K261679 · Deep Breathe, Inc. · POK · Oct 23, 2026 · Radiology
Device Facts
Record ID
K261679
Device Name
BlineSlide
Applicant
Deep Breathe, Inc.
Product Code
POK · Radiology
Decision Date
Oct 23, 2026
Decision
SESE
Submission Type
Traditional
Regulation
21 CFR 892.2060
Device Class
Class 2
Attributes
AI/ML, Software as a Medical Device, Real-World Evidence
Real-World Evidence
Submission
Device
Sponsor
RWD Sources
RWE Use Summary
Key Tags
K261679 · Oct 23, 2026
BlineSlide
Deep Breathe, Inc.
Retrospective clinical lung ultrasound cine loops (cines) from multiple clinical sites
Retrospective clinical data was used to evaluate the standalone performance of the Pleural Line Motion Artifact Detection Module and to conduct a Multiple Reader Multiple Case (MRMC) study to assess the diagnostic accuracy of physicians aided by the device compared to unaided performance.
25 (US Board Certification in Emergency Medicine, Internal Medicine, or Critical Care Medicine)
Indications for Use
BlineSlide is intended for use as a concurrent reading aid to assist interpreting physicians, after lung ultrasound examinations of patients aged 18 years or older, by automatically identifying the absence of pleural line motion (lung sliding or lung pulse). The device provides information on abnormal findings that may be useful in rendering potential diagnoses. Patient management decisions should not be made solely on the results of the BlineSlide's analysis.
Device Story
BlineSlide is a cloud-based CADx software application for lung ultrasound (LUS) analysis. Physicians upload de-identified B-Mode LUS cine loops (MP4) acquired with non-linear transducers. The device uses machine learning to analyze cines for the presence or absence of pleural line motion (lung sliding or lung pulse). It automatically validates input criteria (e.g., frame rate, duration, memory size); if criteria are not met, it returns a 'Cannot Evaluate' output. The software provides a structured summary table of findings to the physician. It is intended for use in clinical settings (ICUs, EDs, clinics) by credentialed physicians (internal, emergency, critical care). The device acts as a concurrent reading aid; it does not replace clinical judgment or dictate acquisition protocols. By identifying the absence of pleural line motion, it assists in rendering potential diagnoses, potentially improving diagnostic accuracy for conditions where lung sliding is absent.
Clinical Evidence
Retrospective MRMC study with 25 board-certified physicians (Emergency, Internal, Critical Care). Dataset included 400 LUS cines (200 positive, 200 negative). Primary endpoint: difference in mean reader AUC between aided and unaided conditions. Results: Aided AUC 0.922 vs. Unaided 0.855 (difference 0.068, 95% CI: 0.036–0.100). Aided sensitivity 0.880 vs. Unaided 0.716. Aided specificity 0.841 vs. Unaided 0.820. Standalone performance on 818 cines showed 0.94 sensitivity and 0.87 specificity.
Technological Characteristics
Cloud-based web application. Machine learning-based computer vision algorithm. Inputs: MP4 B-Mode LUS cines. Compatible with non-linear array transducers (Sonosite, Philips, Butterfly, Mindray, GE). Software developed under FDA Basic Documentation Level. Cybersecurity controls include penetration testing, vulnerability scanning, and CVE analysis. No hardware components.
Indications for Use
Indicated for patients aged 18 years or older who have undergone a lung ultrasound examination. Used as a concurrent reading aid to assist physicians in identifying the absence of pleural line motion (lung sliding or lung pulse). No contraindications.
Regulatory Classification
Identification
A radiological computer-assisted diagnostic software for lesions suspicious of cancer is an image processing prescription device intended to aid in the characterization of lesions as suspicious for cancer identified on acquired medical images such as magnetic resonance, mammography, radiography, or computed tomography. The device characterizes lesions based on features or information extracted from the images and provides information about the lesion(s) to the user. Diagnostic and patient management decisions are made by the clinical user.
Special Controls
A radiological computer-assisted diagnostic (CADx) software for lesions suspicious for cancer must comply with the following special controls: 1. Design verification and validation must include: i. A detailed description of the image analysis algorithms including, but not limited to, a detailed description of the algorithm inputs and outputs, each major component or block, and algorithm limitations. ii. A detailed description of pre-specified performance testing protocols and dataset(s) used to assess whether the device will improve reader performance as intended. iii. Results from performance testing protocols that demonstrate that the device improves reader performance in the intended use population when used in accordance with the instructions for use. The performance assessment must be based on appropriate diagnostic accuracy measures (e.g., receiver operator characteristic plot, sensitivity, specificity, predictive value, and diagnostic likelihood ratio). The test dataset must contain sufficient numbers of cases from important cohorts (e.g., subsets defined by clinically relevant confounders, effect modifiers, concomitant diseases, and subsets defined by image acquisition characteristics) such that the performance estimates and confidence intervals of the device for these individual subsets can be characterized for the intended use population and imaging equipment. iv. Standalone performance testing protocols and results of the device. v. Appropriate software documentation (e.g., device hazard analysis; software requirements specification document; software design specification document; traceability analysis; description of verification and validation activities including system level test protocol, pass/fail criteria, results, and cybersecurity). 2. Labeling must include: i. A detailed description of the patient population for which the device is indicated for use. ii. A detailed description of the intended reading protocol. iii. A detailed description of the intended user and recommended user training. iv. A detailed description of the device inputs and outputs. v. A detailed description of compatible imaging hardware and imaging protocols. vi. Warnings, precautions, and limitations, including situations in which the device may fail or may not operate at its expected performance level (e.g., poor image quality or for certain subpopulations), as applicable. vii. Detailed instructions for use. viii. A detailed summary of the performance testing, including: test methods, dataset characteristics, results, and a summary of sub-analyses on case distributions stratified by relevant confounders (e.g., lesion and organ characteristics, disease stages, and imaging equipment).
