Saige-Density is a software application intended for use with compatible full-field digital mammography (FFDM) and digital breast tomosynthesis (DBT) systems. Saige-Density provides an ACR BI-RADS Atlas 5th Edition breast density category to aid interpreting physicians in the assessment of breast tissue composition. Saige-Density produces adjunctive information. It is not a diagnostic aid.

Device Description

Saige-Density is Software as a Medical Device that processes screening and diagnostic digital mammograms using deep learning techniques and generates outputs that serve as an aid for interpreting radiologists in assessing breast density. The software takes as input a single x-ray mammogram study and processes all acceptable 2D image DICOM files (FFDM and/or 2D synthetics) and generates a single study-level breast density category. Two DICOM files are outputted as a result: 1) a structured report (SR) DICOM object containing the case-level breast density category and 2) a secondary capture (SC) DICOM object containing a summary report with the study-level density category. Both output files contain the same breast density category ranging from "A" through "D" following Breast Imaging Reporting and Data System (BI-RADS) 5th Edition reporting guidelines. The SC report and/or the SR file may be viewed on a mammography viewing workstation.

AI/ML Overview

Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:

Acceptance Criteria and Reported Device Performance

The acceptance criteria are implied by the reported performance metrics of the Saige-Density device. The primary objective of the standalone performance testing was to quantify the accuracy of Saige-Density's density category outputs. The reported performance is the accuracy of the device in classifying breast density into four categories (A, B, C, or D) and two categories (nondense: A, B; dense: C, D) compared to a consensus ground truth.

Acceptance Criteria	Reported Device Performance
Accuracy (Four-class categorization: A, B, C, D) vs. Ground Truth	81.28% (95% CI: 78.42, 83.84)
Accuracy (Two-class categorization: Nondense, Dense) vs. Ground Truth	Implicitly represented by the confusion matrix, not a single percentage explicitly stated for this metric. - Nondense correctly classified: 87.8% - Dense correctly classified: 95.2%

Study Details

Sample size used for the test set and the data provenance:
- Sample Size: A total of 796 mammogram cases (representing 6,170 images) were retrospectively collected for the standalone performance testing.
- Data Provenance: The data was collected from five breast imaging centers in the United States. The collection sites selected for the pivotal study did not overlap with those used previously to collect data for training or testing the Saige-Density AI algorithm.
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts: Five expert radiologists were used to establish the ground truth.
- Qualifications of Experts: The text refers to them as "expert radiologists," implying they are qualified to interpret mammograms, but specific details about their experience (e.g., years of experience) are not provided.
Adjudication method for the test set:
- Adjudication Method: Ground truth for each case was established as the consensus of the five expert radiologists' breast density categories on the same set of cases, and calculated as the median of the reported categories for each case. This suggests a form of consensus-based adjudication, specifically using the median.
If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:
- The provided text does not indicate that an MRMC comparative effectiveness study was conducted to evaluate human readers' improvement with AI assistance. The performance testing described is "Standalone Performance Testing," focusing on the algorithm's performance only.
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, a standalone performance study was explicitly conducted and detailed: "Standalone Performance Testing: A multi-site retrospective study was conducted to evaluate the standalone performance of Saige-Density on DBT and FFDM mammograms."
The type of ground truth used:
- The type of ground truth used was expert consensus of five expert radiologists, based on ACR BI-RADS 5th Edition guidelines.
The sample size for the training set:
- The exact sample size for the training set is not explicitly stated. However, the text mentions that the training data consisted of "four datasets across various geographic locations within the US."
How the ground truth for the training set was established:
- The text does not explicitly describe how the ground truth for the training set was established. It only states that the data used for training the algorithm was distinct from the test set and came from "four datasets across various geographic locations within the US, including racially diverse regions such as New York City and Los Angeles."

Summary

{0}------------------------------------------------

December 16, 2022

Image /page/0/Picture/1 description: The image shows the logo of the U.S. Food and Drug Administration (FDA). On the left is the Department of Health & Human Services logo. To the right of that is the FDA logo, which consists of a blue square with the letters "FDA" in white. To the right of the blue square is the text "U.S. FOOD & DRUG ADMINISTRATION" in blue.

