ChestView US (1.1)

{0} FDA U.S. FOOD & DRUG ADMINISTRATION September 9, 2026 Gleamer Sas % Antoine Tournier Chief Compliance Officer 47bis Rue Des Vinaigriers PARIS, 75010 FRANCE Re: K252486 Trade/Device Name: ChestView US (1.1) Regulation Number: 21 CFR 892.2070 Regulation Name: Medical Image Analyzer Regulatory Class: Class II Product Code: MYN Dated: August 10, 2026 Received: August 10, 2026 Dear Antoine Tournier: We have reviewed your section 510(k) premarket notification of intent to market the device referenced above and have determined the device is substantially equivalent (for the indications for use stated in the enclosure) to legally marketed predicate devices marketed in interstate commerce prior to May 28, 1976, the enactment date of the Medical Device Amendments, or to devices that have been reclassified in accordance with the provisions of the Federal Food, Drug, and Cosmetic Act (the Act) that do not require approval of a premarket approval application (PMA). You may, therefore, market the device, subject to the general controls provisions of the Act. Although this letter refers to your product as a device, please be aware that some cleared products may instead be combination products. The 510(k) Premarket Notification Database available at https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm identifies combination product submissions. The general controls provisions of the Act include requirements for annual registration, listing of devices, good manufacturing practice, labeling, and prohibitions against misbranding and adulteration. Please note: CDRH does not evaluate information related to contract liability warranties. We remind you, however, that device labeling must be truthful and not misleading. If your device is classified (see above) into either class II (Special Controls) or class III (PMA), it may be subject to additional controls. Existing major regulations affecting your device can be found in the Code of Federal Regulations, Title 21, Parts 800 to 898. In addition, FDA may publish further announcements concerning your device in the Federal Register. U.S. Food & Drug Administration 10903 New Hampshire Avenue Silver Spring, MD 20993 www.fda.gov {1} K252486 - Antoine Tournier Page 2 FDA's substantial equivalence determination also included the review and clearance of your Predetermined Change Control Plan (PCCP). Under section 515C(b)(1) of the Act, a new premarket notification is not required for a change to a device cleared under section 510(k) of the Act, if such change is consistent with an established PCCP granted pursuant to section 515C(b)(2) of the Act. Under 21 CFR 807.81(a)(3), a new premarket notification is required if there is a major change or modification in the intended use of a device, or if there is a change or modification in a device that could significantly affect the safety or effectiveness of the device, e.g., a significant change or modification in design, material, chemical composition, energy source, or manufacturing process. Accordingly, if deviations from the established PCCP result in a major change or modification in the intended use of the device, or result in a change or modification in the device that could significantly affect the safety or effectiveness of the device, then a new premarket notification would be required consistent with section 515C(b)(1) of the Act and 21 CFR 807.81(a)(3). Failure to submit such a premarket submission would constitute adulteration and misbranding under sections 501(f)(1)(B) and 502(o) of the Act, respectively. Additional information about changes that may require a new premarket notification are provided in the FDA guidance documents entitled "Deciding When to Submit a 510(k) for a Change to an Existing Device" (https://www.fda.gov/media/99812/download) and "Deciding When to Submit a 510(k) for a Software Change to an Existing Device" (https://www.fda.gov/media/99785/download). Your device is also subject to, among other requirements, the Quality Management System Regulation (QMSR) (21 CFR Part 820), which includes, but is not limited to, ISO 13485 clause 7.3 (Design controls), ISO 13485 clause 8.3 (Nonconforming product), ISO 13485 clause 8.5.2 (Corrective action), and ISO 13485 clause 8.5.3 (Preventative action). Please note that regardless of whether a change requires premarket review, the QMSR requires device manufacturers to review and approve changes to device design and production (ISO 13485 clause 7.3 and ISO 13485 clause 7.5) and document changes and approvals in the Medical Device File (ISO 13485 clause 4.2.3). Please be advised that FDA's issuance of a substantial equivalence determination does not mean that FDA has made a determination that your device complies with other requirements of the Act or any Federal statutes and regulations administered by other Federal agencies. You must comply with all the Act's requirements, including, but not limited to: registration and listing (21 CFR Part 807); labeling (21 CFR Part 