A scoping review of fair machine learning techniques when using real-world data.

OBJECTIVE: The integration of artificial intelligence (AI) and machine learning (ML) in health care to aid clinical decisions is widespread. However, as AI and ML take important roles in health care, there are concerns about AI and ML associated fairness and bias. That is, an AI tool may have a disparate impact, with its benefits and drawbacks unevenly distributed across societal strata and subpopulations, potentially exacerbating existing health inequities. Thus, the objectives of this scoping review were to summarize existing literature and identify gaps in the topic of tackling algorithmic bias and optimizing fairness in AI/ML models using real-world data (RWD) in health care domains. METHODS: We conducted a thorough review of techniques for assessing and optimizing AI/ML model fairness in health care when using RWD in health care domains. The focus lies on appraising different quantification metrics for accessing fairness, publicly accessible datasets for ML fairness research, and bias mitigation approaches. RESULTS: We identified 11 papers that are focused on optimizing model fairness in health care applications. The current research on mitigating bias issues in RWD is limited, both in terms of disease variety and health care applications, as well as the accessibility of public datasets for ML fairness research. Existing studies often indicate positive outcomes when using pre-processing techniques to address algorithmic bias. There remain unresolved questions within the field that require further research, which includes pinpointing the root causes of bias in ML models, broadening fairness research in AI/ML with the use of RWD and exploring its implications in healthcare settings, and evaluating and addressing bias in multi-modal data. CONCLUSION: This paper provides useful reference material and insights to researchers regarding AI/ML fairness in real-world health care data and reveals the gaps in the field. Fair AI/ML in health care is a burgeoning field that requires a heightened research focus to cover diverse applications and different types of RWD.

Identify and mitigate bias in electronic phenotyping: A comprehensive study from computational perspective.

Electronic phenotyping is a fundamental task that identifies the special group of patients, which plays an important role in precision medicine in the era of digital health. Phenotyping provides real-world evidence for other related biomedical research and clinical tasks, e.g., disease diagnosis, drug development, and clinical trials, etc. With the development of electronic health records, the performance of electronic phenotyping has been significantly boosted by advanced machine learning techniques. In the healthcare domain, precision and fairness are both essential aspects that should be taken into consideration. However, most related efforts are put into designing phenotyping models with higher accuracy. Few attention is put on the fairness perspective of phenotyping. The neglection of bias in phenotyping leads to subgroups of patients being underrepresented which will further affect the following healthcare activities such as patient recruitment in clinical trials. In this work, we are motivated to bridge this gap through a comprehensive experimental study to identify the bias existing in electronic phenotyping models and evaluate the widely-used debiasing methods' performance on these models. We choose pneumonia and sepsis as our phenotyping target diseases. We benchmark 9 kinds of electronic phenotyping methods spanning from rule-based to data-driven methods. Meanwhile, we evaluate the performance of the 5 bias mitigation strategies covering pre-processing, in-processing, and post-processing. Through the extensive experiments, we summarize several insightful findings from the bias identified in the phenotyping and key points of the bias mitigation strategies in phenotyping.

Search-based Automatic Repair for Fairness and Accuracy in Decision-making Software.

Decision-making software mainly based on Machine Learning (ML) may contain fairness issues (e.g., providing favourable treatment to certain people rather than others based on sensitive attributes such as gender or race). Various mitigation methods have been proposed to automatically repair fairness issues to achieve fairer ML software and help software engineers to create responsible software. However, existing bias mitigation methods trade accuracy for fairness (i.e., trade a reduction in accuracy for better fairness). In this paper, we present a novel search-based method for repairing ML-based decision making software to simultaneously increase both its fairness and accuracy. As far as we know, this is the first bias mitigation approach based on multi-objective search that aims to repair fairness issues without trading accuracy for binary classification methods. We apply our approach to two widely studied ML models in the software fairness literature (i.e., Logistic Regression and Decision Trees), and compare it with seven publicly available state-of-the-art bias mitigation methods by using three different fairness measurements. The results show that our approach successfully increases both accuracy and fairness for 61% of the cases studied, while the state-of-the-art always decrease accuracy when attempting to reduce bias. With our proposed approach, software engineers that previously were concerned with accuracy losses when considering fairness, are now enabled to improve the fairness of binary classification models without sacrificing accuracy.

