(177 days)
TechCare Trauma is intended to analyze 2D X ray radiographs using techniques to aid in the detection, localization, and characterization of fractures and/or elbow joint effusion during the review of commonly acquired radiographs of: Ankle, Foot, Knee, Leg (includes Tibia/Fibula), Femur, Wrist, Hand/Finger, Elbow, Forearm, Arm (includes Humerus), Shoulder, Clavicle, Pelvis, Hip, Thorax (includes ribs).
TechCare Trauma can provide results for fracture in neonates and infants (from birth to less than 2 years), children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).
TechCare Trauma can provide results for elbow joint effusions in children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).
The intended users of TechCare Trauma are clinicians with the authority to diagnose fractures and/or elbow joint effusions in various settings including primary care (e. g., family practice, internal medicine), emergency medicine, urgent care, and specialty care (e. g. orthopedics), as well as radiologists who review radiographs across settings.
TechCare Trauma results are not intended to be used on a stand-alone basis for clinical decision-making. Primary diagnostic and patient management decisions are made by the clinical user.
The TechCare Trauma device is a software as Medical Device (SaMD). More specifically it is defined as a "radiological computer assisted detection and diagnostic software for suspected fractures".
As a CADe/x software, TechCare Trauma is an image processing device intended to aid in the detection and localization of fractures and elbow joint effusions on acquired medical images (2D X-ray radiographs).
TechCare Trauma uses an artificial intelligence algorithm to analyze acquired medical images (2D X-ray radiographs) for features suggestive of fractures and elbow joint effusions.
TechCare Trauma can provide results for fractures in neonates and infants (from birth to less than 2 years), children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over) regardless of their condition.
TechCare Trauma can provide results for elbow joint effusions in children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).The device detects and identifies fractures and elbow joint effusions based on a visual model's analysis of images and provides information about the presence and location of these prespecified findings to the user.
It relies solely on images provided by DICOM sources. Once integrated into existing networks, TechCare Trauma automatically receives and processes these images without any manual intervention. The processed results, which consist of one or more images derived from the original inputs, are then sent to specified DICOM destinations. This ensures that the results can be seamlessly viewed on any compatible DICOM viewer, allowing smooth into medical imaging workflows.
TechCare Trauma can be deployed on-premises or on cloud and be connected to multiple DICOM sources / destinations (including but not limited to DICOM storage platform, PACS, VNA and radiological equipment, such as X-ray systems), ensuring easy integration into existing clinical workflows.
Here's a detailed breakdown of the acceptance criteria and study findings for the TechCare Trauma device, based on the provided text:
Acceptance Criteria and Device Performance
The acceptance criteria for the TechCare Trauma device appear to be based on achieving high diagnostic accuracy, specifically measured by the Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC) curve for both standalone performance and multi-reader multi-case (MRMC) comparative studies. The study demonstrated successful performance against these implied criteria.
Table of Acceptance Criteria and Reported Device Performance
| Metric | Acceptance Criteria (Implied/Study Goal) | Reported Device Performance (Standalone) | Reported Device Performance (MRMC with AI vs. without AI) |
|---|---|---|---|
| Standalone Performance (Image-level ROC-AUC) | High accuracy (specific threshold not explicitly stated but implied by achievement across all categories) | Fracture - Adult: 0.962 [0.957 - 0.967] Fracture - Pediatric: 0.962 [0.955 - 0.969] EJE - Adult: 0.965 [0.936 - 0.986] EJE - Pediatric: 0.976 [0.963 - 0.986] (Further detailed by anatomical regions, age, gender, image view, and imaging hardware manufacturers, all showing high AUCs.) | Not applicable (standalone algorithm only) |
| Reader Performance (MRMC ROC-AUC) | Superior to unaided reader performance (statistically significant improvement) | Not applicable (human reader performance) | Adult Fracture: Improved from 0.865 to 0.955 (Δ 0.090, p < 0.001) Adult EJE: Improved from 0.851 to 0.914 (Δ 0.064, p < 0.001) Pediatric Fracture: Improved from 0.857 to 0.931 (Δ 0.074, p < 0.001) Pediatric EJE: Improved from 0.877 to 0.941 (Δ 0.063, p = 0.002) |
| Reader Performance (MRMC Sensitivity) | Increased Sensitivity with AI aid | Not applicable (human reader performance) | Adult Fracture: Increased by 21.8% (from 0.807 to 0.983) Adult EJE: Increased by 12.7% (from 0.872 to 0.983) Pediatric Fracture: Increased by 19.9% (from 0.804 to 0.964) Pediatric EJE: Increased by 18.2% (from 0.825 to 0.975) |
| Reader Performance (MRMC Specificity) | Maintained or increased Specificity with AI aid | Not applicable (human reader performance) | Adult Fracture: Increased by 1.47% (from 0.815 to 0.827) Adult EJE: Increased by 1.08% (from 0.738 to 0.746) Pediatric Fracture: Remained the same (0.797) Pediatric EJE: Increased by 1.43% (from 0.839 to 0.851) |
Study Details
2. Sample sizes used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
Standalone Performance Test Set:
- Fracture Detection: 4109 radiographs of US adult patients and 2872 radiographs of US pediatric patients.
- EJE Detection: 280 radiographs of US adult patients and 483 radiographs of US pediatric patients.
- Data Provenance: Retrospective, obtained from various states in the US (at least 4) and various imaging hardware manufacturers (at least 14).
MRMC Comparative Effectiveness Study Test Set:
- Adult US population for fracture detection: 304 radiological cases
- Pediatric US population for fracture detection: 256 radiological cases
- Adult US population for EJE detection: 109 radiological cases
- Pediatric US population for EJE detection: 100 radiological cases
- Data Provenance: Retrospective, external multicenter anonymized datasets obtained from sites that were different from the training data sites, ensuring independence. All data from US patients.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
Standalone Performance Test Set:
- Number of Experts: Three American Board of Radiology (ABR)-certified radiologists for both adult and pediatric cases.
- Qualifications: Minimum of 5 years of experience since ABR certification. Pediatric cases were annotated by pediatric radiologists, and adult cases by musculoskeletal (MSK) radiologists.
MRMC Comparative Effectiveness Study Test Set:
- Number of Experts: Three ABR-certified radiologists.