801); medical device reporting (reporting of medical device-related adverse events) (21 CFR Part 803) for devices or postmarketing safety reporting (21 CFR Part 4, Subpart B) for combination products (see https://www.fda.gov/combination-products/guidance-regulatory-information/postmarketing-safety-reporting-combination-products); good manufacturing practice requirements as set forth in the Quality Management System Regulation (QMSR) (21 CFR Part 820) for devices or current good manufacturing practices (21 CFR Part 4, Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR Parts 1000-1050. All medical devices, including Class I and unclassified devices and combination product device constituent parts are required to be in compliance with the final Unique Device Identification System rule ("UDI Rule"). The UDI Rule requires, among other things, that a device bear a unique device identifier (UDI) on its label and package (21 CFR 801.20(a)) unless an exception or alternative applies (21 CFR 801.20(b)) and that the dates on the device label be formatted in accordance with 21 CFR 801.18. The UDI Rule (21 CFR 830.300(a) and 830.320(b)) also requires that certain information be submitted to the Global Unique Device Identification Database (GUDID) (21 CFR Part 830 Subpart E). For additional information on these {2} K252486 - Antoine Tournier Page 3 requirements, please see the UDI System webpage at https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/unique-device-identification-system-udi-system. Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR 807.97). For questions regarding the reporting of adverse events under the MDR regulation (21 CFR Part 803), please go to https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems. For comprehensive regulatory information about medical devices and radiation-emitting products, including information about labeling regulations, please see Device Advice (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance) and CDRH Learn (https://www.fda.gov/training-and-continuing-education/cdrh-learn). Additionally, you may contact the Division of Industry and Consumer Education (DICE) to ask a question about a specific regulatory topic. See the DICE website (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/contact-us-division-industry-and-consumer-education-dice) for more information or contact DICE by email (DICE@fda.hhs.gov) or phone (1-800-638-2041 or 301-796-7100). Sincerely, Lu Jiang, Ph.D. Assistant Director DHT8B: Division of Radiologic Imaging Devices and Electronic Products OHT8: Office of Radiological Health Office of Product Evaluation and Quality Center for Devices and Radiological Health Enclosure {3} | Indications for Use | | | | --- | --- | --- | | Please type in the marketing application/submission number, if it is known. This textbox will be left blank for original applications/submissions. | K252486 | ? | | Please provide the device trade name(s). | | ? | | ChestView US (1.1) | | | | Please provide your Indications for Use below. | | ? | | ChestView US is a radiological Computer-Assisted Detection (CADe) software device that analyzes frontal and lateral chest radiographs of patients presenting with symptoms (e.g. dyspnea, cough, pain) or suspected for findings related to regions of interest (ROIs) in the lungs, airways, mediastinum/hila and pleural space. The device uses machine learning techniques to identify and produces boxes around the ROIs. The boxes are labeled with one of the following radiographic findings: Nodule, Pleural space abnormality, Mediastinum/Hila abnormality, and Consolidation. | | | | ChestView US is intended for use as a concurrent reading aid for radiologists and emergency medicine physicians. It does not replace the role of radiologists and emergency medicine physicians or of other diagnostic testing in the standard of care. ChestView US is for prescription use only and is indicated for adults only. | | | | Please select the types of uses (select one or both, as applicable). | ☑ Prescription Use (Part 21 CFR 801 Subpart D) ☐ Over-The-Counter Use (21 CFR 801 Subpart C) | ? | {4} 510(k) Summary GLEAMER – ChestView US Gleamer Date prepared: July 26th, 2026 In accordance with 21 CFR 807.87(h) and 21 CFR 807.92 the 510(k) Summary for ChestView US 1.1 is provided below. ## 1. Submitter | Submitter | GLEAMER SAS 47bis, rue des Vinaigriers 75010, Paris 10, FRANCE | | --- | --- | | Primary Contact Person | Antoine Tournier Chief Compliance Officer Tel: 0033 6 15 81 23 45 Email: antoine.tournier@gleamer.ai Alternate email: qara@gleamer.ai | ## 2. Device | Trade Name | ChestView US (1.1) | | --- | --- | | 510(k) Number | K252486 | | Common Name | Medical Image Analyzer | | Classification Name | Medical Image Analyzer | | Regulation | 21 CFR 892.2070 | | Product Code | MYN | | Classification | Class II | ## 3. Predicate Device | Predicate Device | ChestView US | | --- | --- | | 510(k) Number | K241620 | ## 4. Device Description ChestView US is a radiological Computer-Assisted