The time is now: addressing implicit bias in obstetrics and gynecology education.

Obstetrician-gynecologists can improve the learning environment and patient care by addressing implicit bias. Accumulating evidence demonstrates that racial and gender-based discrimination is woven into medical education, formal curricula, patient-provider-trainee interactions in the clinical workspace, and all aspects of learner assessment. Implicit bias negatively affects learners in every space. Strategies to address implicit bias at the individual, interpersonal, institutional, and structural level to improve the well-being of learners and patients are needed. The authors review an approach to addressing implicit bias in obstetrics and gynecology education, which includes: (1) curricular design using an educational framework of antiracism and social justice theories, (2) bias awareness and management pedagogy throughout the curriculum, (3) elimination of stereotypical patient descriptions from syllabi and examination questions, and (4) critical review of epidemiology and evidence-based medicine for underlying assumptions based on discriminatory practices or structural racism that unintentionally reinforce stereotypes and bias. The movement toward competency-based medical education and holistic evaluations may result in decreased bias in learner assessment. Educators may wish to monitor grades and narratives for bias as a form of continuous educational equity improvement. Given that practicing physicians may have little training in this area, faculty development efforts in bias awareness and mitigation strategies may have significant impact on learner well-being.

Evaluating and mitigating bias in machine learning models for cardiovascular disease prediction.

OBJECTIVE: The study aims to investigate whether machine learning-based predictive models for cardiovascular disease (CVD) risk assessment show equivalent performance across demographic groups (such as race and gender) and if bias mitigation methods can reduce any bias present in the models. This is important as systematic bias may be introduced when collecting and preprocessing health data, which could affect the performance of the models on certain demographic sub-cohorts. The study is to investigate this using electronic health records data and various machine learning models. METHODS: The study used large de-identified Electronic Health Records data from Vanderbilt University Medical Center. Machine learning (ML) algorithms including logistic regression, random forest, gradient-boosting trees, and long short-term memory were applied to build multiple predictive models. Model bias and fairness were evaluated using equal opportunity difference (EOD, 0 indicates fairness) and disparate impact (DI, 1 indicates fairness). In our study, we also evaluated the fairness of a non-ML baseline model, the American Heart Association (AHA) Pooled Cohort Risk Equations (PCEs). Moreover, we compared the performance of three different de-biasing methods: removing protected attributes (e.g., race and gender), resampling the imbalanced training dataset by sample size, and resampling by the proportion of people with CVD outcomes. RESULTS: The study cohort included 109,490 individuals (mean [SD] age 47.4 [14.7] years; 64.5% female; 86.3% White; 13.7% Black). The experimental results suggested that most ML models had smaller EOD and DI than PCEs. For ML models, the mean EOD ranged from -0.001 to 0.018 and the mean DI ranged from 1.037 to 1.094 across race groups. There was a larger EOD and DI across gender groups, with EOD ranging from 0.131 to 0.136 and DI ranging from 1.535 to 1.587. For debiasing methods, removing protected attributes didn't significantly reduced the bias for most ML models. Resampling by sample size also didn't consistently decrease bias. Resampling by case proportion reduced the EOD and DI for gender groups but slightly reduced accuracy in many cases. CONCLUSIONS: Among the VUMC cohort, both PCEs and ML models were biased against women, suggesting the need to investigate and correct gender disparities in CVD risk prediction. Resampling by proportion reduced the bias for gender groups but not for race groups.

Confidence, Connection & Collaboration: Creating a Scalable Bias Reduction Improvement Coaching Train-the-Trainer Program to Mitigate Implicit Bias across a Medical Center.