- Qualifications: At least five years of experience. Pediatric cases were annotated by a panel of three pediatric radiologists, while adult cases were reviewed by a panel of three MSK radiologists.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
Both Standalone and MRMC Studies:
- Adjudication Method: Two radiologists independently assessed each case. For cases with disagreement between the first two, a third radiologist independently reviewed the case. The final reference standard (ground truth) was determined by majority consensus (referred to as "2+1" if two agree, or "3+0" if all three agree after subsequent review).
5. 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
-
Yes, a MRMC comparative effectiveness study was done.
-
Effect Size of Improvement (AI vs. without AI assistance):
- Adult Fracture (ROC AUC delta): 0.090 (from 0.865 to 0.955)
- Adult EJE (ROC AUC delta): 0.064 (from 0.851 to 0.914)
- Pediatric Fracture (ROC AUC delta): 0.074 (from 0.857 to 0.931)
- Pediatric EJE (ROC AUC delta): 0.063 (from 0.877 to 0.941)
Additionally, significant improvements in sensitivity were observed:
- Adult Fracture Sensitivity: +21.8%
- Adult EJE Sensitivity: +12.7%
- Pediatric Fracture Sensitivity: +19.9%
- Pediatric EJE Sensitivity: +18.2%
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Yes, a standalone performance study was done.
- Fracture Detection ROC-AUC: 0.962 (Adults) and 0.962 (Pediatrics)
- EJE Detection ROC-AUC: 0.965 (Adults) and 0.976 (Pediatrics)
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
- The ground truth for both standalone and MRMC studies was established by expert consensus of ABR-certified radiologists.
8. The sample size for the training set
- Training Set Sample Size: 95,266 images.
9. How the ground truth for the training set was established
- The document does not explicitly describe how the ground truth for the training set was established. However, given the detailed methodology for the test set ground truth, it is highly probable that a similar expert-driven annotation process (potentially internal and/or external) was followed for the training data as well. The text states the training was performed "from various manufacturers," suggesting a diverse dataset that would necessitate robust ground truthing.
{0}------------------------------------------------
January 17, 2025
Image /page/0/Picture/1 description: The image contains the logo of the U.S. Food and Drug Administration (FDA). On the left, there is a seal with an eagle and text around it. To the right of the seal, there is the FDA logo in blue, with the words "U.S. FOOD & DRUG" stacked on top of the word "ADMINISTRATION". The logo is simple and professional, and it is easily recognizable.
Milvue % John Smith Partner Hogan Lovells US LLP Columbia Square 555 Thirteenth Street NW Washington, DC 20004
Re: K242171
Trade/Device Name: TechCare Trauma Regulation Number: 21 CFR 892.2090 Regulation Name: Radiological computer assisted detection and diagnosis software Regulatory Class: Class II Product Code: QBS Dated: July 24, 2024 Received: December 16, 2024
Dear John Smith:
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 (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.
{1}------------------------------------------------
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 System (QS) regulation (21 CFR Part 820), which includes, but is not limited to, 21 CFR 820.30, Design controls; 21 CFR 820.90, Nonconforming product; and 21 CFR 820.100, Corrective and preventive action. Please note that regardless of whether a change requires premarket review. the OS regulation requires device manufacturers to review and approve changes to device design and production (21 CFR 820.30 and 21 CFR 820.70) and document changes and approvals in the device master record (21 CFR 820.181).
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 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-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 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 Re"). 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-advicecomprehensive-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-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
{2}------------------------------------------------
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,
Samul for
Jessica Lamb, Ph.D. Assistant Director Imaging Software Team DHT8B: Division of Radiological 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
510(k) Number (if known) K242171
Device Name TechCare Trauma
Indications for Use (Describe)
TechCare Trauma is intended to analyze 2D X ray radiographs using techniques to aid in the detection, localization, and characterization of fractures and/or elbow joint effusion during the review of commonly acquired radiographs of:
| Study Type(Anatomic Area ofInterest) | RadiographicView(s) Supported | Findings | Patient Population |
|---|---|---|---|
| Ankle | Frontal, Lateral,Oblique | Fracture | Adults, Infants, Children,Adolescents |
| Foot | Frontal, Lateral,Oblique | Fracture | Adults, Infants, Children,Adolescents |
| Knee | Frontal, Lateral,Oblique | Fracture | Adults, Infants, Children,Adolescents |
| Leg (includesTibia/Fibula) | Frontal, Lateral | Fracture | Adults, Infants, Children,Adolescents |
| Femur | Frontal, Lateral | Fracture | Adults |
| Wrist | Frontal, Lateral,Oblique | Fracture | Adults, Children, Adolescents |
| Hand/Finger | Frontal, Lateral,Oblique | Fracture | Adults, Infants, Children,Adolescents |
| Elbow | Lateral | Elbow JointEffusion | Adults, Children, Adolescents |
| Forearm | Frontal, Lateral | Fracture |
Table 1 : Intended Use Overview
{4}------------------------------------------------
| Study Type(Anatomic Area ofInterest) | RadiographicView(s) Supported | Findings | Patient Population |
|---|---|---|---|
| Lateral | Elbow JointEffusion | Adults, Children, Adolescents | |
| Arm (includesHumerus) | Frontal, Lateral | Fracture | Adults, Neonates, Infants, Children,Adolescents |
| Lateral | Elbow JointEffusion | Adults, Children, Adolescents | |
| Shoulder | Frontal, Lateral | Fracture | Adults, Infants, Children,Adolescents |
| Clavicle | Frontal | Fracture | Adults, Children, Adolescents,Neonates, Infants |
| Pelvis | Frontal | Fracture | Adults, Infants, Children,Adolescents |
| Hip | Frontal, Lateral | Fracture | Adults, Infants, Children,Adolescents |
| Thorax (includes ribs) | Frontal, Lateral, Ribsseries | Fracture | Adults |
TechCare Trauma can provide results for fracture in neonates and infants (from birth to less than 2 years), children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).
TechCare Trauma can provide results for elbow joint effusions in children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).
The intended users of TechCare Trauma are clinicians with the authority to diagnose fractures and/or elbow joint effusions in various settings including primary care (e. g., family practice, internal medicine), emergency medicine, urgent care, and specialty care (e. g. orthopedics), as well as radiologists who review radiographs across settings.
TechCare Trauma results are not intended to be used on a stand-alone basis for clinical decision-making. Primary diagnostic and patient management decisions are made by the clinical user.