Detection (CADe) software device intended to analyze frontal and lateral chest radiographs for suspicious regions of interest (ROIs): Nodule, Consolidation, Pleural Space Abnormality and Mediastinum/Hila Abnormality. The nodule ROI category was developed from images with focal nonlinear opacity with a generally spherical shape situated in the pulmonary interstitium. The consolidation ROI category was developed from images with area of increased attenuation of lung parenchyma due to the replacement of air in the alveoli. The pleural space abnormality ROI category was developed from images with: - Pleural Effusion that is an abnormal presence of fluid in the pleural space {5} 510(k) Summary GLEAMER – ChestView US Gleamer - Pneumothorax that is an abnormal presence of air or gas in the pleural space that separates the parietal and the visceral pleura The mediastinum/hila abnormality ROI category was developed from images with enlargement of the mediastinum or the hilar region with a deformation of its contours. ChestView US can be deployed on the cloud and be connected to several computing platforms and X-ray imaging platforms such as radiographic systems, or PACS. More precisely, ChestView US can be deployed in the cloud connected to a DICOM Source/Destination with a DICOM Viewer, i.e. a PACS. After the acquisition of the radiographs of the patient and their storage in the DICOM Source, the radiographs are automatically received by ChestView US from the user's DICOM Source through intermediate DICOM node(s) (for example, a specific Gateway, or a dedicated API). The DICOM Source can be the user's image storage system (for example, the Picture Archiving and Communication System, or PACS), or other radiological equipment (for example X-ray systems). Once received by ChestView US, the radiographs are automatically processed by the AI algorithm to identify regions of interest. Based on the processing result, ChestView US generates result files in DICOM format. These result files consist of annotated images with boxes drawn around the regions of interest on a copy of all images (as an overlay). ChestView US does not alter the original images, nor does it change the order of original images or delete any image from the DICOM Source. Once available, the result files are sent by ChestView US to the DICOM Destination through the same intermediate DICOM node(s). Similar to the DICOM Source, the DICOM Destination can be the user's image storage system (for example, the Picture Archiving and Communication System, or PACS), or other radiological equipment (for example X-ray systems). The DICOM Source and the DICOM Destination are not necessarily identical. The DICOM Destination can be used to visualize the result files provided by ChestView US or to transfer the results to another DICOM host for visualization. The users are then able to use them as a concurrent reading aid to provide their diagnosis. For each exam analyzed by ChestView US, a DICOM Secondary Capture is generated. If any ROI is detected by ChestView US, the output DICOM image includes a copy of the original images of the study and the following information: - Above the images, a header with the text "CHESTVIEW ROI(s)" and the list of the findings detected in the image. - Around the ROI(s), a bounding box with a solid or dotted line depending on the confidence of the algorithm and the type of ROI written above the box: - Dotted-line Bounding Box: Identified region of interest when the confidence degree of the AI algorithm associated with the possible finding is above "high-sensitivity operating point" and below "high specificity operating point" displayed as a dotted bounding box around the area of interest. {6} 510(k) Summary GLEAMER – ChestView US Gleamer - ○ **Solid-line Bounding Box:** Identified region of interest when the confidence degree of the AI algorithm associated with the finding is above “high-specificity operating point” displayed as a solid bounding box around the area of interest. - ● Below the images, a footer with: - ○ The scope of ChestView US to allow the user to always have available the list of ROI type that are in the indications for use of the device and avoid any risk of confusion or misinterpretation of the types of ROI detected by ChestView US. - ○ The total number of regions of interest identified by ChestView US on the exam (sum of solid-line and dotted-line bounding boxes) If no ROI is detected by ChestView US, the output DICOM image includes a copy of the original images of the study and the text “NO CHESTVIEW ROI” with the scope of ChestView US to allow the user to always have available the list of ROI type that are in the indications for use of the device and avoid any risk of confusion or misinterpretation of the types of ROI detected by ChestView US. Finally, if the processing of the exam by ChestView US is not possible because it is outside the indications for use of the device or some information is missing to allow the processing, the output DICOM image