Real-world Imaging Data Research Articles

Cohort Builder: A Software Pipeline for Generating Patient Cohorts with Predetermined Baseline Characteristics from Medical Records and Raw Ophthalmic Imaging Data.

In clinical research, the analysis of patient cohorts is a widely employed method for investigating relevant healthcare questions. The ability to automatically extract large-scale patient cohorts from hospital systems is vital in order to unlock the potential of real-world clinical data, and answer pivotal medical questions through retrospective research studies. However, existing medical data is often dispersed across various systems and databases, preventing a systematic approach to access and interoperability. Even when the data are readily accessible, clinical researchers need to sift through Electronic Medical Records, confirm ethical approval, verify status of patient consent, check the availability of imaging data, and filter the data based on disease-specific image biomarkers. We present Cohort Builder, a software pipeline designed to facilitate the creation of patient cohorts with predefined baseline characteristics from real-world ophthalmic imaging data and electronic medical records. The applicability of our approach extends beyond ophthalmology to other medical domains with similar requirements such as neurology, cardiology and orthopedics.

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

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. 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. 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. 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. 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.

Generative Artificial Intelligence Enhancements for Reducing Image-based Training Data Requirements

ObjectiveTraining data fuel and shape the development of AI models. Intensive data requirements are a major bottleneck limiting the success of AI tools in sectors with inherently scarce data. In healthcare, training data is difficult to curate, triggering growing concerns that the current lack of access to healthcare by under-privileged social groups will translate into future bias in healthcare AIs. In this report, we developed an autoencoder to grow and enhance inherently scarce datasets to alleviate our dependence on big data. DesignComputational study with open-source data SubjectsThe data was obtained from six open-source datasets comprising patients aged 40 to 80 in Singapore, China, India, and Spain. MethodsThe reported framework generates synthetic images based on real-world patient imaging data. As a test case, we used autoencoder to expand publicly available training sets of optic disc photos, and evaluated the ability of the resultant datasets to train AI models in the detection of glaucomatous optic neuropathy.Main Outcome MeasuresAUC (Area Under the ROC Curve) were used to evaluate the performance of the glaucoma detector. A higher AUC indicates better detection performance. ResultsResults show that enhancing datasets with synthetic images generated by autoencoder led to superior training sets that improved the performance of AI models. ConclusionsOur findings here help address the increasingly untenable data volume and quality requirements for AI model development and have implications beyond healthcare, towards empowering AI adoption for all similarly data-challenged fields.

ASL MRI Denoising via Multi Channel Collaborative Low-Rank Regularization.

Arterial spin labeling (ASL) perfusion MRI is the only non-invasive imaging technique for quantifying regional cerebral blood flow (CBF), which is a fundamental physiological variable. ASL MRI has a relatively low signal-to-noise-ratio (SNR). In this study, we proposed a novel ASL denoising method by simultaneously exploiting the inter- and intra-receive channel data correlations. MRI including ASL MRI data have been routinely acquired with multi-channel coils but current denoising methods are designed for denoising the coil-combined data. Indeed, the concurrently acquired multi-channel images differ only by coil sensitivity weighting and random noise, resulting in a strong low-rank structure of the stacked multi-channel data matrix. In our method, this matrix was formed by stacking the vectorized slices from different channels. Matrix rank was then approximately measured through the logarithm-determinant of the covariance matrix. Notably, our filtering technique is applied directly to complex data, avoiding the need to separate magnitude and phase or divide real and imaginary data, thereby ensuring minimal information loss. The degree of low-rank regularization is controlled based on the estimated noise level, striking a balance between noise removal and texture preservation. A noteworthy advantage of our framework is its freedom from parameter tuning, distinguishing it from most existing methods. Experimental results on real-world imaging data demonstrate the effectiveness of our proposed approach in significantly improving ASL perfusion quality. By effectively mitigating noise while preserving important textural information, our method showcases its potential for enhancing the utility and accuracy of ASL perfusion MRI, paving the way for improved neuroimaging studies and clinical diagnoses.