*Classification.* Class II (special controls). The special controls for this device are:(1) Design verification and validation must include:
(i) A detailed description of the image analysis algorithms including, but not limited to, a detailed description of the algorithm inputs and outputs, each major component or block, and algorithm limitations.
(ii) A detailed description of pre-specified performance testing protocols and dataset(s) used to assess whether the device will improve reader performance as intended.
(iii) Results from performance testing protocols that demonstrate that the device improves reader performance in the intended use population when used in accordance with the instructions for use. The performance assessment must be based on appropriate diagnostic accuracy measures (
*e.g.,* receiver operator characteristic plot, sensitivity, specificity, predictive value, and diagnostic likelihood ratio). The test dataset must contain sufficient numbers of cases from important cohorts (*e.g.,* subsets defined by clinically relevant confounders, effect modifiers, concomitant diseases, and subsets defined by image acquisition characteristics) such that the performance estimates and confidence intervals of the device for these individual subsets can be characterized for the intended use population and imaging equipment.(iv) Standalone performance testing protocols and results of the device.
(v) Appropriate software documentation (
*e.g.,* device hazard analysis; software requirements specification document; software design specification document; traceability analysis; and description of verification and validation activities including system level test protocol, pass/fail criteria, results, and cybersecurity).(2) Labeling must include:
(i) A detailed description of the patient population for which the device is indicated for use.
(ii) A detailed description of the intended reading protocol.
(iii) A detailed description of the intended user and recommended user training.
(iv) A detailed description of the device inputs and outputs.
(v) A detailed description of compatible imaging hardware and imaging protocols.
(vi) Warnings, precautions, and limitations, including situations in which the device may fail or may not operate at its expected performance level (
*e.g.,* poor image quality or for certain subpopulations), as applicable.(vii) Detailed instructions for use.
(viii) A detailed summary of the performance testing, including: Test methods, dataset characteristics, results, and a summary of sub-analyses on case distributions stratified by relevant confounders (
*e.g.,* lesion and organ characteristics, disease stages, and imaging equipment).
{1}
FDA U.S. FOOD & DRUG
ADMINISTRATION
October 23, 2026
Deep Breathe, Inc.
Robert Arntfield
CEO and Founder
45 King St. #300
London, ON N6A1B8
Canada
Re: K261679
Trade/Device Name: BlineSlide
Regulation Number: 21 CFR 892.2060
Regulation Name: Radiological Computer-Assisted Diagnostic Software For Lesions Suspicious Of Cancer
Regulatory Class: Class II
Product Code: POK
Dated: September 17, 2026
Received: September 18, 2026
Dear Robert Arntfield:
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.
Before making any change significantly affecting the safety or effectiveness of the device, you must submit a new premarket notification in accordance with 21 CFR 807.81.
U.S. Food & Drug Administration
10903 New Hampshire Avenue
Silver Spring, MD 20993
www.fda.gov
{2}
K261679 - Robert Arntfield
Page 2
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 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.
{3}
K261679 - Robert Arntfield
Page 3
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,
Jessica Lamb, Ph.D.
Assistant Director
DHT8C: Division of Radiological
Imaging and Radiation Therapy Devices
OHT8: Office of Radiological Health
Office of Product Evaluation and Quality
Center for Devices and Radiological Health
Enclosure
{4}
| 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. | K261679 | ? |
| Please provide the device trade name(s). | | ? |
| BlineSlide | | |
| Please provide your Indications for Use below. | | ? |
| BlineSlide is intended for use as a concurrent reading aid to assist interpreting physicians, after lung ultrasound examinations of patients aged 18 years or older, by automatically identifying the absence of pleural line motion (lung sliding or lung pulse).The device provides information on abnormal findings that may be useful in rendering potential diagnoses. Patient management decisions should not be made solely on the results of the BlineSlide's analysis. | | |
| Please select the types of uses (select one or both, as applicable). | Prescription Use (21 CFR 801 Subpart D)Over-The-Counter Use (21 CFR 801 Subpart C) | ? |
| Please select the age group(s) for which the device(s) is to be used. | Neonates/Newborns (Birth to < 29 days old)Infants (29 days old to < 2 years old)Children (2 years old to < 12 years old)Adolescents (12 years old to < 22 years old)Adults (22 years old and greater) | ? |
{5}
# 510(k) Summary
K261679
## Table of Contents
| Submitter | 1 |
| --- | --- |
| Subject Device | 2 |
| Predicate Device | 2 |
| Reference Device | 3 |
| Device Description | 3 |
| Device Overview | 3 |
| Intended Patient Population | 4 |
| Intended User | 4 |
| Environment of Use | 4 |
| Compatible Ultrasounds | 4 |
| Contraindications | 4 |
| Indications for Use | 4 |
| Comparison of Technological Characteristics with the Predicate Device | 5 |
| Performance Data | 8 |
| Summary of Non-Clinical Tests | 8 |
| Verification and Validation | 8 |
| Risk Assessment | 9 |
| Cybersecurity Assessment | 9 |
| Standalone Performance Assessment | 9 |
| Summary of Clinical Tests | 11 |
| Conclusion | 14 |
## Submitter
Address:
Deep Breathe Inc.