Problem: Academic medical centers need to mitigate the negative effects of implicit bias with approaches that are empirically-based, scalable, sustainable, and specific to departmental needs. Guided by Kotter's Model of Change to create and sustain cultural change, we developed the Bias Reduction Improvement Coaching Program (BRIC), a two-year, train-the-trainer implicit bias coaching program designed to meet the increasing demand for bias training across a university medical center. Intervention: BRIC trained a cohort of faculty and staff as coaches during four quarterly training sessions in Year 1 that covered 1) the science of bias, 2) bias in selection and hiring, 3) bias in mentoring, and 4) bias in promotion, retention, and workplace culture. In Year 2, coaches attended two booster sessions and delivered at least two presentations. BRIC raises awareness of bias mitigation strategies in a scalable way by uniquely building capacity through department-level champions, providing programming that addresses the 'local context,' and setting a foundation for sustained institutional change. Context: In a U.S. academic medical center, 27 faculty and staff from 24 departments were trained as inaugural BRIC coaches. We assessed outcomes at multiple levels: BRIC coach outcomes (feedback on the training sessions; coach knowledge, attitudes, and skills), departmental-level outcomes (program attendee feedback, knowledge, and intentions) and institutional outcomes (activities to sustain change). Impact: After Year 1, coaches reported high satisfaction with BRIC and a statistically significant increase in self-efficacy in their abilities to recognize, mitigate, and teach about implicit bias. In Year 2, attendees at BRIC coach presentations reported an increase in bias mitigation knowledge, and the majority committed to taking follow-up action (e.g., taking an Implicit Association Test). Coaches also launched activities for sustaining change at the broader university and beyond. Lessons Learned: The BRIC Program indicates a high level of interest in receiving bias mitigation training, both among individuals who applied to be BRIC coaches and among presentation attendees. BRIC's initial success supports future expansion. The model appears scalable and sustainable; future efforts will formalize the emerging community of practice around bias mitigation and measure elements of on-going institutional culture change.

Implicit Racial Bias Attitudes and Mitigation in Neonatal Nurse Practitioners: A Quality Improvement Project.

PURPOSE: to assess the impact of education using the implicit bias recognition and management (IBRM) teaching approach. DESIGN: longitudinal quasi-experimental design. Surveys at baseline, immediate postimplementation, and 4-week postimplementation using the modified version of the Attitudes Toward Implicit Bias Instrument (ATIBI). The 4-week survey included items about implicit bias recognition and mitigation strategies. SAMPLE: thirty-six neonatal nurse practitioners assigned to the NICU in a Midwest urban children's hospital. RESULTS: one-way repeated-measures analysis of variance was used, and the score range was 16-96. The results showed a statistically significant model, F (1.49, 707.97) = 34.46, p <.001, partial η2 = 0.496. Pairwise comparisons showed improvement from pre (M = 73.08, SD = 9.36) to immediate postimplementation (M = 80.06, SD = 8.19), p <.001. Scores were sustained at 4-week postimplementation (M = 79.28, SD = 10.39), p = .744. CONCLUSIONS: The IBRM teaching approach improved scores from baseline on a modified ATIBI that remained improved 4 weeks after the education.

Addressing the Need for Controls on Particle Bounce and Re-entrainment in the Cascade Impactor and for the Mitigation of Electrostatic Charge for Aerodynamic Particle Size Assessment of Orally Inhaled Products: an Assessment by the International Consortium on Regulation and Science (IPAC-RS).

Multi-stage cascade impactors (CI) are accepted for the determination of metrics of the drug mass aerodynamic particle size distributions (APSD) of aerosols emitted from orally inhaled products (OIPs). This is particularly important for products where the drug to excipient ratio or particle density may not be the same in each aerodynamic size fraction; examples of such products are carrier-containing dry powder inhalers (DPIs) and suspension pressurized metered-dose inhalers (pMDIs). CI measurements have been used as the "gold standard" for acceptance of alternative methods of APSD assessment, such as laser diffraction for nebulized solutions. Although these apparatus are labor-intensive, they are accepted in regulatory submissions and quality control assessments because the mass of active pharmaceutical ingredient(s) in the aerosol can be quantified by chemical assay and measured particle size is based on the aerodynamic diameter scale that is predictive of deposition in the respiratory tract. Two of the most important factors that modify the ideal operation of an impactor are "particle bounce," that is often accompanied by re-entrainment in the air flow passing the stage of interest, and electrostatic charge acquired by the particles during the preparation and aerosolization of the formulation when the inhaler is actuated. This article reviews how both factors can lead to biased APSD measurements, focusing on measurements involving pMDIs and DPIs, where these sources of error are most likely to be encountered. Recommendations are provided for the mitigation of both factors to assist the practitioner of these measurements.

Towards objective and systematic evaluation of bias in artificial intelligence for medical imaging.