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.
{5}------------------------------------------------
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."
{6}------------------------------------------------
Image /page/6/Picture/0 description: The image shows the word "MILVUE" in large, green letters. Below the word, in smaller letters, is the phrase "NEW TECH NEW CARE". The words "NEW TECH" are in a dark teal color, while the words "NEW CARE" are in the same green color as the word "MILVUE".
510(k) SUMMARY
Milvue's TechCare Trauma
The following 510(k) summary has been prepared pursuant to requirements specified in 21CFR 807.92(a)
Date Prepared: 17th of January 2025
Submitter 1.
Milvue 29 rue du Faubourg Saint-Jacques, 75014 Paris - FRANCE
Primary Contact Person:
Mathieu Quintin Quality and Regulatory Affairs Director Tel: +33 6 64 22 46 28 Email: mathieu@milvue.com
Secondary Contact Person:
Dr Alexandre Parpaleix CEO Tel : +33 6 03 02 47 79 Email: alexandre@milvue.com
2. Subject Device
Trade Name: TechCare Trauma Manufacturer: Milvue Classification Name: Radiological computer assisted detection/diagnosis software for fracture Regulation: Regulation: 21 CFR 892.2090 Regulatory Class: Class II Product Code: QBS
3. Predicate Device
Trade Name: BoneView 1.1-US Manufacturer: Gleamer 510(k) reference: K222176 Classification Name: Radiological computer assisted detection/diagnosis software for fracture Regulatory Class: Class II
{7}------------------------------------------------
Image /page/7/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE". The word "NEW" is in blue, and the words "TECH" and "CARE" are in green.
Product Code: QBS
4. Device Description
The TechCare Trauma device is a software as Medical Device (SaMD). More specifically it is defined as a "radiological computer assisted detection and diagnostic software for suspected fractures".
As a CADe/x software, TechCare Trauma is an image processing device intended to aid in the detection and localization of fractures and elbow joint effusions on acquired medical images (2D X-ray radiographs).
TechCare Trauma uses an artificial intelligence algorithm to analyze acquired medical images (2D X-ray radiographs) for features suggestive of fractures and elbow joint effusions.
TechCare Trauma can provide results for fractures in neonates and infants (from birth to less than 2 years), children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over) regardless of their condition.
TechCare Trauma can provide results for elbow joint effusions in children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).The device detects and identifies fractures and elbow joint effusions based on a visual model's analysis of images and provides information about the presence and location of these prespecified findings to the user.
It relies solely on images provided by DICOM sources. Once integrated into existing networks, TechCare Trauma automatically receives and processes these images without any manual intervention. The processed results, which consist of one or more images derived from the original inputs, are then sent to specified DICOM destinations. This ensures that the results can be seamlessly viewed on any compatible DICOM viewer, allowing smooth into medical imaging workflows.
TechCare Trauma can be deployed on-premises or on cloud and be connected to multiple DICOM sources / destinations (including but not limited to DICOM storage platform, PACS, VNA and radiological equipment, such as X-ray systems), ensuring easy integration into existing clinical workflows.
The training of TechCare Trauma was performed on a training dataset of 95266 images (age: range [0 – 102]; mean 53.3 +/- 26.4, median : 58) for all anatomical areas of interest in the Indications for Use and from various manufacturers.
{8}------------------------------------------------
Image /page/8/Picture/0 description: The image shows the logo for Milvue. The word "MILVUE" is written in large, green, sans-serif letters. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE", with "NEW TECH" in a dark teal color and "NEW CARE" in green.
Intended Use / Indications for Use 5.
TechCare Trauma is intended to analyze 2D X ray radiographs using machine learning techniques to aid in the detection, localization, and characterization of fractures and/or elbow joint effusion during the review of commonly acquired radiographs.
| Study Type (AnatomicArea of Interest) | Radiographic View(s)Supported | Findings | Patient Population |
|---|---|---|---|
| Ankle | Frontal, Lateral, Oblique | Fracture | Adults, Infants, Children, Adolescents |
| Foot | Frontal, Lateral, Oblique | Fracture | Adults, Infants, Children, Adolescents |
| Knee | Frontal, Lateral, Oblique | Fracture | Adults, Infants, Children, Adolescents |
| Leg (includesTibia/Fibula) | Frontal, Lateral | Fracture | Adults, Infants, Children, Adolescents |
| Femur | Frontal, Lateral | Fracture | Adults |
| Wrist | Frontal, Lateral, Oblique | Fracture | Adults, Children, Adolescents |
| Hand/Finger | Frontal, Lateral, Oblique | Fracture | Adults, Infants, Children, Adolescents |
| Elbow | Frontal, Lateral, Oblique | Fracture | Adults, Children, Adolescents |
| Lateral | Elbow JointEffusion | Adults, Children, Adolescents | |
| Forearm | Frontal, Lateral | Fracture | Adults, Infants, Children, Adolescents |
| Lateral | Elbow JointEffusion | Adults, Children, Adolescents | |
| Arm (includes Humerus) | Frontal, Lateral | Fracture | Adults, Neonates, Infants, Children,Adolescents |
| Lateral | Elbow JointEffusion | Adults, Children, Adolescents | |
| Study Type (AnatomicArea of Interest) | Radiographic View(s)Supported | Findings | Patient Population |
| Shoulder | Frontal, Lateral | Fracture | Adults, Infants, Children, Adolescents |
| Clavicle | Frontal | Fracture | Adults, Children, Adolescents,Neonates, Infants |
| Pelvis | Frontal | Fracture | Adults, Infants, Children, Adolescents |
| Hip | Frontal, Lateral | Fracture | Adults, Infants, Children, Adolescents |
| Thorax (includes ribs) | Frontal, Lateral, Ribsseries | Fracture | Adults |
{9}------------------------------------------------
Image /page/9/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in a smaller, teal-colored font. The text is centered and the background is white.
Table 1 : Intended use overview
TechCare Trauma can provide results for fracture in neonates and infants (from birth to less than 2 years), children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).
TechCare Trauma can provide results for elbow joint effusions in children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over).
The intended users of TechCare Trauma are clinicians with the authority to diagnose fractures and/or elbow joint effusions in various settings including primary care (e. g., family practice, internal medicine), emergency medicine, urgent care, and specialty care (e. g. orthopedics), as well as radiologists who review radiographs across settings.