45 King St
Ste. 300
London, ON, Canada
N6A 1B8

1
{6}
Primary Contact Person: Robert Arntfield
CEO and Founder
Deep Breathe Inc.
Email: rob@deepbreathe.ai
Phone: 5196394941
Secondary Contact Person: Nima Akhlaghi
Director, Digital Health and Imaging Center Lead
MCRA
Email: nakhlaghi@mcra.com
Phone: 202-301-4443
Date Prepared: 2026-04-17
## Subject Device
Name of Device: BlineSlide
Manufacturer: Deep Breathe Inc.
Classification Name: Radiological computer-assisted diagnostic software for lesions suspicious of cancer
Regulation Number: 21 CFR 892.2060
Regulatory Class: II
Product Code: POK
## Predicate Device
Name of Device: Sonio Suspect
Premarket Notification: K243614
Manufacturer: Sonio
Classification Name: Radiological computer-assisted diagnostic software for lesions suspicious of cancer
Regulation Number: 21 CFR 892.2060
Regulatory Class: II
Product Code: POK

2
{7}
# Reference Device
| Name of Device: | BlineSlide |
| --- | --- |
| Premarket Notification: | K252007 |
| Manufacturer: | Deep Breathe Inc. |
| Classification Name: | Medical image management and processing system |
| Regulation Number: | 21 CFR 892.2050 |
| Regulatory Class: | II |
| Product Code: | QIH |
# Device Description
## Device Overview
BlineSlide is a computer-aided diagnostic (CADx) cloud service application that assists qualified users with image-based assessment of lung ultrasound (LUS) cines acquired with the transducer placed perpendicular to the ribs, where the pleural line is the primary imaging target, from patients aged 18 years or older. It does not directly interface with ultrasound systems or receive data during the bedside examination.
BlineSlide takes as input user-uploaded de-identified B-Mode LUS cine loops (cines) in MP4 format and automatically identifies abnormal findings that may be useful in rendering potential diagnoses. Key features of the software are:
- **Pleural Line Motion Module:** an AI-assisted tool for detecting the absence of pleural line motion (lung sliding or lung pulse) in LUS cines
BlineSlide is incompatible with:
- Cines that are acquired from *Linear* array ultrasound transducers;
- Cines requiring more than 2048 megabytes of memory;
- Cines with a frame rate less than 18 frames per second;
- Cines that are less than 2600 milliseconds in duration;
- Cines that are greater than 7800 milliseconds in duration; and
- Cines in which the pleural line is not consistently within the field of view.
Each of these criteria are automatically assessed by the software. If detected, an output of *Cannot Evaluate* is returned to the user. A Cannot Evaluate output may occur even for technically adequate cines and does not necessarily indicate poor image quality. It neither indicates nor excludes the presence of pleural line motion. Guidance on how to proceed if a Cannot Evaluate output is returned is provided in the Instructions for Use.

3
{8}
BlineSlide does not perform any function that could not be accomplished by a trained user manually. It does not prescribe an acquisition protocol; cines should be acquired in accordance with standard clinical practice and applicable lung ultrasound guidelines. Patient management decisions should not be made solely on the results of BlineSlide's analysis.
Cines should be uploaded to BlineSlide as soon as practicable after acquisition and no later than 1 hour. Beyond this interval, including where connectivity does not permit upload, cines should be interpreted without device assistance. Care that is clinically indicated, including, but not limited to, intervention for suspected tension physiology, should not be withheld while awaiting a BlineSlide output.
## Intended Patient Population
Patients aged 18 years or older.
## Intended User
The intended user is a physician credentialed to use lung ultrasound clinically, such as internal medicine, emergency medicine, and critical care physicians.
## Environment of Use
The environment of use includes any setting with reliable internet connectivity where lung ultrasound exams may be performed by credentialed physicians. This includes, but is not limited to, hospital-based environments such as intensive care units and emergency departments, and outpatient clinics.
## Compatible Ultrasounds
De-identified cines imaged using non-linear array ultrasound transducers.
BlineSlide has been validated for use with cine loops acquired on the following ultrasound systems: Sonosite, Philips, Butterfly, Mindray, and GE. Performance has not been established for ultrasound systems not listed above.
## Contraindications
None.
## Indications for Use
BlineSlide is intended for use as a concurrent reading aid to assist interpreting physicians, after

4
{9}
lung ultrasound examinations of patients aged 18 years or older, by automatically identifying the absence of pleural line motion (lung sliding or lung pulse).
The device provides information on abnormal findings that may be useful in rendering potential diagnoses. Patient management decisions should not be made solely on the results of the BlineSlide’s analysis.