OBJECTIVE: Artificial intelligence (AI) models trained using medical images for clinical tasks often exhibit bias in the form of subgroup performance disparities. However, since not all sources of bias in real-world medical imaging data are easily identifiable, it is challenging to comprehensively assess their impacts. In this article, we introduce an analysis framework for systematically and objectively investigating the impact of biases in medical images on AI models. MATERIALS AND METHODS: Our framework utilizes synthetic neuroimages with known disease effects and sources of bias. We evaluated the impact of bias effects and the efficacy of 3 bias mitigation strategies in counterfactual data scenarios on a convolutional neural network (CNN) classifier. RESULTS: The analysis revealed that training a CNN model on the datasets containing bias effects resulted in expected subgroup performance disparities. Moreover, reweighing was the most successful bias mitigation strategy for this setup. Finally, we demonstrated that explainable AI methods can aid in investigating the manifestation of bias in the model using this framework. DISCUSSION: The value of this framework is showcased in our findings on the impact of bias scenarios and efficacy of bias mitigation in a deep learning model pipeline. This systematic analysis can be easily expanded to conduct further controlled in silico trials in other investigations of bias in medical imaging AI. CONCLUSION: Our novel methodology for objectively studying bias in medical imaging AI can help support the development of clinical decision-support tools that are robust and responsible.

Drop the shortcuts: image augmentation improves fairness and decreases AI detection of race and other demographics from medical images.

BACKGROUND: It has been shown that AI models can learn race on medical images, leading to algorithmic bias. Our aim in this study was to enhance the fairness of medical image models by eliminating bias related to race, age, and sex. We hypothesise models may be learning demographics via shortcut learning and combat this using image augmentation. METHODS: This study included 44,953 patients who identified as Asian, Black, or White (mean age, 60.68 years ±18.21; 23,499 women) for a total of 194,359 chest X-rays (CXRs) from MIMIC-CXR database. The included CheXpert images comprised 45,095 patients (mean age 63.10 years ±18.14; 20,437 women) for a total of 134,300 CXRs were used for external validation. We also collected 1195 3D brain magnetic resonance imaging (MRI) data from the ADNI database, which included 273 participants with an average age of 76.97 years ±14.22, and 142 females. DL models were trained on either non-augmented or augmented images and assessed using disparity metrics. The features learned by the models were analysed using task transfer experiments and model visualisation techniques. FINDINGS: In the detection of radiological findings, training a model using augmented CXR images was shown to reduce disparities in error rate among racial groups (-5.45%), age groups (-13.94%), and sex (-22.22%). For AD detection, the model trained with augmented MRI images was shown 53.11% and 31.01% reduction of disparities in error rate among age and sex groups, respectively. Image augmentation led to a reduction in the model's ability to identify demographic attributes and resulted in the model trained for clinical purposes incorporating fewer demographic features. INTERPRETATION: The model trained using the augmented images was less likely to be influenced by demographic information in detecting image labels. These results demonstrate that the proposed augmentation scheme could enhance the fairness of interpretations by DL models when dealing with data from patients with different demographic backgrounds. FUNDING: National Science and Technology Council (Taiwan), National Institutes of Health.

Class-incremental learning with Balanced Embedding Discrimination Maximization.

Class incremental learning is committed to solving representation learning and classification assignments while avoiding catastrophic forgetting in scenarios where categories are increasing. In this work, a unified method named Balanced Embedding Discrimination Maximization (BEDM) is developed to make the intermediate embedding more distinctive. Specifically, we utilize an orthogonality constraint based on doubly-blocked Toeplitz matrix to minimize the correlation of convolution kernels, and an algorithm for similarity visualization is introduced. Furthermore, uneven samples and distribution shift among old and new tasks eventuate strongly biased classifiers. To mitigate the imbalance, we propose an adaptive balance weighting in softmax to compensate insufficient categories dynamically. In addition, hybrid embedding learning is introduced to preserve knowledge from old models, which involves less hyper-parameters than conventional knowledge distillation. Our proposed method outperforms the existing approaches on three mainstream benchmark datasets. Moreover, we technically visualize that our method can produce a more uniform similarity histogram and more stable spectrum. Grad-CAM and t-SNE visualizations further confirm its effectiveness. Code is available at https://github.com/wqzh/BEDM.

Mitigating animal methods bias to reduce animal use and improve biomedical translation.

