The field of artificial intelligence (AI) is rapidly influencing health and healthcare, but bias and inadequate subgroup performance persists. Previous work has clearly outlined the need for more rigorous attention to data representativeness and model performance to advance population health and reduce bias. However, there is an opportunity to leverage best practices of social epidemiology, behavior science, and health services research to help develop hypotheses for associations found, particularly when using explainable AI (XAI) methods. This paper introduces a novel human-in-the-loop model that couples generative XAI with a wraparound interdisciplinary expert panel review process. This human-centered approach critically assesses AI model explanations from multiple perspectives, effectively identifying areas of bias and pinpointing directions for future research. We summarize the approach, highlighting processes and findings where the interdisciplinary expert panel produced interpretations which were historically and contextually informed. In particular, the interdisciplinary panel identified potential confounders and brought in scientific evidence from previous studies, which have clear implications to reduce bias. This promising human-centered framework not only maximizes multidisciplinary methodological rigor for AI model improvement and eventual implementation but also proactively fosters team science opportunities in future research.

Precipitation nowcasting, which predicts rainfall up to a few hours ahead, is a critical tool for vulnerable communities in the Global South frequently exposed to intense, rapidly developing storms. Timely forecasts provide a crucial window to protect lives and livelihoods. Traditional numerical weather prediction (NWP) methods suffer from high latency, low spatial and temporal resolution, and significant gaps in accuracy across the world. Recent machine learning-based nowcasting methods, common in the Global North, cannot be extended to the Global South due to extremely sparse radar coverage. We present Global MetNet, an operational global machine learning nowcasting model. It leverages the Global Precipitation Mission's CORRA dataset, geostationary satellite data, and global NWP data to predict precipitation for the next 12 hours. The model operates at a high resolution of approximately 0.05° ( 5km) spatially and 15 minutes temporally. Global MetNet significantly outperforms industry-standard hourly forecasts and achieves significantly higher skill, making forecasts useful over a much larger area of the world than previously available. Our model demonstrates better skill in data-sparse regions than even the best high-resolution NWP models achieve in the US. Validated using ground radar and satellite data, it shows significant improvements across key metrics like the critical success index and fractions skill score for all precipitation rates and lead times. Crucially, our model generates forecasts in under a minute, making it readily deployable for real-time applications. It is already deployed for millions of users on Google Search. This work represents a key step in reducing global disparities in forecast quality and integrating sparse, high-resolution satellite observations into weather forecasting.

BackgroundAI models have shown promise in performing many medical imaging tasks. However, our ability to explain what signals these models have learned is severely lacking. Explanations are needed in order to increase the trust of doctors in AI-based models, especially in domains where AI prediction capabilities surpass those of humans. Moreover, such explanations could enable novel scientific discovery by uncovering signals in the data that aren’t yet known to experts.MethodsIn this paper, we present a workflow for generating hypotheses to understand which visual signals in images are correlated with a classification model’s predictions for a given task. This approach leverages an automatic visual explanation algorithm followed by interdisciplinary expert review. We propose the following 4 steps: (i) Train a classifier to perform a given task to assess whether the imagery indeed contains signals relevant to the task; (ii) Train a StyleGAN-based image generator with an architecture that enables guidance by the classifier (“StylEx”); (iii) Automatically detect, extract, and visualize the top visual attributes that the classifier is sensitive towards. For visualization, we independently modify each of these attributes to generate counterfactual visualizations for a set of images (i.e., what the image would look like with the attribute increased or decreased); (iv) Formulate hypotheses for the underlying mechanisms, to stimulate future research. Specifically, present the discovered attributes and corresponding counterfactual visualizations to an interdisciplinary panel of experts so that hypotheses can account for social and structural determinants of health (e.g., whether the attributes correspond to known patho-physiological or socio-cultural phenomena, or could be novel discoveries).FindingsTo demonstrate the broad applicability of our approach, we present results on eight prediction tasks across three medical imaging modalities—retinal fundus photographs, external eye photographs, and chest radiographs. We showcase examples where many of the automatically-learned attributes clearly capture clinically known features (e.g., types of cataract, enlarged heart), and demonstrate automatically-learned confounders that arise from factors beyond physiological mechanisms (e.g., chest X-ray underexposure is correlated with the classifier predicting abnormality, and eye makeup is correlated with the classifier predicting low hemoglobin levels). We further show that our method reveals a number of physiologically plausible, previously-unknown attributes based on the literature (e.g., differences in the fundus associated with self-reported sex, which were previously unknown).InterpretationOur approach enables hypotheses generation via attribute visualizations and has the potential to enable researchers to better understand, improve their assessment, and extract new knowledge from AI-based models, as well as debug and design better datasets. Though not designed to infer causality, importantly, we highlight that attributes generated by our framework can capture phenomena beyond physiology or pathophysiology, reflecting the real world nature of healthcare delivery and socio-cultural factors, and hence interdisciplinary perspectives are critical in these investigations. Finally, we will release code to help researchers train their own StylEx models and analyze their predictive tasks of interest, and use the methodology presented in this paper for responsible interpretation of the revealed attributes.FundingGoogle.

