENDOANGEL-ED. A traditional sole deep learning (DL) model was trained using the same dataset. The performance of ENDOANGEL-ED and sole DL was evaluated on six levels: internal and external images, internal and external videos, consecutive videos, and man-machine comparison with 77 endoscopists in videos. Furthermore, a multi-reader, multi-case study was conducted to evaluate the ENDOANGEL-ED's effectiveness. A scale was used to compare the overall acceptance of endoscopists to traditional and explainable AI systems. The ENDOANGEL-ED showed high performance in the image and video tests. In man-machine comparison, the accuracy of ENDOANGEL-ED was significantly higher than that of all endoscopists in internal (81.10% vs. 70.61%, p < 0.001) and external videos (88.24% vs. 78.49%, p < 0.001). With ENDOANGEL-ED's assistance, the accuracy of endoscopists significantly improved (70.61% vs. 79.63%, p < 0.001). Compared with the traditional AI, the explainable AI increased the endoscopists' trust and acceptance (4.42 vs. 3.74, p < 0.001; 4.52 vs. 4.00, p < 0.001). In conclusion, we developed a real-time explainable AI that showed high performance, higher clinical credibility, and acceptance than traditional DL models and greatly improved the diagnostic ability of endoscopists.

Detecting the socio-economic drivers of confidence in government with eXplainable Artificial Intelligence

The European Quality of Government Index (EQI) measures the perceived level of government quality by European Union citizens, combining surveys on corruption, impartiality and quality of provided services. It is, thus, an index based on individual subjective evaluations. Understanding the most relevant objective factors affecting the EQI outcomes is important for both evaluators and policy makers, especially in view of the fact that perception of government integrity contributes to determine the level of civic engagement. In our research, we employ methods of Artificial Intelligence and complex systems physics to measure the impact on the perceived government quality of multifaceted variables, describing territorial development and citizen well-being, from an economic, social and environmental viewpoint. Our study, focused on a set of regions in European Union at a subnational scale, leads to identifying the territorial and demographic drivers of citizens' confidence in government institutions. In particular, we find that the 2021 EQI values are significantly related to two indicators: the first one is the difference between female and male labour participation rates, and the second one is a proxy of wealth and welfare such as the average number of rooms per inhabitant. This result corroborates the idea of a central role played by labour gender equity and housing policies in government confidence building. In particular, the relevance of the former indicator in EQI prediction results from a combination of positive conditions such as equal job opportunities, vital labour market, welfare and availability of income sources, while the role of the latter is possibly amplified by the lockdown policies related to the COVID-19 pandemics. The analysis is based on combining regression, to predict EQI from a set of publicly available indicators, with the eXplainable Artificial Intelligence approach, that quantifies the impact of each indicator on the prediction. Such a procedure does not require any ad-hoc hypotheses on the functional dependence of EQI on the indicators used to predict it. Finally, using network science methods concerning community detection, we investigate how the impact of relevant indicators on EQI prediction changes throughout European regions. Thus, the proposed approach enables to identify the objective factors at the basis of government quality perception by citizens in different territorial contexts, providing the methodological basis for the development of a quantitative tool for policy design.

TIRESIA: An eXplainable Artificial Intelligence Platform for Predicting Developmental Toxicity

Herein, a robust and reproducible eXplainable Artificial Intelligence (XAI) approach is presented, which allows prediction of developmental toxicity, a challenging human-health endpoint in toxicology. The application of XAI as an alternative method is of the utmost importance with developmental toxicity being one of the most animal-intensive areas of regulatory toxicology. In this work, the established CAESAR (Computer Assisted Evaluation of industrial chemical Substances According to Regulations) training set made of 234 chemicals for model learning is employed. Two test sets, including as a whole 585 chemicals, were instead used for validation and generalization purposes. The proposed framework favorably compares with the state-of-the-art approaches in terms of accuracy, sensitivity, and specificity, thus resulting in a reliable support system for developmental toxicity ensuring informativeness, uncertainty estimation, generalization, and transparency. Based on the eXtreme Gradient Boosting (XGB) algorithm, our predictive model provides easy interpretative keys based on specific molecular descriptors and structural alerts enabling one to distinguish toxic and nontoxic chemicals. Inspired by the Organisation for Economic Co-operation and Development (OECD) principles for the validation of Quantitative Structure-Activity Relationships (QSARs) for regulatory purposes, the results are summarized in a standard report in portable document format, enclosing also details concerned with a density-based model applicability domain and SHAP (SHapley Additive exPlanations) explainability, the latter particularly useful to better understand the effective roles played by molecular features. Notably, our model has been implemented in TIRESIA (Toxicology Intelligence and Regulatory Evaluations for Scientific and Industry Applications), a free of charge web platform available at http://tiresia.uniba.it.