Nonanimal biomedical research methods have advanced rapidly over the last decade making them the first-choice model for many researchers due to improved translatability and avoidance of ethical concerns. Yet confidence in novel nonanimal methods is still being established and they remain a small portion of nonclinical biomedical research, which can lead peer reviewers to evaluate animal-free studies or grant proposals in a biased manner. This "animal methods bias" is the preference for animal-based research methods where they are not necessary or where nonanimal-based methods are suitable. It affects the fair consideration of animal-free biomedical research, hampering the uptake and dissemination of these approaches by putting pressure on researchers to conduct animal experiments and potentially perpetuating the use of poorly translatable model systems. An international team of researchers and advocates called the Coalition to Illuminate and Address Animal Methods Bias (COLAAB) aims to provide concrete evidence of the existence and consequences of this bias and to develop and implement solutions towards overcoming it. The COLAAB recently developed the first of several mitigation tools: the Author Guide for Addressing Animal Methods Bias in Publishing, which is described herein along with broader implications and future directions of this work.

Balancing Performance and Interpretability in Medical Image Analysis: Case study of Osteopenia.

Multiple studies within the medical field have highlighted the remarkable effectiveness of using convolutional neural networks for predicting medical conditions, sometimes even surpassing that of medical professionals. Despite their great performance, convolutional neural networks operate as black boxes, potentially arriving at correct conclusions for incorrect reasons or areas of focus. Our work explores the possibility of mitigating this phenomenon by identifying and occluding confounding variables within images. Specifically, we focused on the prediction of osteopenia, a serious medical condition, using the publicly available GRAZPEDWRI-DX dataset. After detection of the confounding variables in the dataset, we generated masks that occlude regions of images associated with those variables. By doing so, models were forced to focus on different parts of the images for classification. Model evaluation using F1-score, precision, and recall showed that models trained on non-occluded images typically outperformed models trained on occluded images. However, a test where radiologists had to choose a model based on the focused regions extracted by the GRAD-CAM method showcased different outcomes. The radiologists' preference shifted towards models trained on the occluded images. These results suggest that while occluding confounding variables may degrade model performance, it enhances interpretability, providing more reliable insights into the reasoning behind predictions. The code to repeat our experiment is available on the following link: https://github.com/mikulicmateo/osteopenia .

Mitigating Sociodemographic Bias in Opioid Use Disorder Prediction: Fairness-Aware Machine Learning Framework.

BACKGROUND: Opioid use disorder (OUD) is a critical public health crisis in the United States, affecting >5.5 million Americans in 2021. Machine learning has been used to predict patient risk of incident OUD. However, little is known about the fairness and bias of these predictive models. OBJECTIVE: The aims of this study are two-fold: (1) to develop a machine learning bias mitigation algorithm for sociodemographic features and (2) to develop a fairness-aware weighted majority voting (WMV) classifier for OUD prediction. METHODS: We used the 2020 National Survey on Drug and Health data to develop a neural network (NN) model using stochastic gradient descent (SGD; NN-SGD) and an NN model using Adam (NN-Adam) optimizers and evaluated sociodemographic bias by comparing the area under the curve values. A bias mitigation algorithm, based on equality of odds, was implemented to minimize disparities in specificity and recall. Finally, a WMV classifier was developed for fairness-aware prediction of OUD. To further analyze bias detection and mitigation, we did a 1-N matching of OUD to non-OUD cases, controlling for socioeconomic variables, and evaluated the performance of the proposed bias mitigation algorithm and WMV classifier. RESULTS: Our bias mitigation algorithm substantially reduced bias with NN-SGD, by 21.66% for sex, 1.48% for race, and 21.04% for income, and with NN-Adam by 16.96% for sex, 8.87% for marital status, 8.45% for working condition, and 41.62% for race. The fairness-aware WMV classifier achieved a recall of 85.37% and 92.68% and an accuracy of 58.85% and 90.21% using NN-SGD and NN-Adam, respectively. The results after matching also indicated remarkable bias reduction with NN-SGD and NN-Adam, respectively, as follows: sex (0.14% vs 0.97%), marital status (12.95% vs 10.33%), working condition (14.79% vs 15.33%), race (60.13% vs 41.71%), and income (0.35% vs 2.21%). Moreover, the fairness-aware WMV classifier achieved high performance with a recall of 100% and 85.37% and an accuracy of 73.20% and 89.38% using NN-SGD and NN-Adam, respectively. CONCLUSIONS: The application of the proposed bias mitigation algorithm shows promise in reducing sociodemographic bias, with the WMV classifier confirming bias reduction and high performance in OUD prediction.