includes a copy of the original images of the study and, in a header, the text “OUT OF SCOPE” and a caution message explaining the reason why no result was provided by the device. ## 5. Intended use/ Indications for Use ChestView US is a radiological Computer-Assisted Detection (CADe) software device that analyzes frontal and lateral chest radiographs of patients presenting with symptoms (e.g. dyspnea, cough, pain) or suspected for findings related to regions of interest (ROIs) in the lungs, airways, mediastinum/hila and pleural space. The device uses machine learning techniques to identify and produces boxes around the ROIs. The boxes are labeled with one of the following radiographic findings: Nodule, Pleural space abnormality, Mediastinum/Hila abnormality, and Consolidation. ChestView US is intended for use as a concurrent reading aid for radiologists and emergency medicine physicians. It does not replace the role of radiologists and emergency medicine physicians or of other diagnostic testing in the standard of care. ChestView US is for prescription use only and is indicated for adults only. ## 6. Substantial equivalence The predicate device for ChestView US in its version 1.1 is ChestView US in its version 1.0 (K241620). ChestView US has the following FDA-cleared Indications for Use: *ChestView US is a radiological Computer-Assisted Detection (CADe) software device that analyzes frontal and lateral chest radiographs of patients presenting with symptoms (e.g. dyspnea, cough, pain) or suspected for findings related to regions of interest (ROIs) in the lungs, airways, mediastinum/hila and pleural space. The device uses machine learning techniques to identify and produces boxes around the ROIs. The boxes are* {7} 510(k) Summary GLEAMER – ChestView US Gleamer labeled with one of the following radiographic findings: Nodule, Pleural space abnormality, Mediastinum/Hila abnormality, and Consolidation. ChestView US is intended for use as a concurrent reading aid for radiologists and emergency medicine physicians. It does not replace the role of radiologists and emergency medicine physicians or of other diagnostic testing in the standard of care. ChestView US is for prescription use only and is indicated for adults only. Table 1 provides a comparison of the Indications for Use and Technological Characteristics of ChestView US in its version 1.1 to the predicate ChestView US in its version 1.0. {8} 510(k) Summary GLEAMER – ChestView US [LOGO] Gleamer | Features and Characteristics | Subject Device Gleamer ChestView US – version 1.1 | Predicate Device Gleamer ChestView US – version 1.0 | | --- | --- | --- | | **Regulation Information** | | | | Classification regulation | 21 CFR 892.2070 – Medical image analyzer | Same | | Product Code | MYN | Same | | Regulation Description | Medical image analyzers, including computer-assisted/aided detection (CADe) devices for mammography breast cancer, ultrasound breast lesions, radiograph lung nodules, and radiograph dental caries detection, is a prescription device that is intended to identify, mark, highlight, or in any other manner direct the clinicians' attention to portions of a radiology image that may reveal abnormalities during interpretation of patient radiology images by the clinicians. This device incorporates pattern recognition and data analysis capabilities and operates on previously acquired medical images. | Same | | **Indications for Use** | | | | Image Modality | X-ray | Same | | Study type | Chest | Same | | Clinical Output | When regions of interest (ROIs) are detected: Identify and mark ROIs on chest radiographs and label the box around the ROI as one of the following: Nodule, Pleural space abnormality, Mediastinum/Hila abnormality, and Consolidation. When no ROI is detected: Mark the chest radiographs as “NO CHESTVIEW ROI” and provide the scope of the indications for use of ChestView US. | Same | | Intended Users | Radiologists/emergency medicine physicians | Same | {9} 510(k) Summary GLEAMER – ChestView US [Non-Text] Gleamer | Intended User Workflow | Device intended for use as a concurrent reading aid for users interpreting chest radiographs | Same | | --- | --- | --- | | Patient population | Adults with Chest Radiographs | Same | | Technological Information | | | | Machine Learning Methodology | Supervised Deep learning | Same | | Deployment Platform | Secure cloud-based processing and delivery of chest radiographs | Same | | Image Source | Digital X-ray | Same | | Image Viewing | Image displayed on PACS system | Same | Table 1: Indications for Use and Technological Comparison between ChestView US 1.1 and ChestView US 1.0 As outlined in the comparison table above, ChestView US version 1.1 shares identical technological features, indications for use, intended users, and workflow with its predicate device, ChestView US version 1.0 (K241620). Both versions analyze chest radiographs, identify and categorize regions of interest (ROIs) as Nodule, Pleural space abnormality, Mediastinum/Hila