TechCare Trauma results are not intended to be used on a stand-alone basis for clinical decision-making. Primary diagnostic and patient management decisions are made by the clinical user.
Operating mode & Outputs 6.
TechCare Trauma has been designed to solve the problem of missed fractures/EJE including subtle fractures, and thus detects fractures/EJE with a high sensitivity. In this regard, the display of findings is triggered by a "high-sensitivity operating point" (DOUBT) that will enable the display of a dotted-line bounding box around the region of interest. Additionally, the users need to be confident that when TechCare Trauma identifies a fracture/EJE, it is actually a fracture/EJE. In this regard, additional information is introduced to the user with a "high-specificity operating point" (YES).
Operating points
{10}------------------------------------------------
Image /page/10/Picture/0 description: The image shows the logo for "MILVUE" in large, green, sans-serif font. Below the company name is the text "NEW TECH NEW CARE", with "NEW TECH" in a dark blue color and "NEW CARE" in green. The logo is simple and modern, with a focus on technology and care.
TechCare Trauma operates based on two fixed operating points, which determine the result status of each finding:
- -DOUBT: suspicious area or subtle finding (when the level of confidence of the Al algorithm associated with the finding is above "high-sensitivity operating point" but below "high-specificity operating point").
- -YES: definite or unequivocal finding (when the level of confidence of the Al algorithm associated with the finding is above "high-specificity operating point").
Results status
For each image of the study, TechCare Trauma provides three possible statuses for each finding (such as fractures or elbow joint effusions):
- -YES: A definite or unequivocal finding has been identified by TechCare Trauma.
- -DOUBT: A suspicious area or subtle finding has been identified by TechCare Trauma.
- -NO: No suspicious area has been detected by TechCare Trauma.
Outputs
The outputs generated by TechCare Trauma consist of two individual and newly generated DICOM series associated with the original DICOM study, composed as follows:
Summary Image: the summary image is a newly generated secondary capture image derived from the regions of interest detected in the original images. It provides a study-level overall assessment for the entire set of images in a study. This assessment is based on the most severe findings identified across all the images in the study. TechCare Trauma can deliver four possible assessments for the summary image:
- -SUSPECTED: TechCare Trauma identified at least one finding above the operating point "YES" in any of the images in the study.
- -DOUBT: TechCare Trauma identified at least one finding above the operating point "DOUBT" but below the operating point "YES" in any of the images in the study.
- -NOT SUSPECTED: TechCare Trauma did not identify any findings above the operating point "DOUBT" in any of the images in the study.
- -NOT PROCESSED: All images in the study are out of the scope of TechCare Trauma's Indications for Use.
In cases where the study includes both processed images and at least one image that cannot be processed (e.g., a study with multiple series, where one series can be processed, and another series contains images outside the supported anatomic areas of interest), TechCare Trauma introduces derived assessments to indicate the presence of unprocessed images. These derived assessments are marked with an asterisk (*) to draw attention to the mixed processing status:
SUSPECTED*: TechCare Trauma identified at least one finding "YES", but there is also at least one image in the study that couldn't be processed because it is out of the Indications for Use.
{11}------------------------------------------------
Image /page/11/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in a smaller, teal, sans-serif font. The text is horizontally aligned and centered.
DOUBT*: TechCare Trauma identified at least one finding "DOUBT", but there is also at least one image in the study that couldn't be processed because it is out of the Indications for Use.
NOT SUSPECTED*: TechCare Trauma did not identify any finding above the operating point "DOUBT", but there is also at least one image in the study that couldn't be processed because it is out of the Indications for Use.
Annotated Secondary Capture Images (result images) : the annotated secondary capture images are provided for all the images processed by TechCare Trauma and contain additional information overlaid onto the original images:
- । Findings with Bounding Boxes: If a detection is made, one or more bounding boxes surrounding the finding site is drawn.
- । For a YES status, the bounding boxes are delineated by a green solid line.
- । For a DOUBT status, a white dotted line is used (see Operating points). Below the boxes, the corresponding finding label is displayed.
If no finding is detected, no bounding box is displayed.
DICOM Structured Report (SR)
If configured, TechCare Trauma can send back, for each original image, a related DICOM SR that lists the status and area for each finding.
Summary of Characteristics & Substantial Equivalence 7.
At a high level, the subject and predicate devices are based on the following same technological elements:
| Substantial Equivalence Table | |||
|---|---|---|---|
| Proposed DeviceTechCare Trauma | Predicate DeviceBone View | Similarities &Differences | |
| Number | K242171 | K222176 | NA |
| Applicant | Milvue | Gleamer | NA |
| Device Name | TechCare Trauma | BoneView 1.1-US | NA |
| ClassificationRegulation | 21 CFR 892.2090 | 21 CFR 892.2090 | Same |
| ProductCode | QBS | QBS | Same |
| DeviceClassification | Radiological computer assisteddetection/diagnosis software forfracture Class II | Radiological computer assisteddetection/diagnosis software forfracture Class II | Same |
| ImageModality | 2D Xray Images | 2D Xray Images | Same |
| NEW TECH | NEW CARE | ||
| IntendedUse | TechCare Trauma is intended toanalyze 2D X ray radiographs usingmachine learning techniques to aid inthe detection, localization, andcharacterization of fractures and/orelbow joint effusion during thereview of commonly acquiredradiographs. | BoneView is intended to aid in thedetection,localization, and characterization offractures onacquired medical images | Same |
| AnatomicAreas ofInterest | Fracture, for Adults (greater than 21years of age) andInfants/Children/Adolescents(between 28 days of ageand 21 years of age):AnkleFootKneeLegHand/FingerForearmArmShoulderClaviclePelvisHipFracture, for Adults (greater than 21years of age) andChildren/Adolescents (between 2years of ageand 21 years of age):WristFracture, for Adults only (greaterthan 21 years of age):FemurThorax (includes ribs)Fracture, for neonates (from 0 to 28days of age):Arm (includes Humerus)ClavicleElbow Joint Effusion, for Adults(greater than 21 years of age) andChildren/Adolescents (between 2years of ageand 21 years of age):Elbow (lateral)Forearm (lateral) | Fracture :Adults (greater than 21 years of age)andChildren/Adolescents (between 2years of ageand 21 years of age):AnkleFootKneeTibia/FibulaWristHandElbowForearmHumerusShoulderClavicleAdults (greater than 21 years of age)only:PelvisHipFemurRibsThoracic SpineLumbosacral Spine | Similar |