Microscopic interpretation of histopathology images underlies many important diagnostic and treatment decisions. While advances in vision-language modeling raise new opportunities for analysis of such images, the gigapixel-scale size of whole slide images (WSIs) introduces unique challenges. Additionally, pathology reports simultaneously highlight key findings from small regions while also aggregating interpretation across multiple slides, often making it difficult to create robust image-text pairs. As such, pathology reports remain a largely untapped source of supervision in computational pathology, with most efforts relying on region-of-interest annotations or self-supervision at the patch-level. In this work, we develop a vision-language model based on the BLIP-2 framework using WSIs paired with curated text from pathology reports. This enables applications utilizing a shared image-text embedding space, such as text or image retrieval for finding cases of interest, as well as integration of the WSI encoder with a frozen large language model (LLM) for WSI-based generative text capabilities such as report generation or AI-in-the-loop interactions. We utilize a de-identified dataset of over 350,000 WSIs and diagnostic text pairs, spanning a wide range of diagnoses, procedure types, and tissue types. We present pathologist evaluation of text generation and text retrieval using WSI embeddings, as well as results for WSI classification and workflow prioritization (slide-level triaging). Model-generated text for WSIs was rated by pathologists as accurate, without clinically significant error or omission, for 78% of WSIs on average. This work demonstrates exciting potential capabilities for language-aligned WSI embeddings.

The field of artificial intelligence (AI) is rapidly influencing health and healthcare, but bias and poor performance persists for populations who face widespread structural oppression. Previous work has clearly outlined the need for more rigorous attention to data representativeness and model performance to advance equity and reduce bias. However, there is an opportunity to also improve the explainability of AI by leveraging best practices of social epidemiology and health equity to help us develop hypotheses for associations found. In this paper, we focus on explainable AI (XAI) and describe a framework for interdisciplinary expert panel review to discuss and critically assess AI model explanations from multiple perspectives and identify areas of bias and directions for future research. We emphasize the importance of the interdisciplinary expert panel to produce more accurate, equitable interpretations which are historically and contextually informed. Interdisciplinary panel discussions can help reduce bias, identify potential confounders, and identify opportunities for additional research where there are gaps in the literature. In turn, these insights can suggest opportunities for AI model improvement.

A representation-learning strategy for machine-learning models applied to medical-imaging tasks improves model robustness and training efficiency and mitigates suboptimal out-of-distribution performance. Machine-learning models for medical tasks can match or surpass the performance of clinical experts. However, in settings differing from those of the training dataset, the performance of a model can deteriorate substantially. Here we report a representation-learning strategy for machine-learning models applied to medical-imaging tasks that mitigates such 'out of distribution' performance problem and that improves model robustness and training efficiency. The strategy, which we named REMEDIS (for 'Robust and Efficient Medical Imaging with Self-supervision'), combines large-scale supervised transfer learning on natural images and intermediate contrastive self-supervised learning on medical images and requires minimal task-specific customization. We show the utility of REMEDIS in a range of diagnostic-imaging tasks covering six imaging domains and 15 test datasets, and by simulating three realistic out-of-distribution scenarios. REMEDIS improved in-distribution diagnostic accuracies up to 11.5% with respect to strong supervised baseline models, and in out-of-distribution settings required only 1-33% of the data for retraining to match the performance of supervised models retrained using all available data. REMEDIS may accelerate the development lifecycle of machine-learning models for medical imaging.