abnormality, or Consolidation, and are intended for use as concurrent reading aids to assist radiologists and emergency medicine physicians interpreting chest radiographs. The sole difference between ChestView US version 1.0 and version 1.1 is the retraining of the underlying machine learning algorithm, together with an update of the model architecture and its initialization: the object-detection (OD) backbone was changed to a vision foundation model, and the number of False-Positive Classifiers (FPC) was reduced. This modification is a backbone initialization followed by task-specific fine-tuning performed according to standard, previously established practices. It does not alter the device's intended use, indications, clinical outputs, or technological characteristics, and does not raise new questions regarding safety and effectiveness. Therefore, ChestView US version 1.1 is considered substantially equivalent to its predicate device, ChestView US version 1.0, maintaining the same level of safety and effectiveness. {10} 510(k) Summary GLEAMER – ChestView US Gleamer ## 7. Performance data ### 7.1. Software Verification and Validation Testing Product verification and validation testing were conducted and documented as per the requirements of the FDA guidance “Content of Premarket Submissions for Device Software Functions” for a Basic Documentation Level. Nonclinical tests include unit, integration and system levels testing for the final software version. ChestView US performed as intended and all results observed were as expected. All software requirements and risk analysis have been successfully verified and traced. ### 7.2. Bench testing Gleamer conducted a standalone clinical performance study on 3,837 chest radiograph cases representative of the intended use population. The results of the standalone testing demonstrated that ChestView US detects ROIs with high Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC) curve, high sensitivity and high specificity: Table 2: AUC of the ROC Curve for ChestView US Model Predictions by ROI – Comparison of performances | ROI Type | AUC at the exam/patient level - 95% Bootstrap CI | | | | --- | --- | --- | --- | | | Previous | New | Difference | | PLEURAL SPACE ABNORMALITY | 0.959 [0.954, 0.964] | 0.966 [0.962, 0.970] | 0.007 [-0.002, 0.016] | | CONSOLIDATION | 0.943 [0.938, 0.948] | 0.958 [0.953, 0.964] | 0.015 [0.005, 0.026] | | NODULE | 0.906 [0.898, 0.914] | 0.921 [0.913, 0.929] | 0.015 [-0.001, 0.030] | | MEDIASTINUM/HILA ABNORMALITY | 0.881 [0.871, 0.890] | 0.899 [0.889, 0.908] | 0.018 [-0.001, 0.037] | Table 3: Specificity (with 95% Bootstrap CI, at the high specificity operating point) of ChestView US at the exam/patient level on the standalone dataset – Comparison of performances | ROI Type | Specificity at the exam/patient level - 95% Bootstrap CI | | | | --- | --- | --- | --- | | | Previous | New | Difference | | PLEURAL SPACE ABNORMALITY | 0.959 [0.950, 0.965] | 0.967 [0.960, 0.972] | 0.008 [-0.005, 0.022] | | CONSOLIDATION | 0.983 [0.980, 0.987] | 0.981 [0.976, 0.986] | -0.003 [-0.011, 0.006] | | NODULE | 0.994 [0.990, 0.997] | 0.984 [0.979, 0.988] | -0.009 [-0.018, -0.002] | | MEDIASTINUM/HILA ABNORMALITY | 0.993 [0.991, 0.996] | 0.994 [0.991, 0.996] | 0.001 [-0.005, 0.005] | Table 4: Sensitivity (with 95% Bootstrap CI, at the high sensitivity operating point) of ChestView US at the exam/patient level on the standalone dataset – Comparison of performances | ROI Type | Sensitivity at the exam/patient level - 95% Bootstrap CI | | | | --- | --- | --- | --- | | | Previous | New | Difference | | PLEURAL SPACE ABNORMALITY | 0.875 [0.850, 0.894] | 0.910 [0.891, 0.923] | 0.035 [-0.003, 0.073] | | CONSOLIDATION | 0.831 [0.809, 0.864] | 0.853 [0.829, 0.874] | 0.023 [-0.035, 0.065] | | NODULE | 0.750 [0.711, 0.772] | 0.804 [0.768, 0.824] | 0.054 [-0.004, 0.113] | | MEDIASTINUM/HILA ABNORMALITY | 0.620 [0.588, 0.674] | 0.693 [0.645, 0.732] | 0.073 [-0.029, 0.143] | These standalone performances of the updated version of ChestView US 1.1 were compared to the performances of the predicate device, ChestView US 1.0. Based on pre-specified acceptance criteria, the {11} 510(k) Summary GLEAMER – ChestView US Gleamer comparative standalone performance evaluation demonstrated that ChestView US 1.1 is non-inferior to ChestView US 1.0 for the primary endpoint (AUC) and the secondary endpoints (Sensitivity and Specificity at both operating points), at both the Exam/Patient and ROI levels, using pre-specified non-inferiority margins. Side-by-side performance comparisons at the Exam/Patient level are presented in Table 2, 3, and 4 for the primary endpoint and for the secondary endpoints at their respective operating points; the complete comparative results are reported in the Standalone Performance Study. Descriptive comparison of point estimates indicates that point estimates of version 1.1 are generally favorable to those of version 1.0 across the majority of evaluated regions of interest, with the limited