| Arm, includes Humerus (lateral) | |||
| ClinicalFindings | Fracture and Elbow Joint Effusion(EJE) | Fracture | Similar with theaddition of ElbowJoint Effusions forTechCare Trauma |
| IntendedUser | The intended users of TechCareTrauma are clinicians with theauthority to diagnose fracturesand/or elbow joint effusions invarious settings including primarycare (e. g., family practice, internalmedicine), emergency medicine,urgent care, and specialty care (e. g.orthopedics), as well as radiologistswho review radiographs acrosssettings. | The intended users of BoneView areclinicians with the authority todiagnose fractures in various settingsincluding primary care (e. g., familypractice, internal medicine),emergency medicine, urgent care, andspecialty care (e. g. orthopedics), aswell as radiologists who reviewradiographs across settings. | Same |
| PatientPopulation | For Fracture : Neonates and infants(from birth to less than 2 years),children and adolescents (aged 2 toless than 22 years) and adults (aged22 years and over).For EJE: Children and adolescents(aged 2 to less than 22 years) andadults (aged 22 years and over). | Adults (greater than 21 years of age)and Children/Adolescents (between 2years of age and 21 years of age) | Similar with theaddition ofneonates andinfants (frombirth to less than2 years) forTechCare Trauma |
| MachineLearningMethodology | Supervised Deep Learning | Supervised Deep Learning | Same |
| Imagesource | DICOM node (e.g., imaging device,intermediate DICOM node, PACSsystem, etc.) | DICOM node (e.g., imaging device,intermediate DICOM node, PACSsystem, etc.) | Same |
| ImageViewing | PACS systemImage annotations made on copy oforiginalimage or image annotations toggledon/off | PACS systemImage annotations made on copy oforiginalimage or image annotations toggledon/off | Same |
| DeploymentPlatform | Deployment on-premise or on cloudandconnection to several computingplatforms andX-ray imaging platforms such as X-rayradiographic systems or PACS | Deployment on-premise or on cloudandconnection to several computingplatforms andX-ray imaging platforms such as X-rayradiographic systems or PACS | Same |
{12}------------------------------------------------
Image /page/12/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in a smaller, teal-colored font. The text is horizontally aligned and centered.
{13}------------------------------------------------
Image /page/13/Picture/0 description: The image shows the logo for MILVUE. The word "MILVUE" is written in large, green, sans-serif letters. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in smaller, teal letters. The logo is simple and modern.
There are known minor differences between the TechCare Trauma device and the predicate device:
{14}------------------------------------------------
Image /page/14/Picture/0 description: The image shows the logo for "MILVUE" in large, green, sans-serif font. Below the logo, in a smaller, teal font, is the phrase "NEW TECH NEW CARE". The logo is simple and modern, with a focus on the company name and tagline.
- TechCare Trauma excludes spine analysis, -
- -TechCare Trauma expands indication for assisting clinicians in detecting elbow joint effusion for children and adolescents (aged 2 to less than 22 years) and adults (aged 22 years and over),
- -TechCare Trauma expands indications for assisting clinicians in detecting fractures in all ages, including neonates and infants.
All indications covered by TechCare Trauma, including elbow joint effusion and additional population subgroups (neonates and infants) compared to the predicate device, have been assessed as part of the performance studies (Standalone Model Performance Assessment and Multireader-Multicase studies). The results of these studies, as shown in below tables, demonstrate that the performance levels are maintained across all clinical indications and populations.
Therefore, the above aforementioned differences do not raise new questions of safety or effectiveness.
8. Performance tests data sets
8.1. Software verification and validation testing
TechCare Trauma development, verification, and validation activities have been carried out following FDA guidelines and relevant technical standards.
The software was tested against the established software requirements and specifications to assure the device performances as expected. The device hazard analysis has been completed and risk controls have been implemented to mitigate identified hazards.
The testing results demonstrate that all the software specifications have met the acceptance criteria as predefined in the test plans.
All activities have been appropriately documented in line with the FDA guidance on "Content of Premarket Submissions for Device Software Functions".
8.2. Biocompatibility testing
As a standalone software, TechCare Trauma has no direct patient or user contacting components. Therefore, biocompatibility information is not required for this device.
8.3. Electrical safety and electromagnetic compatibility (EMC)
As a standalone software, BoneView is not subject to electromagnetic compatibility or electrical safety testing activities. Therefore, Electrical safety and Electromagnetic compatibility information is not required for this device.
{15}------------------------------------------------
Image /page/15/Picture/0 description: The image shows the word "MILVUE" in large, green letters. Below that, the words "NEW TECH NEW CARE" are written in a smaller font. The words "NEW TECH" are in a dark teal color, while the words "NEW CARE" are in a green color that matches the color of the word "MILVUE".
8.4. Bench testing
Milvue performed a standalone performance testing for fracture detection on a dataset of 4109 radiographs of US adult patients (age: range [22 – 93]; mean 60.5 +/- 17.5 years) and a dataset of 2872 radiographs of US pediatric patients (age: range [0 – 21.9]; mean 12.2 +/- 5.6 years) , for all anatomical areas and views of interest in the Indications for Use. Milvue also performed a performance testing for EJE detection on a dataset of 280 radiographs of US adult patients (age: range [22 – 93]; mean 63.0 +/- 17.3 years) and a dataset of 483 radiographs of US pediatric patients (age: range [2 – 21.7]; mean 9.2 +/- 5.5 years). Cases were obtained from various states in the US (at least 4), and various imaging hardware manufacturers (at least 14). These datasets were independent of the data used for model training, tuning, and establishment of device operating points.
The ground-truth was established by American Board of Radiology (ABR)-certified radiologists with a minimum of 5 years of experience since ABR certification. Pediatric and adult cases followed two parallel ground-truthing (GT) pathways: the pediatric cases were annotated by a pediatric GT panel made of three ABR-certified pediatric radiologists and the adult cases by an adult GT panel made of three ABR-certified musculoskeletal (MSK) radiologists independently interpreted each case for the presence or absence of fracture and EJE using the standard clinical definitions of these pathologies. The third radiologist independently reviewed the cases where there was disagreement between the first two. The final reference standard was determined by majority consensus.