## Comparison of Technological Characteristics with the Predicate Device
The following table contains a comparison of the key technological features of the subject, predicate, and reference device:
| Feature | Subject Device Blineslide | Predicate Device Sonio Suspect (K243614) | Reference Device BlineSlide (K252007) | Discussion of Differences |
| --- | --- | --- | --- | --- |
| **Intended Use** | BlineSlide is intended for use as a concurrent reading aid to assist interpreting physicians, after lung ultrasound examinations of patients aged 18 years or older, by automatically identifying the absence of pleural line motion (lung sliding or lung pulse). The device provides information on abnormal findings that may be useful in rendering potential | Sonio Suspect is intended to assist interpreting physicians, during or after fetal ultrasound examinations, by automatically identifying and characterizing abnormal fetal ultrasound findings on detected views, using machine learning techniques. The device is intended for use as a concurrent reading aid on acquired images, during | BlineSlide is intended for noninvasive processing of ultrasound images to detect, measure, and calculate relevant medical parameters of structures and function of patients aged 18 years or older with suspected disease. | Equivalent to predicate. Both devices are concurrent reading aids that automatically analyze ultrasound images and provide information to assist interpreting physicians. They differ in target population: The subject device indicated for patients aged 18 years or older having undergone a lung ultrasound examination. The predicate device is intended for adult pregnant female patients |
![Logo of the University of Illinois at Chicago (US) featuring a stylized 'd' and 'h' inside a circle.]()
5
{10}
| | diagnoses. Patient management decisions should not be made solely on the results of the BlineSlide's analysis. | and/or after fetal ultrasound examinations. The device provides information on abnormal findings that may be useful in rendering potential diagnosis. Patient management decisions should not be made solely on the results of the Sonio Suspect analysis. | | having undergone a fetal ultrasound examination. Performance testing results demonstrate that these differences do not raise new questions of safety nor effectiveness. Comparisons with the reference device are included because the subject device uses the same software system. All software architecture, infrastructure, and user interface components are identical; the only modification is the AI detection feature. |
| --- | --- | --- | --- | --- |
| Image Modality | Lung ultrasound (LUS) video clips (cines) | Fetal ultrasound images | LUS cines | Equivalent to predicate. Both devices analyze standard diagnostic ultrasound data. Performance testing results demonstrate that the differences in anatomical application do not raise new questions of safety nor effectiveness. |
db
6
{11}
| | | | | Identical to reference device. |
| --- | --- | --- | --- | --- |
| **Findings Detected** | Absence or presence of pleural line motion artifacts | Absence or presence of abnormal fetal ultrasound findings | Absence or presence of B-lines artifacts | Equivalent. All devices automatically identify predefined ultrasound findings to assist physician interpretation. Performance testing results demonstrate that the differences in the specific findings detected do not raise new questions of safety nor effectiveness. |
| **Algorithm Methodology** | Computer vision Machine Learning-Based Algorithms | Computer vision Machine Learning-Based Algorithm | Computer vision Machine Learning-Based Algorithms | Identical. |
| **Platform** | Secure cloud-based web application compatible with uploaded ultrasound cines | Secure cloud-based and stand-alone software compatible with an ultrasound system | Secure cloud-based web application compatible with ultrasound cines | Equivalent to predicate. Both are cloud-based software devices that accept and analyze ultrasound inputs. Identical to reference. The devices use the same software system. |
| **Report Creation** | Yes. A table summarizing the device outputs for each uploaded cine is | No | Yes. A table summarizing the device outputs for each uploaded cine is | The subject device presents outputs in a structured summary table, |
db
7
{12}
| | displayed to the user. | | displayed to the user. | while the predicate does not include the same reporting format. This reporting feature does not alter Intended Use or device function and does not introduce new risks. Identical to reference. |
| --- | --- | --- | --- | --- |
In summary, the subject and predicate devices share the same Intended Use: concurrent reading aids that assist interpreting physicians, after ultrasound examinations, by automatically identifying abnormal findings that may be useful in rendering potential diagnoses. In both cases, patient management decisions are not to be based solely on the device's analysis. Although the subject device is indicated following lung ultrasound in patients aged 18 years or older and the predicate device is indicated following fetal ultrasound examinations of adult pregnant female patients, both devices interpret ultrasound images using machine-learning based algorithms to detect predefined abnormalities.
Performance testing conducted in accordance with established validation protocols demonstrates that the subject device performs as intended within its intended patient population, and that the differences in target patient population do not raise new questions of safety or effectiveness. The addition of structured summary reporting into the subject device does not alter Intended Use or device function and does not introduce new risks.
Accordingly, the subject device, BlineSlide, is determined to be substantially equivalent to the predicate device, Sonio Suspect.
## Performance Data
### Summary of Non-Clinical Tests
#### Verification and Validation
The software was developed and evaluated in accordance with the FDA's Basic Documentation Level for software of low risk. Verification activities included unit, integration, and system-level testing, which demonstrated that all software requirements were correctly implemented and

8
{13}
performed as intended. Validation testing was conducted on a representative system configuration and confirmed the software's performance in relevant use conditions.
## Risk Assessment
A risk analysis was conducted following the principles of ISO 14971 and applicable FDA guidance. The analysis identified potential hazards, assessed associated risks, and implemented risk control measures. All residual risks were evaluated and deemed acceptable based on the device's intended use and benefit-risk profile.