DeepHealth, Inc. % B. Nathan Hunt VP, Quality Assurance and Regulatory Affairs 1000 Massachusetts Avenue CAMBRIDGE MA 01238

Re: K222275

Trade/Device Name: Saige-Density Regulation Number: 21 CFR 892.2050 Regulation Name: Medical image management and processing system Regulatory Class: Class II Product Code: QIH Dated: November 7, 2022 Received: November 7, 2022

Dear B. Nathan Hunt:

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 (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 located 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.

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

{1}------------------------------------------------

801); medical device reporting of medical device-related adverse events) (21 CFR 803) for devices or postmarketing safety reporting (21 CFR 4, Subpart B) for combination products (see https://www.fda.gov/combination-products/guidance-regulatory-information/postmarketing-safety-reportingcombination-products); good manufacturing practice requirements as set forth in the quality systems (QS) regulation (21 CFR Part 820) for devices or current good manufacturing practices (21 CFR 4. Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR 1000-1050.

Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR Part 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-device-safety/medical-device-reportingmdr-how-report-medical-device-problems.

For comprehensive regulatory information about mediation-emitting products, including information about labeling regulations, please see Device Advice (https://www.fda.gov/medicaldevices/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-device-advice-comprehensive-regulatoryassistance/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,

Yanna S. Kang -S

Yanna Kang, Ph.D. Assistant Director Mammography and Ultrasound Team 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

{2}------------------------------------------------

Indications for Use

510(k) Number (if known) K22275

Device Name Saige-Density

Indications for Use (Describe)

Type of Use (Select one or both, as applicable)
X Prescription Use (Part 21 CFR 801 Subpart D)	Over-The-Counter Use (21 CFR 801 Subpart C)

CONTINUE ON A SEPARATE PAGE IF NEEDED.

This section applies only to requirements of the Paperwork Reduction Act of 1995.

DO NOT SEND YOUR COMPLETED FORM TO THE PRA STAFF EMAIL ADDRESS BELOW.

The burden time for this collection of information is estimated to average 79 hours per response, including the time to review instructions, search existing data sources, gather and maintain the data needed and complete and review the collection of information. Send comments regarding this burden estimate or any other aspect of this information collection, including suggestions for reducing this burden, to:

Department of Health and Human Services Food and Drug Administration Office of Chief Information Officer Paperwork Reduction Act (PRA) Staff PRAStaff(@fda.hhs.gov

"An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB number."

{3}------------------------------------------------

Image /page/3/Picture/0 description: The image contains the logo for DeepHealth. On the left side of the logo is a stack of three squares, each slightly offset from the one below it. To the right of the squares is the word "DeepHealth", with "Deep" in a dark red color and "Health" in gray.

1000 Massachusetts Ave Cambridge. MA 02138 Phone: 424-832-1480 www.deep.health

510(k) Summary DeepHealth, Inc. Saige-Density (K222275)

In accordance with 21 CFR 807.92 the following summary of information is provided, on this date, December 15, 2022:

1. 510(k) SUBMITTER

DeepHealth, Inc. 1000 Massachusetts Avenue Cambridge, MA 02138 Tel: 424-832-1480

Contact Person:

B. Nathan Hunt Vice President, Quality Assurance and Regulatory Affairs DeepHealth, Inc. 1000 Massachusetts Avenue Cambridge, MA 02138 Tel: 424-832-1480 nhunt@deep.health

Date Prepared:

December 15, 2022

2. DEVICE

Trade Name of Device: Saige-Density

Common or Usual Name: Medical Image Software

Regulation Name and Number: Medical Image Management and Processing System (21 CFR 892.2050)

Regulation Class: II

Product Code: QIH

3. PREDICATE DEVICE

Trade Name: Densitas densityai™

Common Name or Usual Name: Medical Image Software

Regulation Name and Number: Picture Archiving and Communication System (21 CFR 892.2050)

Regulation Class: II

Product Code: LLZ

510(K) No.: K192973

{4}------------------------------------------------

Image /page/4/Picture/0 description: The image contains the logo for DeepHealth. On the left side of the logo, there is a stack of three squares, each slightly offset from the one below it. To the right of the squares, the word "DeepHealth" is written in a combination of red and gray letters. The word "Deep" is in red, while the word "Health" is in gray.