Worldwide impact of lifestyle predictors of dementia prevalence: An eXplainable Artificial Intelligence analysis

Introduction

Dementia is an umbrella term indicating a group of diseases that affect the cognitive sphere. Dementia is not a mere individual health issue, since its interference with the ability to carry out daily activities entails a series of collateral problems, comprising exclusion of patients from civil rights and welfare, unpaid caregiving work, mostly performed by women, and an additional burden on the public healthcare systems. Thus, gender and wealth inequalities (both among individuals and among countries) tend to amplify the social impact of such a disease. Since at present there is no cure for dementia but only drug treatments to slow down its progress and mitigate the symptoms, it is essential to work on prevention and early diagnosis, identifying the risk factors that increase the probability of its onset. The complex and multifactorial etiology of dementia, resulting from an interplay between genetics and environmental factors, can benefit from a multidisciplinary approach that follows the "One Health" guidelines of the World Health Organization.

Methods

In this work, we apply methods of Artificial Intelligence and complex systems physics to investigate the possibility to predict dementia prevalence throughout world countries from a set of variables concerning individual health, food consumption, substance use and abuse, healthcare system efficiency. The analysis uses publicly available indicator values at a country level, referred to a time window of 26 years.

Results

Employing methods based on eXplainable Artificial Intelligence (XAI) and complex networks, we identify a group of lifestyle factors, mostly concerning nutrition, that contribute the most to dementia incidence prediction.

Discussion

The proposed approach provides a methodological basis to develop quantitative tools for action patterns against such a disease, which involves issues deeply related with sustainable, such as good health and resposible food consumption.

The impact of harmonization on radiomic features in Parkinson's disease and healthy controls: A multicenter study

Radiomics is a challenging development area in imaging field that is greatly capturing interest of radiologists and neuroscientists. However, radiomics features show a strong non-biological variability determined by different facilities and imaging protocols, limiting the reproducibility and generalizability of analysis frameworks. Our study aimed to investigate the usefulness of harmonization to reduce site-effects on radiomics features over specific brain regions. We selected T1-weighted magnetic resonance imaging (MRI) by using the MRI dataset Parkinson's Progression Markers Initiative (PPMI) from different sites with healthy controls (HC) and Parkinson's disease (PD) patients. First, the investigation of radiomics measure discrepancies were assessed on healthy brain regions-of-interest (ROIs) via a classification pipeline based on LASSO feature selection and support vector machine (SVM) model. Then, a ComBat-based harmonization approach was applied to correct site-effects. Finally, a validation step on PD subjects evaluated diagnostic accuracy before and after harmonization of radiomics data. Results on healthy subjects demonstrated a dependence from site-effects that could be corrected with ComBat harmonization. LASSO regressor after harmonization was unable to select any feature to distinguish controls by site. Moreover, harmonized radiomics features achieved an area under the receiving operating characteristic curve (AUC) of 0.77 (compared to AUC of 0.71 for raw radiomics measures) in distinguish Parkinson's patients from HC. We found a not-negligible site-effect studying radiomics of HC pre- and post-harmonization of features. Our validation study on PD patients demonstrated a significant influence of non-biological noise source in diagnostic performances. Finally, harmonization of multicenter radiomic data represent a necessary step to make analysis pipelines reliable and replicable for multisite neuroimaging studies.

Explainable artificial intelligence (XAI) detects wildfire occurrence in the Mediterranean countries of Southern Europe

The impacts and threats posed by wildfires are dramatically increasing due to climate change. In recent years, the wildfire community has attempted to estimate wildfire occurrence with machine learning models. However, to fully exploit the potential of these models, it is of paramount importance to make their predictions interpretable and intelligible. This study is a first attempt to provide an eXplainable artificial intelligence (XAI) framework for estimating wildfire occurrence using a Random Forest model with Shapley values for interpretation. Our findings accurately detected regions with a high presence of wildfires (area under the curve 81.3%) and outlined the drivers empowering occurrence, such as the Fire Weather Index and Normalized Difference Vegetation Index. Furthermore, our analysis suggests the presence of anomalous hotspots. In contexts where human and natural spheres constantly intermingle and interact, the XAI framework, suitably integrated into decision support systems, could support forest managers to prevent and mitigate future wildfire disasters and develop strategies for effective fire management, response, recovery, and resilience.