Empirical data drift detection experiments on real-world medical imaging data

While it is common to monitor deployed clinical artificial intelligence (AI) models for performance degradation, it is less common for the input data to be monitored for data drift – systemic changes to input distributions. However, when real-time evaluation may not be practical (eg., labeling costs) or when gold-labels are automatically generated, we argue that tracking data drift becomes a vital addition for AI deployments. In this work, we perform empirical experiments on real-world medical imaging to evaluate three data drift detection methods’ ability to detect data drift caused (a) naturally (emergence of COVID-19 in X-rays) and (b) synthetically. We find that monitoring performance alone is not a good proxy for detecting data drift and that drift-detection heavily depends on sample size and patient features. Our work discusses the need and utility of data drift detection in various scenarios and highlights gaps in knowledge for the practical application of existing methods.

Constraint-Aware Learning for Fractional Flow Reserve Pullback Curve Estimation from Invasive Coronary Imaging.

Estimation of the fractional flow reserve (FFR) pullback curve from invasive coronary imaging is important for the intraoperative guidance of coronary intervention. Machine/deep learning has been proven effective in FFR pullback curve estimation. However, the existing methods suffer from inadequate incorporation of intrinsic geometry associations and physics knowledge. In this paper, we propose a constraint-aware learning framework to improve the estimation of the FFR pullback curve from invasive coronary imaging. It incorporates both geometrical and physical constraints to approximate the relationships between the geometric structure and FFR values along the coronary artery centerline. Our method also leverages the power of synthetic data in model training to reduce the collection costs of clinical data. Moreover, to bridge the domain gap between synthetic and real data distributions when testing on real-world imaging data, we also employ a diffusion-driven test-time data adaptation method that preserves the knowledge learned in synthetic data. Specifically, this method learns a diffusion model of the synthetic data distribution and then projects real data to the synthetic data distribution at test time. Extensive experimental studies on a synthetic dataset and a real-world dataset of 382 patients covering three imaging modalities have shown the better performance of our method for FFR estimation of stenotic coronary arteries, compared with other machine/deep learning-based FFR estimation models and computational fluid dynamics-based model. The results also provide high agreement and correlation between the FFR predictions of our method and the invasively measured FFR values. The plausibility of FFR predictions along the coronary artery centerline is also validated.

Modeling tumor size dynamics based on real-world electronic health records and image data in advanced melanoma patients receiving immunotherapy.

The development of immune checkpoint inhibitors (ICIs) has revolutionized cancer therapy but only a fraction of patients benefits from this therapy. Model-informed drug development can be used to assess prognostic and predictive clinical factors or biomarkers associated with treatment response. Most pharmacometric models have thus far been developed using data from randomized clinical trials, and further studies are needed to translate their findings into the real-world setting. We developed a tumor growth inhibition model based on real-world clinical and imaging data in a population of 91 advanced melanoma patients receiving ICIs (i.e., ipilimumab, nivolumab, and pembrolizumab). Drug effect was modeled as an ON/OFF treatment effect, with a tumor killing rate constant identical for the three drugs. Significant and clinically relevant covariate effects of albumin, neutrophil to lymphocyte ratio, and Eastern Cooperative Oncology Group (ECOG) performance status were identified on the baseline tumor volume parameter, as well as NRAS mutation on tumor growth rate constant using standard pharmacometric approaches. In a population subgroup (n = 38), we had the opportunity to conduct an exploratory analysis of image-based covariates (i.e., radiomics features), by combining machine learning and conventional pharmacometric covariate selection approaches. Overall, we demonstrated an innovative pipeline for longitudinal analyses of clinical and imaging RWD with a high-dimensional covariate selection method that enabled the identification of factors associated with tumor dynamics. This study also provides a proof of concept for using radiomics features as model covariates.

Multiresolution Quality Inspection of Layerwise Builds for Metal 3D Printer and Scanner