## Cybersecurity Assessment
A comprehensive cybersecurity assessment was performed in accordance with FDA's premarket cybersecurity guidance and industry best practices. The device underwent penetration testing, vulnerability scanning, and a Common Vulnerabilities and Exposures (CVE) analysis. Identified threats were reviewed and mitigated using accepted controls, and no unmitigated high-severity vulnerabilities remained at the time of release.
## Standalone Performance Assessment
The performance of the Pleural Line Motion Artifact Detection Module was successfully evaluated on a test dataset of 818 cines encompassing diverse demographic and imaging variables (Table 1). Cines were acquired from a variety of clinical sites in cities with diverse race and ethnicity populations.
| Characteristic | Distribution |
| --- | --- |
| Age (years) | 18 to 21 (3); 22 to 39 (30); 40 to 64 (116); Over 65 (182); Unknown (487) |
| Sex | Male (204); Female (120); Unknown (494) |
| BMI (kg/m²) | Under 25 (49); 25 to 30 (40); Over 30 (57); Unknown (672) |
| Acquisition Region | Anterior (380); Anterolateral (72); Posterior (9); Unknown (357) |
| Manufacturer | Sonosite (572); Philips (135); Butterfly (60); GE (26); Mindray (24); Konica Minolta (1) |
| Transducer | Phased Array (656); Curved Linear (162) |
| Image Quality | Good (532); Poor (286) |
Table 1. Standalone Dataset Characteristics.
The test data was entirely separated from that used for development. To ensure such data separation and generalizability, the data sources used in the test set were entirely different and geographically distinct from the data sources used in the development set.
db
9
{14}
Ground truth for the presence or absence of pleural line motion artifacts was determined by consensus agreement of two or more experts. In rare cases where consensus could not be reached due to poor image quality, clips were excluded from the performance assessment. In total, 36 clips met this criterion, representing only 4.2% of all cases considered. After exclusions, the final dataset comprised 202 positive class examples with Pleural Line Motion Artifacts Absent and 616 negative class examples with Pleural Line Motion Artifacts Present. Performance was assessed by measuring agreement using sensitivity and specificity as co-primary endpoints with Cannot Evaluate outputs scored as false predictions. The reliability of the module compared to reference data is summarized in Table 2 below.
| | Reliability |
| --- | --- |
| Sensitivity | 0.94 (95% CI: 0.89 – 0.96) |
| Specificity | 0.87 (95% CI: 0.84 – 0.89) |
Table 2. Blineslide's Reliability for Detecting the Absence of Pleural Line Motion Artifacts.
Module performance was also assessed across a wide range of key cohorts, including imaging characteristics (ultrasound vendor, transducer type, image quality, acquisition region), demographic variables (sex, age, body mass index), and artifact patterns (lung pulse, lung point, lung sliding). The evaluation concluded that performance was consistent across clinically meaningful cohorts.
| | | Sensitivity | | Specificity | |
| --- | --- | --- | --- | --- | --- |
| Cohort | | Sample Size | Estimate (95% CI) | Sample Size | Estimate (95% CI) |
| Image Quality | Good | 130 | 0.95 (0.89 – 0.97) | 402 | 0.89 (0.85 – 0.92) |
| | Poor | 72 | 0.92 (0.83 – 0.96) | 214 | 0.84 (0.78 – 0.88) |
| Vendor | Sonosite | 146 | 0.94 (0.89 – 0.97) | 426 | 0.85 (0.82 – 0.88) |
| | Philips | 47 | 0.96 (0.86 – 0.99) | 88 | 0.85 (0.76 – 0.91) |
| | Butterfly | 5 | 0.60 (0.23 – 0.88) | 55 | 0.93 (0.83 – 0.97) |
| | Mindray | 3 | 1.00 (0.44 – 1.00) | 21 | 1.00 (0.85 – 1.00) |
| | GE | 1 | 1.00 (0.21 – 1.00) | 25 | 1.00 (0.87 – 1.00) |
db
10
{15}
| Transducer | Phased | 144 | 0.92 (0.87 – 0.96) | 512 | 0.88 (0.85 – 0.90) |
| --- | --- | --- | --- | --- | --- |
| | Curved Linear | 58 | 0.97 (0.88 – 0.99) | 104 | 0.83 (0.74 – 0.89) |
| Acquisition Region | Anterior | 121 | 0.96 (0.91 – 0.98) | 259 | 0.88 (0.84 – 0.92) |
| | Anterolateral | 12 | 1.00 (0.76 – 1.00) | 60 | 0.85 (0.74 – 0.92) |
| | Posterior | 6 | 1.00 (0.61 – 1.00) | 3 | 0.67 (0.21 – 0.94) |
Table 3. BlineSlide's Standalone Performance Stratified Across Imaging Characteristics.
## Summary of Clinical Tests
The diagnostic accuracy of readers aided by BlineSlide's Pleural Line Motion Module was successfully evaluated against unaided reader accuracy in a retrospective multiple reader multiple case (MRMC) study. The study dataset encompassed diverse patient demographics and image characteristics, was entirely separated from the development datasets by geography, and was not used for any part of the module's development. Readers were representative of the intended user population of the device across U.S. medical practice.