Bias reduction using combined stain normalization and augmentation for AI-based classification of histological images.

Artificial intelligence (AI)-assisted diagnosis is an ongoing revolution in pathology. However, a frequent drawback of AI models is their propension to make decisions based rather on bias in training dataset than on concrete biological features, thus weakening pathologists' trust in these tools. Technically, it is well known that microscopic images are altered by tissue processing and staining procedures, being one of the main sources of bias in machine learning for digital pathology. So as to deal with it, many teams have written about color normalization and augmentation methods. However, only a few of them have monitored their effects on bias reduction and model generalizability. In our study, two methods for stain augmentation (AugmentHE) and fast normalization (HEnorm) have been created and their effect on bias reduction has been monitored. Actually, they have also been compared to previously described strategies. To that end, a multicenter dataset created for breast cancer histological grading has been used. Thanks to it, classification models have been trained in a single center before assessing its performance in other centers images. This setting led to extensively monitor bias reduction while providing accurate insight of both augmentation and normalization methods. AugmentHE provided an 81% increase in color dispersion compared to geometric augmentations only. In addition, every classification model that involved AugmentHE presented a significant increase in the area under receiving operator characteristic curve (AUC) over the widely used RGB shift. More precisely, AugmentHE-based models showed at least 0.14 AUC increase over RGB shift-based models. Regarding normalization, HEnorm appeared to be up to 78x faster than conventional methods. It also provided satisfying results in terms of bias reduction. Altogether, our pipeline composed of AugmentHE and HEnorm improved AUC on biased data by up to 21.7% compared to usual augmentations. Conventional normalization methods coupled with AugmentHE yielded similar results while being much slower. In conclusion, we have validated an open-source tool that can be used in any deep learning-based digital pathology project on H&E whole slide images (WSI) that efficiently reduces stain-induced bias and later on might help increase pathologists' confidence when using AI-based products.

Evaluation and Assessment of the ABATE Framework to Enhance Implicit Bias Training for Virtual Interviews in Medical Schools.

Introduction: The influence of implicit biases in virtual interviews must be addressed to ensure equity within the admissions process. ABATE is a mnemonic framework of five specific categories of implicit bias (affinity-based, backdrop-based, appearance-based, technology and media-based, and enunciation-based biases) that should be anticipated and mitigated for faculty, staff, health professionals, and medical students who conduct virtual interviews at medical schools. Methods: A 60-minute workshop was developed to educate medical school admissions interviewers about the ABATE model and strategies to mitigate implicit bias during virtual interviews. Four workshops were held over 1 year totaling 217 individual attendees. The workshops were evaluated using a single-group, pre-post questionnaire designed with the Kirkpatrick evaluation model. Results: Attendees reported that they found the ABATE workshop useful and relevant to improving their ability to minimize implicit bias during virtual interviews. Significant improvements were found in attendee reactions to the utility of implicit bias training (M pre = 2.6, M post = 3.1, p = .002). Significant changes were also reported in attendees' attitudes about interviewing confidence (M pre = 3.0, M post = 3.2, p = .04), bias awareness (M pre = 3.0, M post = 3.4, p = .002), and identifying and applying bias mitigation solutions (M pre = 2.5, M post = 3.0, p = .003). Knowledge specific to backdrop-based biases also significantly increased (M pre = 3.2, M post = 3.4, p = .04). Discussion: The ABATE workshop demonstrates promise in mitigating implicit bias in virtual medical school interviews.

Fairness and bias correction in machine learning for depression prediction across four study populations.

A significant level of stigma and inequality exists in mental healthcare, especially in under-served populations. Inequalities are reflected in the data collected for scientific purposes. When not properly accounted for, machine learning (ML) models learned from data can reinforce these structural inequalities or biases. Here, we present a systematic study of bias in ML models designed to predict depression in four different case studies covering different countries and populations. We find that standard ML approaches regularly present biased behaviors. We also show that mitigation techniques, both standard and our own post-hoc method, can be effective in reducing the level of unfair bias. There is no one best ML model for depression prediction that provides equality of outcomes. This emphasizes the importance of analyzing fairness during model selection and transparent reporting about the impact of debiasing interventions. Finally, we also identify positive habits and open challenges that practitioners could follow to enhance fairness in their models.