Abstract Automated optical inspection (AOI) is increasingly advocated for in situ quality monitoring of additive manufacturing (AM) processes. The availability of layerwise imaging data improves the information visibility during fabrication processes and is thus conducive to performing online certification. However, few, if any, have investigated the high-speed contact image sensors (CIS) (i.e., originally developed for document scanners and multifunction printers) for AM quality monitoring. In addition, layerwise images show complex patterns and often contain hidden information that cannot be revealed in a single scale. A new and alternative approach will be to analyze these intrinsic patterns with multiscale lenses. Therefore, the objective of this article is to design and develop an AOI system with contact image sensors for multiresolution quality inspection of layerwise builds in additive manufacturing. First, we retrofit the AOI system with contact image sensors in industrially relevant 95 mm/s scanning speed to a laser-powder-bed-fusion (LPBF) machines. Then, we design the experiments to fabricate nine parts under a variety of factor levels (e.g., gas flow blockage, re-coater damage, laser power changes). In each layer, the AOI system collects imaging data of both recoating powder beds before the laser fusion and surface finishes after the laser fusion. Second, layerwise images are pre-preprocessed for alignment, registration, and identification of regions of interests (ROIs) of these nine parts. Then, we leverage the wavelet transformation to analyze ROI images in multiple scales and further extract salient features that are sensitive to process variations, instead of extraneous noises. Third, we perform the paired comparison analysis to investigate how different levels of factors influence the distribution of wavelet features. Finally, these features are shown to be effective in predicting the extent of defects in the computed tomography (CT) data of layerwise AM builds. The proposed framework of multiresolution quality inspection is evaluated and validated using real-world AM imaging data. Experimental results demonstrated the effectiveness of the proposed AOI system with contact image sensors for online quality inspection of layerwise builds in AM processes.

Privacy risks of whole-slide image sharing in digital pathology

Access to large volumes of so-called whole-slide images—high-resolution scans of complete pathological slides—has become a cornerstone of the development of novel artificial intelligence methods in pathology for diagnostic use, education/training of pathologists, and research. Nevertheless, a methodology based on risk analysis for evaluating the privacy risks associated with sharing such imaging data and applying the principle “as open as possible and as closed as necessary” is still lacking. In this article, we develop a model for privacy risk analysis for whole-slide images which focuses primarily on identity disclosure attacks, as these are the most important from a regulatory perspective. We introduce a taxonomy of whole-slide images with respect to privacy risks and mathematical model for risk assessment and design . Based on this risk assessment model and the taxonomy, we conduct a series of experiments to demonstrate the risks using real-world imaging data. Finally, we develop guidelines for risk assessment and recommendations for low-risk sharing of whole-slide image data.

Enhanced Pathology Image Quality with Restore–Generative Adversarial Network

Whole slide imaging is becoming a routine procedure in clinical diagnosis. Advanced image analysis techniques have been developed to assist pathologists in disease diagnosis, staging, subtype classification, and risk stratification. Recently, deep learning algorithms have achieved state-of-the-art performances in various imaging analysis tasks, including tumor region segmentation, nuclei detection, and disease classification. However, widespread clinical use of these algorithms is hampered by their performances often degrading due to image quality issues commonly seen in real-world pathology imaging data such as low resolution, blurring regions, and staining variation. Restore-Generative Adversarial Network (GAN), a deep learning model, was developed to improve the imaging qualities by restoring blurred regions, enhancing low resolution, and normalizing staining colors. The results demonstrate that Restore-GAN can significantly improve image quality, which leads to improved model robustness and performance for existing deep learning algorithms in pathology image analysis. Restore-GAN has the potential to be used to facilitate the applications of deep learning models in digital pathology analyses.

Predicting OCT images of short-term response to anti-VEGF treatment for retinal vein occlusion using generative adversarial network.

To generate and evaluate post-therapeutic optical coherence tomography (OCT) images based on pre-therapeutic images with generative adversarial network (GAN) to predict the short-term response of patients with retinal vein occlusion (RVO) to anti-vascular endothelial growth factor (anti-VEGF) therapy. Real-world imaging data were retrospectively collected from 1 May 2017, to 1 June 2021. A total of 515 pairs of pre-and post-therapeutic OCT images of patients with RVO were included in the training set, while 68 pre-and post-therapeutic OCT images were included in the validation set. A pix2pixHD method was adopted to predict post-therapeutic OCT images in RVO patients after anti-VEGF therapy. The quality and similarity of synthetic OCT images were evaluated by screening and evaluation experiments. We quantitatively and qualitatively assessed the prognostic accuracy of the synthetic post-therapeutic OCT images. The post-therapeutic OCT images generated by the pix2pixHD algorithm were comparable to the actual images in edema resorption response. Retinal specialists found most synthetic images (62/68) difficult to differentiate from the real ones. The mean absolute error (MAE) of the central macular thickness (CMT) between the synthetic and real OCT images was 26.33 ± 15.81 μm. There was no statistical difference in CMT between the synthetic and the real images. In this retrospective study, the application of the pix2pixHD algorithm objectively predicted the short-term response of each patient to anti-VEGF therapy based on OCT images with high accuracy, suggestive of its clinical value, especially for screening patients with relatively poor prognosis and potentially guiding clinical treatment. Importantly, our artificial intelligence-based prediction approach’s non-invasiveness, repeatability, and cost-effectiveness can improve compliance and follow-up management of this patient population.

