Implicity

The cleared algorithms are the engine of the product's main promise, and the platform carries other value. Implicity aggregates and normalizes raw data transmitted by cardiac implantable devices from Abbott, Biotronik, Boston Scientific, Medtronic and MicroPort into one interface, automates remote monitoring billing workflows and reports, and manages patient connectivity; on top sit three algorithms that decide what a clinician reviews: the ILR ECG Analyzer for insertable cardiac monitor detections, the AF alert management module, and SignalHF for heart failure decompensation.

The company states the value in model terms (79 percent fewer false positives at 99 percent sensitivity, 85 percent fewer unnecessary AF alerts). Without the algorithms a clinic would still have one normalized view of every manufacturer's data and billing documentation, which is useful on its own, so the models are the engine of a core module rather than the whole product.

The autonomy is a filter, and its tuning is published; what happens to what it filters is not. The models decide which device alerts reach a clinician, and a suppressed alert is an event nobody sees, so sensitivity is the safety question: the company reports 98.3 and 94.3 percent across monitor generations while accepting specificity of 61.6 and 75.6 percent, passing non actionable alerts through rather than risking a missed one.

AF episodes are classified against European Society of Cardiology recommendations, so the line between actionable and non actionable follows a published guideline, and the products are described as clinical decision support and an adjunct to remote monitoring that does not replace the practitioner's judgment. What is missing is the control on suppressed alerts: whether they stay retrievable, whether a clinic can audit what was filtered, and whether any sampled review of filtered events occurs.

Performance and inputs are documented in detail, without the model and version discipline behind them. The company publishes both sides of its claims with denominators (ILR ECG Analyzer sensitivity 98.64 percent, 509 of 516, false positive rate 24.03 percent, 68 of 283), footnotes each figure to a study, product code or dated internal data, publishes intended use text verbatim, names SignalHF's inputs (arrhythmia burden, physical activity, nighttime heart rate and thoracic impedance), anchors AF classification to European Society of Cardiology recommendations, and names France's Health Data Hub as its development data source.

What is missing is the model itself and its change practice: no architecture or training set size is stated, and the platform runs the current version for every clinic with no customer managed updates, so how algorithm changes are versioned and announced is not published. The ILR ECG Analyzer received a new clearance in July 2026.

The model chain is structurally short and its data sources are named. The three algorithms are Implicity's own, cleared or CE marked in its name, developed with access to France's Health Data Hub, a government platform with its own published authorization rules, and fed by raw transmissions from five named device manufacturers (Abbott, Biotronik, Boston Scientific, Medtronic, MicroPort), whose interface changes are a known, traceable dependency.

The platform is offered through a public cloud marketplace. The site names no hosting provider, subprocessor list, third party framework or pretrained component; the privacy policy refers to hosting and system administration providers without naming them.

The filtering claims rest on peer reviewed, multi manufacturer studies with the method available, and both sides of the tradeoff are published. A retrospective study of more than 4,000 patients across four device manufacturers, published in the Cardiovascular Digital Health Journal, evaluated the AF alert management algorithm (the company reports 85 percent fewer unnecessary AF alerts); a study on more than 2,800 arrhythmia episodes reported nearly 80 percent fewer false positives from the insertable monitor analyzer; and EHRA 2026 work reported sensitivity of 98.3 and 94.3 percent with specificity of 61.6 and 75.6 percent and positive predictive value around 74 percent across monitor generations.

The company also reports 28 percent of disconnected patients reconnected within two days and more than 90 percent billing compliance from internal data, and more than 110,000 patients monitored at over 250 facilities in the United States and Europe. The home page's 26 percent mortality and 4 percent length of stay reductions are stated against conventional monitoring without a cited study on the page. The designs are retrospective, which fits an accuracy claim on recorded device data.

The development data source is named and the handling of customer data is not. The company states it develops its AI with access to France's Health Data Hub, a government platform combining patient databases for research under its own authorization and ethics process, which gives the development corpus a governance a reader can examine.

For the data clinics send, the security page describes encryption of health data and flows, two factor authentication, access controls, activity tracking and an impact assessment, and says the company is designated a health data host in France; the website privacy policy says personal health data is not transferred outside the European Economic Area.

Nothing states whether customer transmissions are retained beyond clinical use, whether they feed continued model development, how they are de-identified, or whether a clinic can decline secondary use, for raw device data on more than 110,000 patients. No safety engineering is published.

HIPAA compliance is claimed and no document reaching the US service is published. The company's security page says it has implemented procedures to comply with HIPAA requirements to protect US patient data, and describes encryption, two factor authentication, access controls and activity tracking; its privacy policy covers website visitors and contacts only.

No business associate status, agreement, tier or subcontractor terms are published for US clinics, for a platform that ingests identifiable device transmissions for more than 110,000 patients and generates billing documentation. For France, the company is a designated health data host, and it states that personal health data is not transferred outside the European Economic Area; how a French company handles US patient data, and under what agreement, is not described.