BACKGROUND:Photographs of the external eye were recently shown to reveal signs of diabetic retinal disease and elevated glycated haemoglobin. This study aimed to test the hypothesis that external eye photographs contain information about additional systemic medical conditions. METHODS:We developed a deep learning system (DLS) that takes external eye photographs as input and predicts systemic parameters, such as those related to the liver (albumin, aspartate aminotransferase [AST]); kidney (estimated glomerular filtration rate [eGFR], urine albumin-to-creatinine ratio [ACR]); bone or mineral (calcium); thyroid (thyroid stimulating hormone); and blood (haemoglobin, white blood cells [WBC], platelets). This DLS was trained using 123 130 images from 38 398 patients with diabetes undergoing diabetic eye screening in 11 sites across Los Angeles county, CA, USA. Evaluation focused on nine prespecified systemic parameters and leveraged three validation sets (A, B, C) spanning 25 510 patients with and without diabetes undergoing eye screening in three independent sites in Los Angeles county, CA, and the greater Atlanta area, GA, USA. We compared performance against baseline models incorporating available clinicodemographic variables (eg, age, sex, race and ethnicity, years with diabetes). FINDINGS:Relative to the baseline, the DLS achieved statistically significant superior performance at detecting AST >36·0 U/L, calcium <8·6 mg/dL, eGFR <60·0 mL/min/1·73 m2, haemoglobin <11·0 g/dL, platelets <150·0 × 103/μL, ACR ≥300 mg/g, and WBC <4·0 × 103/μL on validation set A (a population resembling the development datasets), with the area under the receiver operating characteristic curve (AUC) of the DLS exceeding that of the baseline by 5·3-19·9% (absolute differences in AUC). On validation sets B and C, with substantial patient population differences compared with the development datasets, the DLS outperformed the baseline for ACR ≥300·0 mg/g and haemoglobin <11·0 g/dL by 7·3-13·2%. INTERPRETATION:We found further evidence that external eye photographs contain biomarkers spanning multiple organ systems. Such biomarkers could enable accessible and non-invasive screening of disease. Further work is needed to understand the translational implications. FUNDING:Google.

Biological age, distinct from an individual's chronological age, has been studied extensively through predictive aging clocks. However, these clocks have limited accuracy in short time-scales. Here we trained deep learning models on fundus images from the EyePACS dataset to predict individuals' chronological age. Our retinal aging clocking, 'eyeAge', predicted chronological age more accurately than other aging clocks (mean absolute error of 2.86 and 3.30 years on quality-filtered data from EyePACS and UK Biobank, respectively). Additionally, eyeAge was independent of blood marker-based measures of biological age, maintaining an all-cause mortality hazard ratio of 1.026 even when adjusted for phenotypic age. The individual-specific nature of eyeAge was reinforced via multiple GWAS hits in the UK Biobank cohort. The top GWAS locus was further validated via knockdown of the fly homolog, Alk, which slowed age-related decline in vision in flies. This study demonstrates the potential utility of a retinal aging clock for studying aging and age-related diseases and quantitatively measuring aging on very short time-scales, opening avenues for quick and actionable evaluation of gero-protective therapeutics.

Retinal fundus photographs can be used to detect a range of retinal conditions. Here we show that deep-learning models trained instead on external photographs of the eyes can be used to detect diabetic retinopathy (DR), diabetic macular oedema and poor blood glucose control. We developed the models using eye photographs from 145,832 patients with diabetes from 301 DR screening sites and evaluated the models on four tasks and four validation datasets with a total of 48,644 patients from 198 additional screening sites. For all four tasks, the predictive performance of the deep-learning models was significantly higher than the performance of logistic regression models using self-reported demographic and medical history data, and the predictions generalized to patients with dilated pupils, to patients from a different DR screening programme and to a general eye care programme that included diabetics and non-diabetics. We also explored the use of the deep-learning models for the detection of elevated lipid levels. The utility of external eye photographs for the diagnosis and management of diseases should be further validated with images from different cameras and patient populations.

External eye photos were recently shown to reveal signs of diabetic retinal disease and elevated HbA1c. In this paper, we evaluate if external eye photos contain information about additional systemic medical conditions. We developed a deep learning system (DLS) that takes external eye photos as input and predicts multiple systemic parameters, such as those related to the liver (albumin, AST); kidney (eGFR estimated using the race-free 2021 CKD-EPI creatinine equation, the urine ACR); bone & mineral (calcium); thyroid (TSH); and blood count (Hgb, WBC, platelets). Development leveraged 151,237 images from 49,015 patients with diabetes undergoing diabetic eye screening in 11 sites across Los Angeles county, CA. Evaluation focused on 9 pre-specified systemic parameters and leveraged 3 validation sets (A, B, C) spanning 28,869 patients with and without diabetes undergoing eye screening in 3 independent sites in Los Angeles County, CA, and the greater Atlanta area, GA. We compared against baseline models incorporating available clinicodemographic variables (e.g. age, sex, race/ethnicity, years with diabetes). Relative to the baseline, the DLS achieved statistically significant superior performance at detecting AST>36, calcium<8.6, eGFR<60, Hgb<11, platelets<150, ACR>=300, and WBC<4 on validation set A (a patient population similar to the development sets), where the AUC of DLS exceeded that of the baseline by 5.2-19.4%. On validation sets B and C, with substantial patient population differences compared to the development sets, the DLS outperformed the baseline for ACR>=300 and Hgb<11 by 7.3-13.2%. Our findings provide further evidence that external eye photos contain important biomarkers of systemic health spanning multiple organ systems. Further work is needed to investigate whether and how these biomarkers can be translated into clinical impact.