The primary endpoint was the image-level Area Under The Curve (AUC) of the Receiver Operating Characteristic (ROC). All statistical analyses were performed separately for the adult and for the pediatric population.
The results of standalone testing demonstrated a high accuracy and homogeneous performance in detecting fractures and elbow joint effusion (EJE) in both US adult and pediatric populations :
- -For fracture detection, the ROC-AUC was 0.962 [0.957; 0.967] in adults and 0.962 [0.955; 0.969] in pediatrics, demonstrating that TechCare Trauma detects fractures in radiographs with similar performances on the adult population and on the pediatric population.
- -For EJE detection, the ROC-AUC was 0.965 [0.936; 0.986] in adults and 0.976 [0.963; 0.986] in pediatrics, indicating a high level of accuracy, similar to the performance for fracture detection.
- -The testing demonstrated high image-level ROC-AUC with narrow 95% confidence intervals for both fracture and elbow joint effusion (EJE) detection in adult and pediatric populations, demonstrating that there are no additional risks to the user.
- -TechCare Trauma performs with high accuracy across study types (anatomic areas of interest, views, patient age and sex and machine) and across potential confounders including different imaging hardware manufacturers and displaced/non-displaced fractures.
{16}------------------------------------------------
Image /page/16/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in a smaller, teal-colored font. The background of the image is white.
Image-level ROC-AUC Summary for TechCare Trauma
| Population (Nb of images) | Image-level ROC-AUC [95% CI] |
|---|---|
| Fracture - Adult (4109) | 0.962 [0.957 - 0.967] |
| Fracture - Pediatric (2872) | 0.962 [0.955 - 0.969] |
| EJE - Adult (280) | 0.965 [0.936 - 0.986] |
| EJE - Pediatric (483) | 0.976 [0.963 - 0.986] |
Image-level ROC-AUC Summary for fracture detection
| Category (Nb of imagesadult/pediatric) | Adult ROC-AUC [95% CI] | Pediatric ROC-AUC [95% CI] |
|---|---|---|
| By Anatomical Regions | ||
| Ankle (293 / 283) | 0.941 [0.913; 0.965] | 0.948 [0.916; 0.976] |
| Arm (197 / 183) | 0.981 [0.962; 0.996] | 0.988 [0.978; 0.996] |
| Clavicle (210 / 171) | 0.998 [0.993; 1.000] | 0.985 [0.965; 1.000] |
| Elbow (364 / 326) | 0.933 [0.899; 0.960] | 0.942 [0.911; 0.970] |
| Femur (244 / NA) | 0.968 [0.945; 0.988] | NA |
| Foot (298 / 299) | 0.972 [0.955; 0.988] | 0.964 [0.937; 0.985] |
| Forearm (197 / 195) | 0.969 [0.946; 0.988] | 0.998 [0.995; 1.000] |
| Hand/Fingers (284 / 281) | 0.951 [0.923; 0.976] | 0.934 [0.899; 0.964] |
| Hip/Pelvis (221 - pediatric) | 0.945 [0.898; 0.979] | 0.893 [0.849; 0.931] |
| Hip (327 - adult) | ||
| Pelvis (163 - adult) | 0.955 [0.923; 0.979] | |
| Knee (284 / 189) | 0.947 [0.921; 0.969] | 0.948 [0.891; 0.990] |
| Leg (216 / 226) | 0.979 [0.957; 0.995] | 0.959 [0.930; 0.983] |
| Shoulder (324 / 238) | 0.962 [0.940; 0.980] | 0.991 [0.981; 0.998] |
| Thorax/Rib (414 / NA) | 0.939 [0.912; 0.961] | NA |
| Wrist (264 / 260) | 0.946 [0.920; 0.974] | 0.973 [0.952; 0.990] |
| Female (2556 / 1090) | 0.962 [0.955; 0.969] | 0.964 [0.952; 0.975] |
| Male (1553 / 1766) | 0.962 [0.953; 0.970] | 0.961 [0.952; 0.970] |
| By Image View | ||
| Frontal (1874 / 1530) | 0.971 [0.964; 0.978] | 0.966 [0.955; 0.975] |
| Lateral (1161 / 805) | 0.955 [0.942; 0.965] | 0.961 [0.946; 0.972] |
| Oblique (771 / 297) | 0.946 [0.932; 0.962] | 0.965 [0.944; 0.983] |
| By Age | ||
| 22 to <65 years (2205) | 0.972 [0.965; 0.977] | NA |
| ≥ 65 years (1904) | 0.950 [0.940; 0.959] | NA |
| <2 years (96) | NA | 0.948 [0.888; 0.992] |
| 2 to <12 years (998) | NA | 0.971 [0.960; 0.981] |
| 12 to <18 years (1218) | NA | 0.955 [0.943; 0.967] |
| 18 to <22 years (560) | NA | 0.965 [0.949; 0.979] |
| By Imaging Hardware Manufacturer | ||
| Konica Minolta (2617 / 709) | 0.952 [0.944; 0.959] | 0.956 [0.941; 0.969] |
| Siemens (1107 / 657) | 0.974 [0.963; 0.984] | 0.953 [0.935; 0.969] |
| Samsung (136 / 257) | 0.978 [0.948; 1.000] | 0.944 [0.912; 0.971] |
| Agfa (NA / 889) | NA | 0.972 [0.957; 0.985] |
| Carestream (NA / 59) | NA | 1.000 [1.000; 1.000] |
| Philips (NA / 81) | NA | 0.933 [0.870; 0.981] |
| RamSoft (NA / 69) | NA | 0.996 [0.982; 1.000] |
| Fujifilm (67 / NA) | 0.993 [0.969; 1.000] | NA |
| Other (181 / 148) | 0.962 [0.931; 0.986] | 0.969 [0.943; 0.988] |
| By Particular Groups | ||
| Displaced (3434 / 2066) | 0.970 [0.964; 0.975] | 0.969 [0.960; 0.977] |
{17}------------------------------------------------
Image /page/17/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in a smaller, teal-colored font. The text is horizontally aligned and centered.
{18}------------------------------------------------
Image /page/18/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE", with "NEW TECH" in a dark teal color and "NEW CARE" in green. The text is horizontally aligned and centered.