A recognized security certification is named in the company's own material without the artifact. The home page lists ISO 27001, alongside ISO 13485 (a device quality standard) and France's HDS health data hosting certification, and the company's security page states it is ISO 27001 certified for the management of its information system security and designated a health data host under French law, and describes encryption of health data and flows, data integrity checks, two factor authentication, activity and incident tracking, access controls and an impact analysis.

No certificate, scope, date or certifying body is given, and no SOC 2 report, penetration testing statement or vulnerability disclosure route is published, so the artifact has to be requested. The platform holds raw implanted device data on more than 110,000 patients in a data path shared with five device manufacturers.

The regulatory position is precise and traceable. FDA's database lists three 510(k) clearances for Implicity, Inc.: IM007, the ILR ECG Analyzer (K210543, November 2021, and a new version, K253463, 2 July 2026, product code DQK, 21 CFR 870.1425), and SignalHF (IM008, K230842, October 2023, product code QNL, 870.2210).

The company states the ILR ECG Analyzer is FDA cleared Class II and CE marked Class I, the AF alert management module CE marked Class I, and SignalHF FDA cleared Class II and not available in Europe, and publishes intended use text that limits the analyzer to insertable cardiac monitor data supplied by remote monitoring platforms, initially from Medtronic devices, for review by qualified professionals. Class I CE marking under the old medical device directive is self certified and the framework has since been replaced, so European status may need refreshing.

Performance is reported by device generation and manufacturer, not by patient group. Work presented at EHRA 2026 reports the insertable monitor algorithm's sensitivity as 98.3 percent on devices with their own onboard AI and 94.3 percent on devices without it, with specificity of 61.6 and 75.6 percent and positive predictive value around 74 percent in both, which shows the filter behaves differently depending on the hardware generation feeding it; the AF study spans more than 4,000 patients across four manufacturers.

No results by sex, age or comorbidity are published, no training population composition is given, and no governance framework or model change policy is published, although the development data source (France's Health Data Hub) is named. Cardiac device populations skew older and male, and atrial fibrillation burden and heart failure decompensation patterns differ by sex, so a filter tuned on that distribution may behave differently for less represented patients.

The company publishes an error rate for its filter with its denominators. Its instructions for use material for the ILR ECG Analyzer states sensitivity of 98.64 percent (509 of 516) with a false positive rate of 24.03 percent (68 of 283), and its site reports 79 percent fewer false positives while keeping 99 percent sensitivity on Medtronic insertable monitor recordings; EHRA 2026 work gives sensitivity of 98.3 and 94.3 percent and specificity of 61.6 and 75.6 percent across monitor generations.

These are test results, not performance in use. The failure that matters is suppression: an event filtered out is reviewed by nobody, so the published sensitivity lets a clinic size the residual risk. The products are described as clinical decision support that does not replace the practitioner's responsibility or judgment, and as an adjunct to remote monitoring.

No indemnity, performance warranty or service level is published, and nothing says whether filtered alerts remain retrievable for audit, the control a clinic would use to investigate a suspected miss.

Manufacturer interoperability is named and built at the data level, and record system integration is not described. The platform aggregates, normalizes and standardizes raw discrete transmissions from Abbott, Biotronik, Boston Scientific, Medtronic and MicroPort cardiac implantable devices, rather than the summary documents their portals produce, which lets one algorithm run across a clinic's mixed device population; SignalHF is described as compatible across manufacturers. Output reaches billing workflows and generated reports. No electronic health record is named, no standard is cited, and nothing says how a reviewed transmission or report reaches the patient's chart.

The service is a cloud platform with no local installation, available through a public cloud marketplace, and European health data residency is committed: the company is designated a health data host in France under article L.1111-8 of the Public Health Code, and its privacy policy says there is no transfer of personal health data to a country outside the European Economic Area. Much else is left open.

The hosting provider goes unnamed, the location of US patients' data is not stated, no tenancy or segregation model is described for clinics on two continents, and there is no availability commitment for an alert stream in a diagnostic pathway.

No price is published, and how the service is paid for is discoverable. The company cites the article of the French social security code under which its AF alert management module is covered, a provision a French clinic can check; the platform automates remote monitoring billing workflows and the company reports customers reaching more than 90 percent billing compliance, which a practice can test against its own billing data; and the platform is available through a public cloud marketplace.

No per patient or per clinic price, implementation cost, term or minimum is published, and nothing says whether the three cleared algorithms are included in a base subscription or licensed separately, which matters because SignalHF is not available in Europe and the AF module is CE marked only.

Coverage is named precisely, with validation behind part of it. The platform supports devices from all five major cardiac implantable device manufacturers (Abbott, Biotronik, Boston Scientific, Medtronic, MicroPort) across pacemakers, defibrillators, resynchronization devices and insertable monitors, addresses arrhythmia and heart failure, serves clinics in the United States and Europe, and is designed for physicians and for the nurses and technicians who do most remote monitoring review.

Validation stands behind part of it: the AF study spans four manufacturers, the insertable monitor analyzer's clearance initially covered Medtronic devices, SignalHF is cleared in the US and not available in Europe, and the AF module is CE marked only.