The evaluation dataset comprised 400 lung ultrasound cines (200 positive, 200 negative), sourced from multiple clinical sites (58.3% U.S.), spanning multiple ultrasound vendors, transducer types, image quality levels, and patient demographic variables. Complete demographic data was not available for all patients in this dataset, with BMI being unknown in 60% of cases, age in 65% of cases, and sex in 67% of cases. The chest region scanned (anterior, anterolateral, or posterior) was unknown in 48% of cases. Ground truth was determined by consensus of two or more clinical experts. Twenty-five readers completed the study, all holding U.S. Board Certification in Emergency Medicine, Internal Medicine, or Critical Care Medicine.
The primary endpoint was the difference in mean reader AUC between aided and unaided conditions. The predefined success criterion was met. Mean reader performance under aided and unaided conditions is summarized in Table 4 and average ROC curves are presented in Figure 1. No adverse effects or patient safety events were observed.
| | AUC (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) |
| --- | --- | --- | --- |
| Aided | 0.922 (0.892 - 0.953) | 0.880 (0.856 - 0.904) | 0.841 (0.811 - 0.872) |
| Unaided | 0.855 (0.807 - 0.902) | 0.716 (0.690 - 0.742) | 0.820 (0.798 - 0.842) |
db
11
{16}
| Difference | 0.068 (0.036 - 0.100) | 0.164 (0.142 - 0.187) | 0.021 (-0.001 - 0.044) |
| --- | --- | --- | --- |
Table 4. Improvements in Reader Average Performance when Aided by BlineSlide.

Figure 1. Average Unaided (dashed) and Aided (solid) ROC Curves.
The improvement in reader specificity, 0.021 (95% CI: -0.001 – 0.044), was not statistically significant. However, reader specificity was not a primary endpoint, and the benefit-risk profile of the device remains favourable: device assistance significantly improves overall accuracy, principally through the detection of absent pleural line motion (Table 4). Absent pleural line motion is the finding readers identified least accurately unaided, and the finding whose non-recognition carries the greater clinical consequence.
BlineSlide's effect on reader diagnostic accuracy was also evaluated across key patient, imaging, and reader-level cohorts. These analyses are not powered to detect statistically significant differences, and the results are not intended to statistically support performance claims for any individual cohort. Rather, they are intended to illustrate how aided performance varied across the intended patient population and user base and should not replace clinical judgment when interpreting device output for an individual patient. Aided performance across patient and imaging characteristics is summarized in Table 5, and across reader characteristics in Table 6.
Aided AUC, sensitivity, and specificity confidence intervals overlapped or exceeded the corresponding overall study confidence intervals in all evaluable cohorts (Table 4), indicating performance consistent with the overall study results. Aided AUC and sensitivity exceeded unaided metrics in all evaluable cohorts. Aided specificity exceeded unaided specificity in most
db
12
{17}
cohorts. The exceptions were patients aged 40 to 64 years (aided 0.79, unaided 0.82), patients with a BMI under 25 (0.85, 0.89) or over 30 (0.71, 0.78), female patients (0.81, 0.81), and emergency medicine physicians (0.88, 0.89). This is consistent with the overall study specificity finding described above and does not change the overall risk/benefit assessment for the device's intended use.
No estimate is reported where a cohort contained fewer than two cines of the relevant class: all metrics for patients aged 18 to 21 years (1 cine); AUC and specificity in the posterior acquisition region (1 negative class cine); and AUC and sensitivity for GE ultrasound systems (1 positive class cine).
| Variable | Cohort | AUC | Sensitivity | Specificity |
| --- | --- | --- | --- | --- |
| Age (years) | 22 to 39 | 0.96 (0.93-0.99, 13) | 0.93 (0.85-0.97, 6) | 0.91 (0.82-0.96, 7) |
| | 40 to 64 | 0.93 (0.88-0.98, 54) | 0.90 (0.83-0.94, 39) | 0.79 (0.64-0.89, 15) |
| | Over 65 | 0.95 (0.93-0.98, 72) | 0.88 (0.83-0.92, 57) | 0.92 (0.88-0.95, 15) |
| BMI (kg/m^{2}) | Under 25 | 0.92 (0.83-1.00, 24) | 0.82 (0.69-0.90, 17) | 0.85 (0.59-0.95, 7) |
| | 25 to 30 | 0.96 (0.92-1.00, 15) | 0.90 (0.82-0.94, 10) | 0.96 (0.91-0.98, 5) |
| | Over 30 | 0.89 (0.74-1.00, 21) | 0.87 (0.79-0.92, 16) | 0.71 (0.38-0.91, 5) |
| Sex | Female | 0.93 (0.89-0.98, 38) | 0.92 (0.88-0.95, 22) | 0.81 (0.67-0.90, 16) |
| | Male | 0.94 (0.92-0.97, 95) | 0.86 (0.81-0.90, 74) | 0.90 (0.85-0.93, 21) |
| Acquisition Region | Anterior | 0.94 (0.91-0.97, 168) | 0.90 (0.87-0.92, 119) | 0.85 (0.78-0.90, 49) |
| | Anterolateral | 0.95 (0.91-0.99, 34) | 0.93 (0.89-0.95, 12) | 0.83 (0.73-0.90, 22) |
| | Posterior | Insufficient Data | 0.95 (0.94-0.96, 6) | Insufficient Data |
| Ultrasound Vendor | Butterfly | 0.94 (0.84-1.00, 29) | 0.79 (0.48-0.94, 5) | 0.94 (0.91-0.96, 24) |
| | GE | Insufficient Data | Insufficient Data | 0.95 (0.92-0.97, 20) |
| | Mindray | 0.96 (0.89-1.00, 18) | 0.91 (0.74-0.97, 3) | 0.91 (0.85-0.95, 15) |
![Logo of the University of California, Los Angeles, CA, featuring a stylized 'd' and 'b' inside a circle.]()
13
{18}
| | Philips | 0.93 (0.89-0.98, 86) | 0.90 (0.85-0.94, 47) | 0.83 (0.74-0.90, 39) |
| --- | --- | --- | --- | --- |
| | Sonosite | 0.90 (0.87-0.93, 245) | 0.87 (0.84-0.90, 144) | 0.79 (0.74-0.83, 101) |
| **Transducer Type** | Curved Linear | 0.94 (0.90-0.98, 98) | 0.92 (0.88-0.94, 58) | 0.85 (0.77-0.90, 40) |
| | Phased Array | 0.92 (0.88-0.95, 302) | 0.87 (0.83-0.89, 142) | 0.84 (0.80-0.87, 160) |
| **Image Quality** | Good | 0.94 (0.90-0.97, 253) | 0.90 (0.87-0.92, 129) | 0.87 (0.82-0.90, 124) |
| | Poor | 0.90 (0.86-0.94, 147) | 0.85 (0.79-0.89, 71) | 0.80 (0.74-0.85, 76) |
Table 5. Average reader performance when aided by BlineSlide across patient and imaging characteristics. Metrics are reported as [point estimate (95% confidence interval, sample size)]. Analyses are not powered to detect statistically significant differences and are not intended to support performance claims for any individual cohort (see accompanying text).