exceptions visible in the tables above; this descriptive observation is not interpreted as a superiority claim, as superiority was not formally tested as a hypothesis. ### 7.3. Clinical Testing No clinical studies were conducted in support of the 510(k) submission of ChestView US 1.1. ### 8. Predetermined Change Control Plan This submission is including a Predetermined Change Control Plan (PCCP). Modifications to the algorithm of ChestView US will be made in accordance with the guidelines outlined in the “Marketing Submission Recommendations for a Predetermined Change Control Plan for Artificial Intelligence-Enabled Device Software Functions”. The PCCP describes the modifications that can occur in a controlled manner ensuring the continued safety, effectiveness and performance of the device. There are two modifications outlined in the PCCP, and accordingly, any change will only occur after training, testing and validating the algorithm after which the change will be locked. When the modification is locked this will be communicated to all users through updated labelling via the Instructions for use. Verification and Validation activities will be conducted in accordance with the PCCP to compare the performance of the modified ChestView US algorithms to both the original cleared model and the most recently deployed model to ensure the safety and effectiveness. The allowed modifications are detailed in Table 5 below. Table 5: Details of modifications of ChestView US allowed in the PCCP | Modification #1: Model Re-training with New or Expanded Data | | | --- | --- | | Characteristics | Retrain or fine-tune the ChestView US AI model using an expanded training and/or tuning dataset of chest X-ray images limited to the data categories enumerated in the PCCP, without any change to the model architecture or pre-/post-processing pipeline. The algorithmic pipeline and output remain the same: the software processes frontal/lateral chest X-rays and outputs the likelihood and location of existing ROI categories only. No new imaging modality or hardware is introduced, and input image requirements (DICOM chest radiographs in specified orientation) remain unchanged. | | Rationale | This modification allows Gleamer to leverage new data and clinical experience to improve the algorithm's accuracy and robustness over time. By training on larger | {12} 510(k) Summary GLEAMER – ChestView US Gleamer | | and more representative datasets (for instance, including images from additional hospitals or increasing dataset sizes), the AI can better recognize subtle findings and reduce false positives/negatives. The rationale is to continuously enhance clinical performance within the cleared indication, thereby improving radiologists' detection support without changing the device's fundamental function. New data becomes available through ongoing use and research, and retraining ensures the algorithm keeps pace with real-world variability (e.g. imaging technique differences or emerging patterns) while staying within the original intended use. | | --- | --- | | Performance Specifications | The modified model is deployed only if it meets all comparison non-inferiority criteria against both comparators and all primary-endpoint standalone criteria; failure of any of these is recorded as an unresolvable failure and the model is not deployed. If a standalone secondary-endpoint minimum or a subgroup criterion is not met, deployment is governed by pre-specified risk-benefit assessment: the model is not deployed if the deviation may compromise safety or effectiveness. | | Implementation | Manual update: model retraining is performed by the manufacturer's development team; no autonomous/self-learning occurs on deployed units) Homogeneous deployment: the updated model version is rolled out universally to all users/platforms of ChestView US, so that every user operates the same algorithm version. | | Impact Assessment | Benefits: Improved detection performance for existing ROIs; increased robustness through exposure to a broader dataset; insurance that the algorithm stays current with any shifts in imaging practices or patient populations over time Risks: Potential unintended performance degradation, overfitting, or uneven performance across subgroups Risk mitigation: Use of representative datasets, comparative performance testing against both the original cleared version and the most recently deployed version, predefined acceptance criteria, subgroup analyses, locked models post-validation | | Modification #2: Algorithm Architecture or Parameter Updates | | | Characteristics | Modify the AI algorithm's model architecture or tuning parameters in order to maintain or improve performance on the existing detection tasks. These changes are limited to the specific modifications enumerated in the PCCP, each mapped to a component of the current cleared multi-stage detection pipeline, which consists of an object detection (OD) stage and two False Positive Classifier (FPC) stages, whose