Recent progress in Medical Artificial Intelligence (AI) has delivered systems that can reach clinical expert level performance. However, such systems tend to demonstrate sub-optimal "out-of-distribution" performance when evaluated in clinical settings different from the training environment. A common mitigation strategy is to develop separate systems for each clinical setting using site-specific data [1]. However, this quickly becomes impractical as medical data is time-consuming to acquire and expensive to annotate [2]. Thus, the problem of "data-efficient generalization" presents an ongoing difficulty for Medical AI development. Although progress in representation learning shows promise, their benefits have not been rigorously studied, specifically for out-of-distribution settings. To meet these challenges, we present REMEDIS, a unified representation learning strategy to improve robustness and data-efficiency of medical imaging AI. REMEDIS uses a generic combination of large-scale supervised transfer learning with self-supervised learning and requires little task-specific customization. We study a diverse range of medical imaging tasks and simulate three realistic application scenarios using retrospective data. REMEDIS exhibits significantly improved in-distribution performance with up to 11.5% relative improvement in diagnostic accuracy over a strong supervised baseline. More importantly, our strategy leads to strong data-efficient generalization of medical imaging AI, matching strong supervised baselines using between 1% to 33% of retraining data across tasks. These results suggest that REMEDIS can significantly accelerate the life-cycle of medical imaging AI development thereby presenting an important step forward for medical imaging AI to deliver broad impact.

BACKGROUND:Diabetic retinopathy screening is instrumental to preventing blindness, but scaling up screening is challenging because of the increasing number of patients with all forms of diabetes. We aimed to create a deep-learning system to predict the risk of patients with diabetes developing diabetic retinopathy within 2 years.METHODS:We created and validated two versions of a deep-learning system to predict the development of diabetic retinopathy in patients with diabetes who had had teleretinal diabetic retinopathy screening in a primary care setting. The input for the two versions was either a set of three-field or one-field colour fundus photographs. Of the 575 431 eyes in the development set 28 899 had known outcomes, with the remaining 546 532 eyes used to augment the training process via multitask learning. Validation was done on one eye (selected at random) per patient from two datasets: an internal validation (from EyePACS, a teleretinal screening service in the USA) set of 3678 eyes with known outcomes and an external validation (from Thailand) set of 2345 eyes with known outcomes.FINDINGS:The three-field deep-learning system had an area under the receiver operating characteristic curve (AUC) of 0·79 (95% CI 0·77-0·81) in the internal validation set. Assessment of the external validation set-which contained only one-field colour fundus photographs-with the one-field deep-learning system gave an AUC of 0·70 (0·67-0·74). In the internal validation set, the AUC of available risk factors was 0·72 (0·68-0·76), which improved to 0·81 (0·77-0·84) after combining the deep-learning system with these risk factors (p<0·0001). In the external validation set, the corresponding AUC improved from 0·62 (0·58-0·66) to 0·71 (0·68-0·75; p<0·0001) following the addition of the deep-learning system to available risk factors.INTERPRETATION:The deep-learning systems predicted diabetic retinopathy development using colour fundus photographs, and the systems were independent of and more informative than available risk factors. Such a risk stratification tool might help to optimise screening intervals to reduce costs while improving vision-related outcomes.FUNDING:Google.