| Variable | Cohort | Readers | AUC | Sensitivity | Specificity |
| --- | --- | --- | --- | --- | --- |
| **Specialty** | Critical Care | 9 | 0.93 (0.90 - 0.96) | 0.90 (0.87 - 0.92) | 0.84 (0.80 - 0.87) |
| | Emergency | 7 | 0.94 (0.91 - 0.97) | 0.89 (0.86 - 0.91) | 0.88 (0.84 - 0.90) |
| | Internal | 9 | 0.90 (0.82 - 0.98) | 0.85 (0.82 - 0.88) | 0.82 (0.78 - 0.85) |
| **Experience** | ≤ 10 years | 14 | 0.94 (0.92 - 0.96) | 0.91 (0.88 - 0.93) | 0.87 (0.83 - 0.90) |
| | >10 years | 11 | 0.90 (0.84 - 0.96) | 0.84 (0.81 - 0.87) | 0.81 (0.77 - 0.84) |
Table 6. Average reader performance when aided by BlineSlide across reader characteristics. Metrics are reported as [point estimate (95% confidence interval)]. Analyses are not powered to detect statistically significant differences and are not intended to support performance claims for any individual cohort (see accompanying text).
## Conclusion
BlineSlide is substantially equivalent in intended use, design, principles of operation, technological characteristics, and safety features to the predicate device.
The special controls for 21 CFR 892.2060 are satisfied by demonstrating the effectiveness of the device in both standalone and clinical testing, showing the superiority of aided versus

14
{19}
unaided reader performance in the clinical testing, providing appropriate software and cybersecurity documentation, and communicating appropriate information in the labeling.
BlineSlide is as safe and as effective as the predicate device when used as intended.

15
Predicate graph will load when search results are available.
Embedding visualization will load when search results are available.
PDF viewer will load when search results are available.
Loading panels...
Select an item from Submissions
Click any panel, subpart, regulation, product code, or device to see details here.
Section Matches
Results will appear here.
Product Code Matches
Results will appear here.
Special Control Matches
Results will appear here.
Loading collections...
Loading
My Alerts
You will receive email notifications based on the filters and frequency you set for each alert.
Sort by:
Guidance alerts
New and changed FDA guidance documents, emailed as a daily digest.
Create Alert
Search Filters
Agent Token
Create a read-only bearer token for Claude, ChatGPT, or other agents that can call HTTP APIs.
Copy this now. It will not be shown again.
Connected apps
Apps you authorized through browser sign-in. Disconnecting revokes their access immediately.
Learn the FDA Browser
Two short videos show you everything — or skip straight to the written tutorial if you'd rather read. You can reopen this any time from the ? button in the top bar.
Part 1 — Search, results, and everyday workflows 16 min
Part 2 — Embeddings: the galaxy map 3 min
1. Search: exact and fuzzy
Type a phrase like "coronary artery calcification" into the search box. You get two kinds of results. Exact results match the literal phrase — prefix searches work ("coronary artery calcificati") but suffix searches do not. Fuzzy results match on the meaning and intent of your phrase rather than the exact words, and are sorted by relevance score. Hover over the Exact or Fuzzy badge on any row to see exactly why it matched.
Use the checkboxes above the results to narrow: SaMD keeps only Software as a Medical Device, AI / ML keeps only devices with AI.
Exact vs. fuzzy search: what's the difference?
Exact matches on the literal phrase (prefix search works, suffix does not). Fuzzy matches on the meaning and intent of the phrase rather than the exact words. Hover over the badge on any row to see why it matched.
You search "coronary artery calcification" and want only software devices with AI. What two filters do you apply?
Narrow by SaMD (Software as a Medical Device), then narrow by AI/ML (devices with AI).