ensemble yields the final detection confidence. An architecture modification might involve replacing/improving one of these components (such as using a more powerful object detection network, using an additional FPC with a different design, replacing the backbone of one component) to boost accuracy. Importantly, these changes will not alter the device's inputs or outputs: ChestView US will continue to take the same chest X-ray images and output the same four types of ROI findings, with the only difference being how the software's internal model processes the images to identify abnormalities. Any such architectural | {13} 510(k) Summary GLEAMER – ChestView US [LOGO] Gleamer | | update will be designed to fit within the existing algorithm pipeline and adhere to the same input/output format and operating principles. Architectural updates under this PCCP are limited to modifications that do not change the fundamental model paradigm, output semantics, or clinical behavior of the device. The algorithm will remain a CADe detection system and will not introduce classification-only outputs, generative inference mechanisms, self-supervised inference at deployment, or changes to confidence score interpretation. | | --- | --- | | Rationale | Advancements in machine learning techniques occur rapidly. This modification allows ChestView US to incorporate technological improvements (such as more effective network architectures or improved training methodologies) to enhance accuracy, robustness, and generalization over time. By integrating architectural refinements, the algorithm can better recognize subtle findings, adapt to variability in imaging techniques, and reduce false positives and false negatives. For example, introducing a new deep learning architecture may improve the recognition of complex radiographic patterns or enable more refined analysis of object detection proposals. The rationale is to continuously strengthen clinical performance within the cleared indication, ensuring the device keeps pace with real-world variability while maintaining its fundamental function and rigorously controlling for safety. | | Performance Specifications | The modified model is deployed only if it meets all comparison non-inferiority criteria against both comparators and all primary-endpoint standalone criteria; failure of any of these is recorded as an unresolvable failure and the model is not deployed. If a standalone secondary-endpoint minimum or a subgroup criterion is not met, deployment is governed by pre-specified risk-benefit assessment: the model is not deployed if the deviation may compromise safety or effectiveness. | | Implementation | Manual update: model retraining is performed by the manufacturer's development team; no autonomous/self-learning occurs on deployed units) Homogeneous deployment: the updated model version is rolled out universally to all users/platforms of ChestView US, so that every user operates the same algorithm version. | | Impact Assessment | Benefits: Improved detection performance for existing ROIs; increased robustness through the use of state-of-the-art AI trainings Risks: Potential unintended performance degradation, unintended changes in algorithm behavior or impact on integrated software operation Risk mitigation: Use of representative datasets, comparative performance testing against both the original cleared version and the most recently deployed version, predefined acceptance criteria, re-execution of relevant system-level V&V testing, locked models post-validation | {14} 510(k) Summary GLEAMER – ChestView US Gleamer ## 9. Conclusion ChestView US version 1.1 is substantially equivalent to its predicate device, ChestView US version 1.0 (K241620). A comparative standalone study demonstrated that version 1.1 performs as well as the predicate device across the evaluated regions of interest, with non-inferiority to version 1.0 demonstrated for the primary endpoint (AUC) and the secondary endpoints (Sensitivity and Specificity at both operating points) at both the Exam/Patient and ROI levels, using pre-specified non-inferiority margins. Descriptive comparison of point estimates indicates that for the primary endpoint (AUC) at the Exam/Patient level, point estimates of version 1.1 are favorable to those of version 1.0 across the evaluated regions of interest. For the secondary endpoints (Sensitivity and Specificity) at both the Exam/Patient and ROI levels, point estimates of version 1.1 are generally favorable to those of version 1.0 across the majority of evaluated regions of interest, with the limited exceptions presented for full transparency in Table 3 and 4. This descriptive observation is presented for transparency and is not interpreted as a superiority claim, as superiority was not formally tested as a hypothesis. The modification implemented in ChestView US 1.1 does not affect the device's intended use, clinical outputs, or technological characteristics, confirming continued compliance with FDA regulatory requirements.