Background: Patients with neovascular age-related macular degeneration (AMD) can avoid vision loss via certain therapy. However, methods to predict the progression to neovascular age-related macular degeneration (nvAMD) are lacking. Purpose: To develop and validate a deep learning (DL) algorithm to predict 1-year progression of eyes with no, early, or intermediate AMD to nvAMD, using color fundus photographs (CFP). Design: Development and validation of a DL algorithm. Methods: We trained a DL algorithm to predict 1-year progression to nvAMD, and used 10-fold cross-validation to evaluate this approach on two groups of eyes in the Age-Related Eye Disease Study (AREDS): none/early/intermediate AMD, and intermediate AMD (iAMD) only. We compared the DL algorithm to the manually graded 4-category and 9-step scales in the AREDS dataset. Main outcome measures: Performance of the DL algorithm was evaluated using the sensitivity at 80% specificity for progression to nvAMD. Results: The DL algorithm's sensitivity for predicting progression to nvAMD from none/early/iAMD (78+/-6%) was higher than manual grades from the 9-step scale (67+/-8%) or the 4-category scale (48+/-3%). For predicting progression specifically from iAMD, the DL algorithm's sensitivity (57+/-6%) was also higher compared to the 9-step grades (36+/-8%) and the 4-category grades (20+/-0%). Conclusions: Our DL algorithm performed better in predicting progression to nvAMD than manual grades. Future investigations are required to test the application of this DL algorithm in a real-world clinical setting.

Mapping the spatial distribution of poverty in developing countries remains an important and costly challenge. These "poverty maps" are key inputs for poverty targeting, public goods provision, political accountability, and impact evaluation, that are all the more important given the geographic dispersion of the remaining bottom billion severely poor individuals. In this paper we train Convolutional Neural Networks (CNNs) to estimate poverty directly from high and medium resolution satellite images. We use both Planet and Digital Globe imagery with spatial resolutions of 3-5 sq. m. and 50 sq. cm. respectively, covering all 2 million sq. km. of Mexico. Benchmark poverty estimates come from the 2014 MCS-ENIGH combined with the 2015 Intercensus and are used to estimate poverty rates for 2,456 Mexican municipalities. CNNs are trained using the 896 municipalities in the 2014 MCS-ENIGH. We experiment with several architectures (GoogleNet, VGG) and use GoogleNet as a final architecture where weights are fine-tuned from ImageNet. We find that 1) the best models, which incorporate satellite-estimated land use as a predictor, explain approximately 57% of the variation in poverty in a validation sample of 10 percent of MCS-ENIGH municipalities; 2) Across all MCS-ENIGH municipalities explanatory power reduces to 44% in a CNN prediction and landcover model; 3) Predicted poverty from the CNN predictions alone explains 47% of the variation in poverty in the validation sample, and 37% over all MCS-ENIGH municipalities; 4) In urban areas we see slight improvements from using Digital Globe versus Planet imagery, which explain 61% and 54% of poverty variation respectively. We conclude that CNNs can be trained end-to-end on satellite imagery to estimate poverty, although there is much work to be done to understand how the training process influences out of sample validation.

The text localization and recognition in real time scene text images is still a big issue in current application. Mobile application and digitization in real world gives a vital and broad impact on real time scene text images. However, the efficiency of recognition rate depends upon the text localization, i.e., higher the purity of text background segmentation and decomposition, higher the rate of accuracy for the image recognition. In this paper, we present a new scene text detection algorithm based on Stroke detection and Hog Transform method. The method introduces an approach for character detection and recognition which combines the advantages of Hog Transform and Connected Component methods. Characters are detected and recognized on the basis of image regions which contain strokes of specific orientations in a specific relative position, where the strokes are efficiently detected by convolving the image gradient field with a set of oriented bar filters. The method was evaluated on a standard dataset consisting mostly real time images where it achieves state-of-the-art results in both text localization and recognition. The results clearly depict the higher bit of accuracy in terms of localization and recognition for a collected dataset.

Statistical machine learning techniques have transformed computer vision research in the last two decades, and have led to many breakthroughs in object detection, recognition and tracking. Such data-driven methods extrapolate rules from a set of labeled examples, freeing us from designing and tuning a system by hand for a particular application or domain. Discriminative learning methods, which directly learn to differentiate categories of data rather than modeling the data itself, have been shown to be particularly effective. However, the requirement of a large set of labeled examples becomes prohibitively expensive, especially if we consider scaling to a wide range of domains and applications. In this dissertation we explore weakly supervised methods of training discriminative models for a number of computer vision applications. These methods require weaker forms of annotation that are easier and/or cheaper to obtain, and can learn in situations where the ground truth is inherently ambiguous. Many of the algorithms in this dissertation are based on a particular form of weakly supervised learning called Multiple Instance Learning (MIL). Our final contribution is a theoretical analysis of MIL that takes into account the characteristics of applications in computer vision and related areas.