2. The results table
Scroll right in the results table. The intended use is extracted for you — no need to open the PDF. The device story gives a high-level snapshot of what the device does and how it's used. The AI Performance sub-table shows each output name, acceptance criteria, observed values, and development/test dataset descriptions — the same format Innolitics uses for regulatory strategy outputs, and the fastest high-level fingerprint of an AI device. It is AI-generated but has been very reliable in practice.
Where do you find a device's intended use without opening the PDF?
Scroll right in the search results table. The intended use column is extracted for you; no need to dig into the 510(k) summary PDF.
What does the AI Performance sub-table show, and why is it useful?
Output name, acceptance criteria, observed values, development dataset description, and test dataset description. It's the same format we use for regulatory strategy output and Fast 510(k) input, and the fastest high-level fingerprint of an AI device. AI-generated but reliable in practice.
3. Judging fuzzy relevance
Fuzzy results trail off in relevance as you scroll. Use three signals to decide how far down to go: the fuzzy badge explanations, the intended use column, and whether your target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, you're past the relevant zone. A top hit with a low score (~0.4) and a stretched explanation is a hint the closest predicates are far away — the project may be headed for De Novo. Note the fuzzy search is a pattern match: it doesn't handle negation ("not") well, and hardware devices can appear — filter by SaMD/AI ML to cut them.
How do you judge how far down fuzzy search results to go?
Use the relevancy signals: the fuzzy badge explanations, the intended use column, and whether the target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, results are trailing off in relevancy.
4. Device detail page: chat and citations
Click a device name to open its detail page: device facts on the left, a chat window on the right. Ask something like "Describe the training data". The answer carries little citation bubbles — click one to jump to the highlighted passage in the source PDF, so you can verify every AI answer against the document. There's also a Download PDF button for sharing.
How do you verify an AI chat answer on the device detail page?
Click the citation bubbles to jump to the relevant highlight in the source document.
Reading rule for every project: how many summaries do you read in full?
At least the three most relevant 510(k) or De Novo summaries, in full. After that, use targeted chat questions to confirm your memory quickly. The tool supports this professional habit — it doesn't replace it.
5. Side-by-side comparison
Select multiple rows in the results table (aim for under ~10), then open the PDF Viewer tab. Ask one question — it goes to all selected devices in parallel, each with citations. This is the fastest way to compare and contrast devices: training data, PCCP scope, how they handled adding new scanners, and so on.
What does the side-by-side PDF viewer mode do?
Select multiple devices, open the PDF viewer tab, and ask one question (e.g., "Describe the training data"). It queries all selected devices simultaneously with citations, so you can compare and contrast quickly.
6. Collections
With rows selected, go to the Collections tab and create a labeled collection (e.g., "Cobb Angle Project"). Reload that selection any time — before a client call, pull up the collection and ask questions across all of its devices at once.
How do you save a set of selected devices for later use?
Select the rows, go to the Collections tab, and create a labeled collection (e.g., "Cobb Angle Project"). You can reload the selection anytime and carry it into the PDF viewer and other tabs that support selections.
7. Product codes and the regulations tree
Click a product code in the results to jump to it in the regulations tree — identification text, sibling product codes, and devices you can open in a PDF viewer on the right. Click a regulation number to see its identification, special controls, and related product codes. You can also search by product code or regulation number at the top of the tree. Always read the special controls if any exist for your device — it broadens your search and sharpens pre-kickoff research.
What can you do from the regulations tree view?
Browse product codes and regulation numbers, read the identification text and special controls, browse sibling product codes, open device PDFs on the right, and search by product code or regulation number at the top of the tree.
8. Chart view
Click Show Chart and segment by regulation number (or product code) to see which regulations dominate your result set. Clicking a regulation takes you into the regulations tree. Great for spotting that most matches are, say, hardware laparoscopic devices — a cue to go back and filter.
How do you see which regulations dominate a search result set?
Click "Show Chart" and segment by Regulation Number. Clicking a regulation takes you to the regulations tree.
9. The predicate graph
Open the Predicates tab for a family-tree view of predicate relationships. Click a node to trace its parents and children; selections from search carry over pre-selected. Commonly predicated devices are worth reading — a lot of people predicated them for a reason. The visual lineage is also handy on client calls, e.g. to show how a predicate family evolved and justify why your predicate still holds.
In the predicate graph, why are commonly predicated devices worth reading?
A lot of people predicated them for a reason. Clicking a node traces parents and children, and selections from search carry over pre-selected.
10. Embeddings: the galaxy map
The Embeddings tab plots every matching document in a 2-D "galaxy map" where semantically similar devices cluster together. Hover or click clusters to explore, and let AI label the clusters for you. Embeddings beat product codes for grouping: two devices can carry different product codes (LLZ vs. QIH) yet do the same thing — the embedding captures the meaning of the intended use and device story. This is also exactly how retrieval-augmented generation (RAG) works under the hood, and it makes a great visual on client calls.
Try it yourself
Head to the search page and work through a few of these AI/ML fuzzy searches to build intuition: perivascular fat on CT · aortic valve calcification opportunistic screening on noncontrast CT · breast cancer prediction on digital pathology slides · autism detection · gestational age prediction · a hearing aid that can also detect a pulse · foundation model based analysis of ECG · large language models · penetration test. Watch how the relevance scores, intended use, and AI Performance tables tell you when results stop being meaningful.