CureMetrix was created to address a mission-critical need in the industry – to help radiologists better detect breast cancer with the objective, data-driven answers they need to quickly support patients and their healthcare teams. To support that mission, cmAssist, an investigational CAD, uses AI intended to help the radiologist identify, mark, and score regions of interest on screening mammograms.

Value Proposition

The number one benefit that cmAssist offers radiologists, health care practices and patients is peace of mind by delivering CAD that Works to help deliver:,
--Fewer unnecessary patient recalls – 69% reduction in false positives vs. traditional CAD, leaving more time for accurate diagnoses using cmAssist
--Improved cancer detection – 27% increase in early cancer detection without an increase in false-positive recalls

cmAssist demonstrates the ability to find cancers up to six years before first detection, and help radiologists improve their detection on average 27%, without increasing recalls.

The model works across all breast densities and lesion sizes, and it works on masses and calcifications for:
--Improved sensitivity
--Improved specificity


The operational flow of the cmAssist begins when digital two-dimensional (2D) mammograms are captured by a Full-Field Digital Mammography (FFDM) system and the processed – For Presentation – images are deposited on the PACS.

The CureMetrix cmEdge client receives the new mammography DICOM image(s) from the PACS, creates a local copy of the image(s), de-identifies the local copy, encrypts the local copy, transmits the local copy to the CureMetrix cloud, and then deletes the local copy.

Within the CureMetrix cloud, cmAssist receives the DICOM image(s), decrypts the image(s), groups them by study, analyzes the image(s) within the study, and then produces results for each study. The results are transmitted to the PACS, the results are decrypted and re-associated with the exam. The results are produced in the form of DICOM Mammography CAD Structured Report (SR) overlay file that includes the type and location of CAD marks and the neuScore for each ROI.

The results file is encrypted and transmitted by the cmAssist from the CureMetrix cloud back to the cmEdge client where it is decrypted and sent (via DICOM) to the PACS. The CAD SR is then routed to the viewing workstation.

After the radiologist has reviewed the mammograms on the viewing workstation and formed an initial interpretation, the radiologist can enable the display of the CAD markings through the viewer functionality as well as the display the cmAssist results. Using the viewing workstation functions, the radiologist can turn on and off the display of the cmAssist results to inspect the mammograms with and without the ROIs marked.

Intended Use

cmAssist™ is a software-as-a-service (SaaS) intended to identify, mark, and score regions of interest on Full Field Digital Mammography (FFDM) screening and nonstandard views  and bring these regions to the attention of the radiologist after the initial reading has been completed. cmAssist also provides a quantified indication of the level of suspicion for each region of interest to assist the radiologist as a decision support tool with clinical determination.


There are no known direct risks to safety or health of the operator or the patient that are related to the use of the Service. There is no direct contact with the patient.

Indirect, inherent risks to the patient are: a) that the Service may not mark all actionable areas; and b) that the Service may mark regions that are not actionable. Additionally, cmAssist may increase the false-positive rates for screening mammography. Increased false-positives may lead to recalls and work-ups that may include additional imaging, biopsies, and other procedures.

Model performance metrics

With a sensitivity at 90%, the algorithm produced a .05 false positives per image (FPPI) for calcifications, .06 FPPI for mass lesions and .123 FPPI overall.

At the sensitivity of 90%, the specificity is 87% for calcifications, 80% for mass lesions and 71% overall.

Performance Curves

Performance Curves

The following figure represents the ROC of the algorithm along with the 95% confidence intervals and the mean AUC.



CureMetrix was established in 2014 with the goal of early and accurate detection of breast cancer in mammograms. Radiologists have needed a CAD that Works® for 20 years, and CureMetrix has delivered:
--cmTriage™, the first FDA-cleared AI-based triage software for mammography screening in the U.S. and
--cmAssist® (investigational software) to identify, mark and score anomalies in breast cancer screening

In studies published in the Journal of Digital Imaging, cmAssist was able to demonstrate the ability to find cancers up to six years before first detection, and help radiologists improve their detection on average 27% with less than a 1% increase in recall rates.¹

Further, in a study that included a head-to-head comparison with traditional computer aided detection (CAD) software, CureMetrix AI-driven CAD cmAssist was able to demonstrate 69% fewer false markings.²

Learn more