1000 Massachusetts Ave Cambridge. MA 02138 Phone: 424-832-1480 www.deep.health

4. DEVICE DESCRIPTION

Saige-Density is Software as a Medical Device that processes screening and diagnostic digital mammograms using deep learning techniques and generates outputs that serve as an aid for interpreting radiologists in assessing breast density. The software takes as input a single x-ray mammogram study and processes all acceptable 2D image DICOM files (FFDM and/or 2D synthetics) and generates a single study-level breast density category. Two DICOM files are outputted as a result: 1) a structured report (SR) DICOM object containing the case-level breast density category and 2) a secondary capture (SC) DICOM object containing a summary report with the study-level density category. Both output files contain the same breast density category ranging from "A" through "D" following Breast Imaging Reporting and Data System (BI-RADS) 5ª Edition reporting guidelines. The SC report and/or the SR file may be viewed on a mammography viewing workstation.

5. INDICATIONS FOR USE

Saige-Density is intended for use with compatible full-field digital mammography (FFDM) and digital breast tomosynthesis (DBT) systems. Saige-Density provides an ACR BI-RADS Atlas 5th Edition breast density category to aid interpreting physicians in the assessment of breast tissue composition. Saige-Density produces adjunctive information. It is not a diagnostic aid.

Intended User Population

The intended users of Saige-Density are interpreting physicians qualified to read mammography exams.

Intended Patient Populations

The device is intended to be used on female patients thirty-five (35) years of age or older undergoing mammography.

Warnings and Precautions

Saige-Density is an adjunct tool and is not intended to replace a physician's own review of a mammogram. Decisions should not be made solely based on analysis by Saige-Density.

6. PREDICATE DEVICE COMPARISON

Saige-Density and the predicate device have similar indications for use, patient population, technical characteristics, and principles of operation. The differences between Saige-Density and the predicate device do not alter the suitability of the subject device for its intended use, and do not raise different questions of safety or effectiveness.

The devices are intended to be used by physicians to aid in the assessment of breast density for mammograms. The devices are not intended to be used as a replacement of a physician's own clinical judgment.

The design of Saige-Density is similar to that of the predicate device. Both devices are compatible with FFDM and DBT mammograms, utilize deep learning to produce an equivalent main output of a patient-level breast density category, and are not diagnostic aids that are intended to be used by physicians interpreting either screening or diagnostic mammograms. As both devices use

{5}------------------------------------------------

Image /page/5/Picture/0 description: The image shows the logo for DeepHealth. The logo consists of a stack of three squares on the left, followed by the word "DeepHealth" in a combination of red and gray. The word "Deep" is in red, while the word "Health" is in gray.

1000 Massachusetts Ave Cambridge. MA 02138 Phone: 424-832-1480

proprietary algorithms, there are assumed differences in the algorithmic components, as well as minor differences in the specific formats of the outputs provided to users.

Non-clinical and clinical testing has been completed ensuring that the differences do not affect the safety and effectiveness of the proposed subject device.

7. PERFORMANCE DATA

Saige-Density is a software device and has been determined to be of Moderate Level of Concern. Verification testing included software unit testing, software integration testing, and system testing. Testing confirmed that the software, as designed and implemented, satisfies the software requirements.

Validation of the software was performed using a retrospective study as described below. The data used in the validation testing was obtained from different clinical sites than those used to develop the Saige-Density algorithm. DeepHealth ensured that there was no overlap between the data used to train and test the Saige-Density algorithm. The data used to train the Saige-Density algorithm consisted of four datasets across various geographic locations within the US, including racially diverse regions such as New York City and Los Angeles.