Improving Equity in Deep Learning Medical Applications with the Gerchberg-Saxton Algorithm.

Deep learning (DL) has gained prominence in healthcare for its ability to facilitate early diagnosis, treatment identification with associated prognosis, and varying patient outcome predictions. However, because of highly variable medical practices and unsystematic data collection approaches, DL can unfortunately exacerbate biases and distort estimates. For example, the presence of sampling bias poses a significant challenge to the efficacy and generalizability of any statistical model. Even with DL approaches, selection bias can lead to inconsistent, suboptimal, or inaccurate model results, especially for underrepresented populations. Therefore, without addressing bias, wider implementation of DL approaches can potentially cause unintended harm. In this paper, we studied a novel method for bias reduction that leverages the frequency domain transformation via the Gerchberg-Saxton and corresponding impact on the outcome from a racio-ethnic bias perspective.

Using Simulation-Based Learning with Standardized Patients (SP) in an Implicit Bias Mitigation Clinician Training Program.

Objectives: To describe the development and refinement of an implicit bias recognition and management training program for clinical trainees. Methods: In the context of an NIH-funded clinical trial to address healthcare disparities in hypertension management, research and education faculty at an academic medical center used a participatory action research approach to engage local community members to develop and refine a "knowledge, awareness, and skill-building" bias recognition and mitigation program. The program targeted medical residents and Doctor of Nursing Practice students. The content of the two-session training included: didactics about healthcare disparities, racism and implicit bias; implicit association test (IAT) administration to raise awareness of personal implicit bias; skill building for bias-mitigating communication; and case scenarios for skill practice in simulation-based encounters with standardized patients (SPs) from the local community. Results: The initial trial year enrolled n = 65 interprofessional participants. Community partners and SPs who engaged throughout the design and implementation process reported overall positive experiences, but SPs expressed need for greater faculty support during in-person debriefings following simulation encounters to balance power dynamics. Initial year trainee participants reported discomfort with intensive sequencing of in-person didactics, IATs, and SP simulations in each of the two training sessions. In response, authors refined the training program to separate didactic sessions from IAT administration and SP simulations, and to increase safe space, and trainee and SP empowerment. The final program includes more interactive discussions focused on identity, race and ethnicity, and strategies to address local health system challenges related to structural racism. Conclusion: It is possible to develop and implement a bias awareness and mitigation skills training program that uses simulation-based learning with SPs, and to engage with local community members to tailor the content to address the experience of local patient populations. Further research is needed to measure the success and impact of replicating this approach elsewhere.

Bias Analysis in Healthcare Time Series (BAHT) Decision Support Systems from Meta Data.

One of the hindrances in the widespread acceptance of deep learning-based decision support systems in healthcare is bias. Bias in its many forms occurs in the datasets used to train and test deep learning models and is amplified when deployed in the real world, leading to challenges such as model drift. Recent advancements in the field of deep learning have led to the deployment of deployable automated healthcare diagnosis decision support systems at hospitals as well as tele-medicine through IoT devices. Research has been focused primarily on the development and improvement of these systems leaving a gap in the analysis of the fairness. The domain of FAccT ML (fairness, accountability, and transparency) accounts for the analysis of these deployable machine learning systems. In this work, we present a framework for bias analysis in healthcare time series (BAHT) signals such as electrocardiogram (ECG) and electroencephalogram (EEG). BAHT provides a graphical interpretive analysis of bias in the training, testing datasets in terms of protected variables, and analysis of bias amplification by the trained supervised learning model for time series healthcare decision support systems. We thoroughly investigate three prominent time series ECG and EEG healthcare datasets used for model training and research. We show the extensive presence of bias in the datasets leads to potentially biased or unfair machine-learning models. Our experiments also demonstrate the amplification of identified bias with an observed maximum of 66.66%. We investigate the effect of model drift due to unanalyzed bias in datasets and algorithms. Bias mitigation though prudent is a nascent area of research. We present experiments and analyze the most prevalently accepted bias mitigation strategies of under-sampling, oversampling, and the use of synthetic data for balancing the dataset through augmentation. It is important that healthcare models, datasets, and bias mitigation strategies should be properly analyzed for a fair unbiased delivery of service.