Image-level Sensitivity and Specificity for Fracture detection - Adult population (N = 4109)
| Se (High Sp) | Sp (High Sp) | Se (High Se) | Sp (High Se) | |
|---|---|---|---|---|
| All | 0.906 [0.891; 0.919] | 0.900 [0.887; 0.912] | 0.940 [0.928; 0.951] | 0.845 [0.830; 0.859] |
Image-level Sensitivity and Specificity for EJE detection - Adult population (N = 280)
| Se (High Sp) | Sp (High Sp) | Se (High Se) | Sp (High Se) | |
|---|---|---|---|---|
| All | 0.821 [0.719; 0.893] | 0.980 [0.949; 0.993] | 0.859 [0.764; 0.923] | 0.955 [0.917; 0.978] |
Image-level Sensitivity and Specificity for Fracture detection – Pediatric population (N = 2872)
| Se (High Sp) | Sp (High Sp) | Se (High Se) | Sp (High Se) | |
|---|---|---|---|---|
| All | 0.900 [0.882; 0.916] | 0.927 [0.913; 0.938] | 0.930 [0.914; 0.943] | 0.875 [0.859; 0.891] |
Image-level Sensitivity and Specificity for EJE detection – Pediatric population (N = 483)
| Se (High Sp) | Sp (High Sp) | Se (High Se) | Sp (High Se) | |
|---|---|---|---|---|
| All | 0.669 [0.587; 0.744] | 0.977 [0.954; 0.989] | 0.901 [0.839; 0.941] | 0.933 [0.901; 0.955] |
8.5. Animal testing
Not applicable. Animal studies are not necessary to establish the substantial equivalence of this device.
8.6. Clinical data
Milvue conducted a fully-crossed multiple readers, multiple case (MRMC) retrospective reader study to determine the impact of TechCare Trauma on reader performance in diagnosing fractures and EJE.
The study was based on external multicenter anonymized datasets of conventional radiographs of all anatomic areas of interest of the appendicular and chest skeleton of pediatric and adult US patients.
The primary objective of the study was to determine whether the diagnostic accuracy of readers aided by TechCare Trauma is superior to the diagnostic accuracy of readers unaided by TechCare Trauma, as determined by the case-level AUC of the ROC curves (primary endpoint).
An overview of the design and results is provided below:
{19}------------------------------------------------
Image /page/19/Picture/0 description: The image shows the word "MILVUE" in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE" in a smaller, teal-colored font. The text is simple and modern, and the colors are bright and eye-catching.
- 4 dataset populations were individualized, obtained from sites that were different from the training data sites to ensure the independence of the test data from training data :
- -Adult US population for fracture detection (304 radiological cases)
- -Pediatric US population for fracture detection (256 radiological cases)
- -Adult US population for EJE detection (109 radiological cases)
- । Pediatric US population for EJE detection (100 radiological cases)
- -All examinations were independently interpreted by readers for the presence or absence of fracture and EJE, both with and without CADe assistance. The reading scenario (initial reading with or without CADe assistance) was randomly assigned for each case. A washout period of at least one month was observed between scenarios.
- -Each reader provided a binary determination and localization of fractures using bounding boxes and assigned an ordinal report score, which was used for ROC data analysis.
- -The ground truth was established by ABR-certified radiologists with at least five years of experience. Pediatric cases were annotated by a panel of three pediatric radiologists, while adult cases were reviewed by a panel of three MSK radiologists. Two radiologists independently assessed each case, with a third reviewing cases with initial disagreements. The final reference standard was determined by majority consensus.
- -All statistical analyses were performed separately for the adult and for the pediatric population. The Obuchowski-Rockette-Hillis (ORH) modeling approach was employed to compare the diagnostic accuracy of readers with and without assistance, accounting for the variability among multiple readers and cases.
The results of the study found that the diagnostic accuracy of readers in the intended use population is superior when aided by TechCare Trauma than when unaided by TechCare Trauma, as measured at the task of fracture and/or EJE detection using the AUC of the ROC curve.
In particular, the study results demonstrated that reader demonstrated significantly improved performance for detecting fractures and elbow joint effusions (EJE) in both adult and pediatric populations with :
For adult fracture detection
- Reader ROC AUC significantly improved from 0.865 [0.822; 0.907] to 0.955 [0.924; 0.979], a delta of 0.090 [0.060; 0.121] (p < 0.001).
- Reader Sensitivity significantly improved from 0.807 (95% Cl: 0.737-0.865) to 0.983 (95% Cl: 0.966-0.996): +21.8% increase of the Sensitivity
- Reader Specificity improved from 0.815 (95% Cl: 0.724-0.886) to 0.827 (95% Cl: 0.734-0.901): +1.47% increase of the Specificity
For adult EJE detection
- Reader ROC AUC significantly improved from 0.851 [0.783; 0.913] to 0.914 [0.862; 0.959], a delta of 0.064 [0.026; 0.101] (p < 0.001).
- Reader Sensitivity significantly improved from 0.872 (95% Cl: 0.789-0.940) to 0.983 (95% Cl: ● 0.954-1.000): +12.7% increase of the Sensitivity
{20}------------------------------------------------
Image /page/20/Picture/0 description: The image shows the logo for MILVUE. The text "MILVUE" is in large, green, sans-serif font. Below the main text, "NEW TECH" is in dark blue, and "NEW CARE" is in green.
- . Reader Specificity improved from 0.738 (95% Cl: 0.635-0.835) to 0.746 (95% Cl: 0.645-0.845): +1.08% increase of the Specificity
For pediatric fracture detection
- Reader ROC AUC significantly improved from 0.857 [0.810; 0.899] to 0.931 [0.892; 0.963], a delta of 0.074 [0.044; 0.104] (p<0.001)
- Reader Sensitivity significantly improved from 0.804 (95% Cl: 0.747-0.864) to 0.964 (95% Cl: 0.940-0.984): +19.9% increase of the Sensitivity
- Reader Specificity remained the same at 0.797 (95% Cl: 0.700-0.873) to 0.797 (95% Cl: 0.700-0.882): 0% increase of the Specificity
For pediatric EJE detection
- Reader ROC AUC significantly improved from 0.877 [0.824; 0.929] to 0.941 [0.890; 0.978], a delta of 0.063 [0.025; 0.101] (p = 0.002)
- . Reader Sensitivity significantly improved from 0.825 (95% Cl: 0.742-0.895) to 0.975 (95% Cl: 0.939-1.000): +18.2% increase of the Sensitivity
- Reader Specificity improved from 0.839 (95% Cl: 0.764-0.914) to 0.851 (95% Cl: 0.769-0.926) : +1.43% increase of the Specificity
Additionally, subgroup/subpopulation analysis was carried out by age, gender, ethnicity, US state, displaced and non-displaced fractures, imaging hardware manufacturer, readers qualification.