Prediction of Incomplete Response of Primary Tumour Based on Clinical and Radiomics Features in Inoperable Head and Neck Cancers after Definitive Treatment.

Radical treatment of patients diagnosed with inoperable and locally advanced head and neck cancers (LAHNC) is still a challenge for clinicians. Prediction of incomplete response (IR) of primary tumour would be of value to the treatment optimization for patients with LAHNC. Aim of this study was to develop and evaluate models based on clinical and radiomics features for prediction of IR in patients diagnosed with LAHNC and treated with definitive chemoradiation or radiotherapy. Clinical and imaging data of 290 patients were included into this retrospective study. Clinical model was built based on tumour and patient related features. Radiomics features were extracted based on imaging data, consisting of contrast- and non-contrast-enhanced pre-treatment CT images, obtained in process of diagnosis and radiotherapy planning. Performance of clinical and combined models were evaluated with area under the ROC curve (AUROC). Classification performance was evaluated using 5-fold cross validation. Model based on selected clinical features including ECOG performance, tumour stage T3/4, primary site: oral cavity and tumour volume were significantly predictive for IR, with AUROC of 0.78. Combining clinical and radiomics features did not improve model’s performance, achieving AUROC 0.77 and 0.68 for non-contrast enhanced and contrast-enhanced images respectively. The model based on clinical features showed good performance in IR prediction. Combined model performance suggests that real-world imaging data might not yet be ready for use in predictive models.

Prediction of the Short-Term Therapeutic Effect of Anti-VEGF Therapy for Diabetic Macular Edema Using a Generative Adversarial Network with OCT Images.

Purpose: To generate and evaluate individualized post-therapeutic optical coherence tomography (OCT) images that could predict the short-term response of anti-vascular endothelial growth factor (VEGF) therapy for diabetic macular edema (DME) based on pre-therapeutic images using generative adversarial network (GAN). Methods: Real-world imaging data were collected at the Department of Ophthalmology, Qilu Hospital. A total of 561 pairs of pre-therapeutic and post-therapeutic OCT images of patients with DME were retrospectively included in the training set, 71 pre-therapeutic OCT images were included in the validation set, and their corresponding post-therapeutic OCT images were used to evaluate the synthetic images. A pix2pixHD method was adopted to predict post-therapeutic OCT images in DME patients that received anti-VEGF therapy. The quality and similarity of synthetic OCT images were evaluated independently by a screening experiment and an evaluation experiment. Results: The post-therapeutic OCT images generated by the GAN model based on big data were comparable to the actual images, and the response of edema resorption was also close to the ground truth. Most synthetic images (65/71) were difficult to differentiate from the actual OCT images by retinal specialists. The mean absolute error (MAE) of the central macular thickness (CMT) between the synthetic OCT images and the actual images was 24.51 ± 18.56 μm. Conclusions: The application of GAN can objectively demonstrate the individual short-term response of anti-VEGF therapy one month in advance based on OCT images with high accuracy, which could potentially help to improve treatment compliance of DME patients, identify patients who are not responding well to treatment and optimize the treatment program.

Measuring variability of local brain volume using improved volume preserved warping.