A robust framework to investigate the reliability and stability of explainable artificial intelligence markers of Mild Cognitive Impairment and Alzheimer's Disease

In clinical practice, several standardized neuropsychological tests have been designed to assess and monitor the neurocognitive status of patients with neurodegenerative diseases such as Alzheimer's disease. Important research efforts have been devoted so far to the development of multivariate machine learning models that combine the different test indexes to predict the diagnosis and prognosis of cognitive decline with remarkable results. However, less attention has been devoted to the explainability of these models. In this work, we present a robust framework to (i) perform a threefold classification between healthy control subjects, individuals with cognitive impairment, and subjects with dementia using different cognitive indexes and (ii) analyze the variability of the explainability SHAP values associated with the decisions taken by the predictive models. We demonstrate that the SHAP values can accurately characterize how each index affects a patient's cognitive status. Furthermore, we show that a longitudinal analysis of SHAP values can provide effective information on Alzheimer's disease progression.

Accurate Evaluation of Feature Contributions for Sentinel Lymph Node Status Classification in Breast Cancer

The current guidelines recommend the sentinel lymph node biopsy to evaluate the lymph node involvement for breast cancer patients with clinically negative lymph nodes on clinical or radiological examination. Machine learning (ML) models have significantly improved the prediction of lymph nodes status based on clinical features, thus avoiding expensive, time-consuming and invasive procedures. However, the classification of sentinel lymph node status represents a typical example of an unbalanced classification problem. In this work, we developed a ML framework to explore the effects of unbalanced populations on the performance and stability of feature ranking for sentinel lymph node status classification in breast cancer. Our results indicate state-of-the-art AUC (Area under the Receiver Operating Characteristic curve) values on a hold-out set (67%) while providing particularly stable features related to tumor size, histological subtype and estrogen receptor expression, which should therefore be considered as potential biomarkers.

Explanations of Machine Learning Models in Repeated Nested Cross-Validation: An Application in Age Prediction Using Brain Complexity Features

SHAP (Shapley additive explanations) is a framework for explainable AI that makes explanations locally and globally. In this work, we propose a general method to obtain representative SHAP values within a repeated nested cross-validation procedure and separately for the training and test sets of the different cross-validation rounds to assess the real generalization abilities of the explanations. We applied this method to predict individual age using brain complexity features extracted from MRI scans of 159 healthy subjects. In particular, we used four implementations of the fractal dimension (FD) of the cerebral cortex—a measurement of brain complexity. Representative SHAP values highlighted that the most recent implementation of the FD had the highest impact over the others and was among the top-ranking features for predicting age. SHAP rankings were not the same in the training and test sets, but the top-ranking features were consistent. In conclusion, we propose a method—and share all the source code—that allows a rigorous assessment of the SHAP explanations of a trained model in a repeated nested cross-validation setting.

A multivariate multi-step LSTM forecasting model for tuberculosis incidence with model explanation in Liaoning Province, China

BACKGROUND

Tuberculosis (TB) is the respiratory infectious disease with the highest incidence in China. We aim to design a series of forecasting models and find the factors that affect the incidence of TB, thereby improving the accuracy of the incidence prediction.

RESULTS

In this paper, we developed a new interpretable prediction system based on the multivariate multi-step Long Short-Term Memory (LSTM) model and SHapley Additive exPlanation (SHAP) method. Four accuracy measures are introduced into the system: Root Mean Square Error, Mean Absolute Error, Mean Absolute Percentage Error, and symmetric Mean Absolute Percentage Error. The Autoregressive Integrated Moving Average (ARIMA) model and seasonal ARIMA model are established. The multi-step ARIMA-LSTM model is proposed for the first time to examine the performance of each model in the short, medium, and long term, respectively. Compared with the ARIMA model, each error of the multivariate 2-step LSTM model is reduced by 12.92%, 15.94%, 15.97%, and 14.81% in the short term. The 3-step ARIMA-LSTM model achieved excellent performance, with each error decreased to 15.19%, 33.14%, 36.79%, and 29.76% in the medium and long term. We provide the local and global explanation of the multivariate single-step LSTM model in the field of incidence prediction, pioneering.