{21}------------------------------------------------
Image /page/21/Picture/0 description: The image shows the logo for MILVUE. The text "MILVUE" is in large, green, sans-serif font. Below the main text, the words "NEW TECH" are in blue, and the words "NEW CARE" are in green.
Clinical Reader Study Results TechCare Trauma-Aided vs TechCare Trauma-Unaided ROC Curves
Image /page/21/Figure/2 description: The image shows two Receiver Operating Characteristic (ROC) curves, one representing a model without AI and the other with AI. The ROC curve for the model without AI has an area under the curve (AUC) of 0.865, while the model with AI has a higher AUC of 0.955. The x-axis represents the false positive rate, and the y-axis represents the true positive rate. The ROC curve with AI is higher than the ROC curve without AI, indicating that the model with AI has better performance.
Adult - Fracture Population ●
Image /page/21/Figure/4 description: The image shows a bullet point followed by the text "Adult - EJE Population". The text is written in a clear, sans-serif font and is left-aligned. The bullet point is a solid black circle.
Image /page/21/Figure/5 description: The image shows two Receiver Operating Characteristic (ROC) curves, one representing a model without AI and the other with AI. The x-axis represents the False Positive Rate, while the y-axis represents the True Positive Rate. The area under the ROC curve (AUC) for the model without AI is 0.851, while the AUC for the model with AI is 0.914, indicating that the AI model has better performance.
{22}------------------------------------------------
Image /page/22/Picture/0 description: The image shows the logo for MILVUE. The word "MILVUE" is written in large, green, sans-serif font. Below the word "MILVUE" is the phrase "NEW TECH NEW CARE", which is written in a smaller, blue, sans-serif font.
● Pediatric - Fracture Population
Image /page/22/Figure/2 description: The image is a plot of the true positive rate versus the false positive rate. There are two ROC curves on the plot, one for average ROC curve without AI which has an area of 0.857, and one for average ROC curve with AI which has an area of 0.931. The ROC curve with AI is higher than the ROC curve without AI, indicating that the AI improves performance. A dashed line is also plotted as a reference.
● Pediatric - EJE Population
Image /page/22/Figure/4 description: The image shows two ROC curves, one for a model without AI and one for a model with AI. The x-axis represents the false positive rate, and the y-axis represents the true positive rate. The area under the ROC curve for the model without AI is 0.877, while the area under the ROC curve for the model with AI is 0.940. The ROC curve for the model with AI is higher than the ROC curve for the model without AI, indicating that the model with AI has better performance.
Conclusions
TechCare Trauma demonstrates substantial equivalence to the predicate device.
TechCare Trauma has the same intended uses and similar indications, technological characteristics, and principles of operation as its predicate device. The minor differences in indications do not alter the intended use of the device and do not affect its safety and effectiveness when used as labeled. Performance testing was conducted to validate the performance of TechCare Trauma on the new patient population and suspected findings. Performance data demonstrates the subject device's (TechCare Trauma) substantial equivalence to the predicate device (BoneView 1.1-US).
§ 892.2090 Radiological computer-assisted detection and diagnosis software.
(a)
Identification. A radiological computer-assisted detection and diagnostic software is an image processing device intended to aid in the detection, localization, and characterization of fracture, lesions, or other disease-specific findings on acquired medical images (e.g., radiography, magnetic resonance, computed tomography). The device detects, identifies, and characterizes findings based on features or information extracted from images, and provides information about the presence, location, and characteristics of the findings to the user. The analysis is intended to inform the primary diagnostic and patient management decisions that are made by the clinical user. The device is not intended as a replacement for a complete clinician's review or their clinical judgment that takes into account other relevant information from the image or patient history.(b)
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 algorithm, including a description of the algorithm inputs and outputs, each major component or block, how the algorithm and output affects or relates to clinical practice or patient care, and any algorithm limitations.
(ii) A detailed description of pre-specified performance testing protocols and dataset(s) used to assess whether the device will provide improved assisted-read detection and diagnostic performance as intended in the indicated user population(s), and to characterize the standalone device performance for labeling. Performance testing includes standalone test(s), side-by-side comparison(s), and/or a reader study, as applicable.
(iii) Results from standalone performance testing used to characterize the independent performance of the device separate from aided user performance. The performance assessment must be based on appropriate diagnostic accuracy measures (
e.g., receiver operator characteristic plot, sensitivity, specificity, positive and negative predictive values, and diagnostic likelihood ratio). Devices with localization output must include localization accuracy testing as a component of standalone testing. The test dataset must be representative of the typical patient population with enrichment made only to ensure that the test dataset contains a sufficient number of cases from important cohorts (e.g., subsets defined by clinically relevant confounders, effect modifiers, concomitant disease, 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) Results from performance testing that demonstrate that the device provides improved assisted-read detection and/or diagnostic performance as intended in the indicated user population(s) when used in accordance with the instructions for use. The reader population must be comprised of the intended user population in terms of clinical training, certification, and years of experience. The performance assessment must be based on appropriate diagnostic accuracy measures (
e.g., receiver operator characteristic plot, sensitivity, specificity, positive and negative predictive values, and diagnostic likelihood ratio). Test datasets must meet the requirements described in paragraph (b)(1)(iii) of this section.(v) Appropriate software documentation, including device hazard analysis, software requirements specification document, software design specification document, traceability analysis, system level test protocol, pass/fail criteria, testing results, and cybersecurity measures.
(2) Labeling must include the following:
(i) A detailed description of the patient population for which the device is indicated for use.
(ii) A detailed description of the device instructions for use, including the intended reading protocol and how the user should interpret the device output.
(iii) A detailed description of the intended user, and any user training materials or programs that address appropriate reading protocols for the device, to ensure that the end user is fully aware of how to interpret and apply the device output.
(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 must include 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) 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, such as anatomical characteristics, patient demographics and medical history, user experience, and imaging equipment.