Measuring local brain volume is clinically important in neuroimaging studies. Voxel preserved warping (VPW) and Jacobian determinant are effective methods for studying local brain volume changes and variations (LBVCV) across multiple brains. However, these LBVCV methods typically depend on the local deformation without using the global deformation, while both deformations are needed in co-registering the brains under examination so that the brains can be compared on a common and fair basis. However, instead of employing a uniformed strategy, different co-registration methods have developed their own unique strategy in performing global and local transformation of the co-registration of the brains, and how the global and local transformations may combine to achieve the final goal of co-registration is not their concern, as long as the final registration may accomplish the co-registering job satisfactorily. The aforementioned inconsistency thus makes the LBVCV measurement that relies on the registration methods for studying local brain volumes totally unstable and actually unreliable. To address the uncertainty in measuring local brain volume variability caused by the un-uniqueness of performing global and local deformations during co-registration, the present study proposes new VPW approaches (VPWα and VPWβ), which no longer require the separation of the global and local transformation components but employ only the general deformation concatenating both components, as long as the general registration may achieve the task of co-registering brain images. The new VPW methods are validated in theory and in practice, using both simulated and real-world imaging data, respectively, based on two registration methods popularly in use by the neuroimaging research community, i.e., the Automatic Registration Toolbox (ART) and Symmetric Image Normalization Method (SyN) registration methods. Experiments using simulated data demonstrated that the proposed new VPW methods may reliably measure local brain volume changes and variability. In contrast, traditional methods typically may result in LBVCV maps containing significantly inconsistent even false findings. In the experiments using real neuroimaging datasets from a schizophrenia study, the results based on the proposed new VPW methods were highly consistent, no matter which registration method was employed. Otherwise, the LBVCV results based on traditional approaches would show significant difference, depending on the individual registration method that the analysis employed. LBVCV assessments based on traditional methods appear to be unreliable. The proposed new VPW methods for measuring local volume changes is independent of registration methods, and therefore can serve as alternative approaches for assessing LBVCV reliably.

Multicentre Study Using Machine Learning Methods in Clinical Diagnosis of Knee Osteoarthritis.

Knee osteoarthritis (OA) is one of the most common musculoskeletal disorders. OA diagnosis is currently conducted by assessing symptoms and evaluating plain radiographs, but this process suffers from the subjectivity of doctors. In this study, we retrospectively compared five commonly used machine learning methods, especially the CNN network, to predict the real-world X-ray imaging data of knee joints from two different hospitals using Kellgren-Lawrence (K-L) grade of knee OA to help doctors choose proper auxiliary tools. Furthermore, we present attention maps of CNN to highlight the radiological features affecting the network decision. Such information makes the decision process transparent for practitioners, which builds better trust towards such automatic methods and, moreover, reduces the workload of clinicians, especially for remote areas without enough medical staff.

Predicting Post-Therapeutic Visual Acuity and OCT Images in Patients With Central Serous Chorioretinopathy by Artificial Intelligence.

To predict visual acuity (VA) and post-therapeutic optical coherence tomography (OCT) images 1, 3, and 6 months after laser treatment in patients with central serous chorioretinopathy (CSC) by artificial intelligence (AI). Real-world clinical and imaging data were collected at Zhongshan Ophthalmic Center (ZOC) and Xiamen Eye Center (XEC). The data obtained from ZOC (416 eyes of 401 patients) were used as the training set; the data obtained from XEC (64 eyes of 60 patients) were used as the test set. Six different machine learning algorithms and a blending algorithm were used to predict VA, and a pix2pixHD method was adopted to predict post-therapeutic OCT images in patients after laser treatment. The data for VA predictions included clinical features obtained from electronic medical records (20 features) and measured features obtained from fundus fluorescein angiography, indocyanine green angiography, and OCT (145 features). The data for OCT predictions included 480 pairs of pre- and post-therapeutic OCT images. The VA and OCT images predicted by AI were compared with the ground truth. In the VA predictions of XEC dataset, the mean absolute errors (MAEs) were 0.074–0.098 logMAR (within four to five letters), and the root mean square errors were 0.096–0.127 logMAR (within five to seven letters) for the 1-, 3-, and 6-month predictions, respectively; in the post-therapeutic OCT predictions, only about 5.15% (5 of 97) of synthetic OCT images could be accurately identified as synthetic images. The MAEs of central macular thickness of synthetic OCT images were 30.15 ± 13.28 μm and 22.46 ± 9.71 μm for the 1- and 3-month predictions, respectively. This is the first study to apply AI to predict VA and post-therapeutic OCT of patients with CSC. This work establishes a reliable method of predicting prognosis 6 months in advance; the application of AI has the potential to help reduce patient anxiety and serve as a reference for ophthalmologists when choosing optimal laser treatments.