CONCLUSIONS

The multivariate 2-step LSTM model is suitable for short-term prediction and obtained a similar performance as previous studies. The 3-step ARIMA-LSTM model is appropriate for medium-to-long-term prediction and outperforms these models. The SHAP results indicate that the five most crucial features are maximum temperature, average relative humidity, local financial budget, monthly sunshine percentage, and sunshine hours.

Validity of Machine Learning in Predicting Giant Cell Arteritis Flare After Glucocorticoids Tapering

Background

Inferential statistical methods failed in identifying reliable biomarkers and risk factors for relapsing giant cell arteritis (GCA) after glucocorticoids (GCs) tapering. A ML approach allows to handle complex non-linear relationships between patient attributes that are hard to model with traditional statistical methods, merging them to output a forecast or a probability for a given outcome.

Objective

The objective of the study was to assess whether ML algorithms can predict GCA relapse after GCs tapering.

Methods

GCA patients who underwent GCs therapy and regular follow-up visits for at least 12 months, were retrospectively analyzed and used for implementing 3 ML algorithms, namely, Logistic Regression (LR), Decision Tree (DT), and Random Forest (RF). The outcome of interest was disease relapse within 3 months during GCs tapering. After a ML variable selection method, based on a XGBoost wrapper, an attribute core set was used to train and test each algorithm using 5-fold cross-validation. The performance of each algorithm in both phases was assessed in terms of accuracy and area under receiver operating characteristic curve (AUROC).

Results

The dataset consisted of 107 GCA patients (73 women, 68.2%) with mean age ( ± SD) 74.1 ( ± 8.5) years at presentation. GCA flare occurred in 40/107 patients (37.4%) within 3 months after GCs tapering. As a result of ML wrapper, the attribute core set with the least number of variables used for algorithm training included presence/absence of diabetes mellitus and concomitant polymyalgia rheumatica as well as erythrocyte sedimentation rate level at GCs baseline. RF showed the best performance, being significantly superior to other algorithms in accuracy (RF 71.4% vs LR 70.4% vs DT 62.9%). Consistently, RF precision (72.1%) was significantly greater than those of LR (62.6%) and DT (50.8%). Conversely, LR was superior to RF and DT in recall (RF 60% vs LR 62.5% vs DT 47.5%). Moreover, RF AUROC (0.76) was more significant compared to LR (0.73) and DT (0.65).

Conclusions

RF algorithm can predict GCA relapse after GCs tapering with sufficient accuracy. To date, this is one of the most accurate predictive modelings for such outcome. This ML method represents a reproducible tool, capable of supporting clinicians in GCA patient management.

Application of Deep Learning to Ischemic and Hemorrhagic Stroke Computed Tomography and Magnetic Resonance Imaging

Deep Learning (DL) algorithm holds great potential in the field of stroke imaging. It has been applied not only to the "downstream" side such as lesion detection, treatment decision making, and outcome prediction, but also to the "upstream" side for generation and enhancement of stroke imaging. This paper aims to comprehensively overview the common applications of DL to stroke imaging. In the future, more standardized imaging datasets and more extensive studies are needed to establish and validate the role of DL in stroke imaging.

Enhancing the impact of Artificial Intelligence in Medicine: A joint AIFM-INFN Italian initiative for a dedicated cloud-based computing infrastructure

Artificial Intelligence (AI) techniques have been implemented in the field of Medical Imaging for more than forty years. Medical Physicists, Clinicians and Computer Scientists have been collaborating since the beginning to realize software solutions to enhance the informative content of medical images, including AI-based support systems for image interpretation. Despite the recent massive progress in this field due to the current emphasis on Radiomics, Machine Learning and Deep Learning, there are still some barriers to overcome before these tools are fully integrated into the clinical workflows to finally enable a precision medicine approach to patients' care. Nowadays, as Medical Imaging has entered the Big Data era, innovative solutions to efficiently deal with huge amounts of data and to exploit large and distributed computing resources are urgently needed. In the framework of a collaboration agreement between the Italian Association of Medical Physicists (AIFM) and the National Institute for Nuclear Physics (INFN), we propose a model of an intensive computing infrastructure, especially suited for training AI models, equipped with secure storage systems, compliant with data protection regulation, which will accelerate the development and extensive validation of AI-based solutions in the Medical Imaging field of research. This solution can be developed and made operational by Physicists and Computer Scientists working on complementary fields of research in Physics, such as High Energy Physics and Medical Physics, who have all the necessary skills to tailor the AI-technology to the needs of the Medical Imaging community and to shorten the pathway towards the clinical applicability of AI-based decision support systems.