Standalone Performance Testing:

A multi-site retrospective study was conducted to evaluate the standalone performance of Saige-Density on DBT and FFDM mammograms. The primary objective was to quantify the accuracy of Saige-Density's density category outputs with respect to a consensus of five expert radiologists. The statistical analysis was performed by an independent biostatistician.

A total of 796 mammogram cases, representing 6,170 images, were retrospectively collected from five breast imaging centers in the United States. Importantly, the collection sites selected for the pivotal study did not overlap with those used previously to collect data for training or testing the Saige-Density AI algorithm. The cases were gathered from unique female patients 35 years of age or older and included different modalities (DBT and FFDM), manufacturers (Hologic and GE), and exam types (screening and diagnostic). Each mammogram included in the study had one ground truth status for density: A (Fatty), B (Scattered Fibroglandular), C (Heterogeneously Dense), or D (Extremely Dense) as defined by ACR's BI-RADS 5th Edition guidelines. Ground truth was established for each case as the consensus of five expert radiologists' breast density categories on the same set of cases, and calculated as the median of the reported categories for each case.

Saige-Density's accuracy was computed as the percentage of cases that Saige-Density assigned the same density category to as ground truth. Saige-Density's accuracy on four-class categorization (density categories A, B, C, or D) was 81.28% (95% CI: 78.42, 83.84). Results for four-class and two-class (nondense: A, B; dense: C,D) categorization with respect to ground truth are summarized in Figures 1 and 2, respectively.

{6}------------------------------------------------

Image /page/6/Picture/0 description: The image contains the logo for DeepHealth. On the left side of the logo, there is a stack of three squares, each slightly offset from the one below it. The squares are a dark red color. To the right of the squares, the word "DeepHealth" is written in a sans-serif font. The word "Deep" is dark red, while the word "Health" is gray.

1000 Massachusetts Ave. Cambridge, MA 02138 Phone: 424-832-1480 www.deep.health

Image /page/6/Figure/2 description: The image shows a confusion matrix with four categories labeled A, B, C, and D on both axes. The matrix displays the counts and percentages of predicted versus actual classifications. For example, category A has 62 true positives (77.5%) and 18 false negatives (22.5%). Category D has 74 true positives (97.4%) and 1 false negative (1.3%).

Saige-Density

Image /page/6/Figure/4 description: The image is a title for a figure. The title reads "Figure 1. Four-class confusion matrix for all pivotal cases." The title is written in a clear, sans-serif font and is left-aligned. The text suggests that the figure is a confusion matrix, which is a table used to evaluate the performance of a classification model.

Image /page/6/Figure/5 description: The image shows a confusion matrix with two classes: "nondense" and "dense". The matrix displays the counts and percentages for each combination of predicted and actual classes. For the "nondense" class, 351 instances are correctly classified (87.8%), while 49 instances are misclassified as "dense" (12.2%). For the "dense" class, 377 instances are correctly classified (95.2%), while 19 instances are misclassified as "nondense" (4.8%).

Saige-Density

Figure 2. Two-class confusion matrix for all pivotal cases.

{7}------------------------------------------------

Image /page/7/Picture/0 description: The image shows the logo for DeepHealth. The logo consists of a stack of three squares on the left, followed by the word "DeepHealth" in a combination of red and gray. The word "Deep" is in red, while the word "Health" is in gray.

1000 Massachusetts Ave. Cambridge, MA 02138 Phone: 424-832-1480 www.deep.health

8. CONCLUSION

The non-clinical and clinical testing conducted to support this submission confirm that Saige-Density is safe and effective. The minor differences, including technological differences, between Saige-Density and the predicate do not alter the intended use of the device and do not affect its safety and effectiveness when used as labeled. Therefore, the information presented in this 510(k) submission demonstrates that Saige-Density is substantially equivalent to the predicate device.

Regulation Number and Section

§ 892.2050 Medical image management and processing system.

(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).