P10.14 ctDNA and Real-World Response (rwR) in Patients With Lung Cancer From A Prospective Real-World Clinico-Genomic (PCG) Study

Prospective real-world data studies integrated into routine clinical workflows can capture multimodal longitudinal data in patients (pts) with cancer treated in community settings. One modality is circulating tumor DNA (ctDNA), a minimally invasive, readily available test for genomic profiling, investigating resistance or monitoring disease. Using the PCG study platform to prospectively collect and link real-world clinico-genomic, imaging and outcomes data, this exploratory analysis assessed the association of longitudinal ctDNA measurement and clinical response.

P09.55 A Platform to Prospectively Link Real-World Clinico-Genomic, Imaging, and Outcomes Data for Patients With Lung Cancer

P09.55 A Platform to Prospectively Link Real-World Clinico-Genomic, Imaging, and Outcomes Data for Patients With Lung Cancer

Multicenter, Head-to-Head, Real-World Validation Study of Seven Automated Artificial Intelligence Diabetic Retinopathy Screening Systems.

OBJECTIVEWith rising global prevalence of diabetic retinopathy (DR), automated DR screening is needed for primary care settings. Two automated artificial intelligence (AI)–based DR screening algorithms have U.S. Food and Drug Administration (FDA) approval. Several others are under consideration while in clinical use in other countries, but their real-world performance has not been evaluated systematically. We compared the performance of seven automated AI-based DR screening algorithms (including one FDA-approved algorithm) against human graders when analyzing real-world retinal imaging data.RESEARCH DESIGN AND METHODSThis was a multicenter, noninterventional device validation study evaluating a total of 311,604 retinal images from 23,724 veterans who presented for teleretinal DR screening at the Veterans Affairs (VA) Puget Sound Health Care System (HCS) or Atlanta VA HCS from 2006 to 2018. Five companies provided seven algorithms, including one with FDA approval, that independently analyzed all scans, regardless of image quality. The sensitivity/specificity of each algorithm when classifying images as referable DR or not were compared with original VA teleretinal grades and a regraded arbitrated data set. Value per encounter was estimated.RESULTSAlthough high negative predictive values (82.72–93.69%) were observed, sensitivities varied widely (50.98–85.90%). Most algorithms performed no better than humans against the arbitrated data set, but two achieved higher sensitivities, and one yielded comparable sensitivity (80.47%, P = 0.441) and specificity (81.28%, P = 0.195). Notably, one had lower sensitivity (74.42%) for proliferative DR (P = 9.77 × 10−4) than the VA teleretinal graders. Value per encounter varied at $15.14–$18.06 for ophthalmologists and $7.74–$9.24 for optometrists.CONCLUSIONSThe DR screening algorithms showed significant performance differences. These results argue for rigorous testing of all such algorithms on real-world data before clinical implementation.

A multi-stakeholder platform to prospectively link longitudinal real-world clinico-genomic, imaging, and outcomes data for patients with metastatic lung cancer.

TPS2087 Background: Making personalized diagnostics and treatments a reality for every cancer patient necessitates comprehensively capturing the patient journey. Real-world data has shown promise for the future of clinical research and advancing precision medicine. However, certain limitations exist such as data quality management as well as bias and confounding factors associated with retrospective analyses. We present a multi-stakeholder platform to prospectively collect and link real-world clinico-genomic, imaging, and outcomes data to longitudinal blood genomic profiling for lung cancer. Methods: This study is enrolling approximately 1000 patients with metastatic non-small cell lung cancer or extensive-stage small cell lung cancer who will initiate standard-of-care systemic anti-neoplastic treatment, regardless of line of therapy, at 20 community oncology and academic practices within the Flatiron Health network. Relevant clinical data points from both structured and unstructured fields will be collected through the electronic health records via technology-enabled abstraction, eliminating the need for case report forms. Digital pathology and clinical images at standard-of-care visits will be collected. Blood samples for circulating tumor DNA (ctDNA) profiling using FoundationOne Liquid will be collected at three timepoints: enrollment, first tumor assessment, and end of treatment. Tumor tissue samples may be submitted at baseline for genomic profiling using FoundationOne CDx. Overall survival follow-up will occur until death, withdrawal of consent, loss to follow-up, or end of study. The objectives are to evaluate 1) the feasibility of building a scalable, prospective platform and 2) the associations between ctDNA and real-world clinical outcomes, including overall survival. Enrollment is ongoing. Clinical trial information: NCT04180176.