The Free Testosterone AccuBind® ELISA Test System is an Enzyme Immunoassay (EIA) for the quantitative measurement of free testosterone in human serum. Measurement of free testosterone is used in the diagnosis and treatment of disorders involving the male sex hormones (androgens), including primary and secondary hypogonadism, impotence in males and in females; hirsutism (excessive hair) and virilization (masculinization) due to tumors, polycystic ovaries and adrogenital syndromes.
Device Story
Competitive enzyme immunoassay (EIA) for quantitative measurement of free testosterone in human serum; utilizes immobilized testosterone antibody, enzyme-antigen conjugate (HRP-labeled), and native antigen from patient sample. Competitive reaction occurs between native free antigen and enzyme-antigen conjugate for limited binding sites on antibody-coated microplate. After equilibrium, unbound antigen is removed via decantation/aspiration. Substrate (TMB/hydrogen peroxide) added to produce blue color, followed by acid stop solution to yield yellow color. Microplate reader (450nm/620nm) measures absorbance; dose-response curve generated from serum calibrators determines unknown sample concentration. Used in clinical laboratories by trained personnel. Output assists clinicians in diagnosing/treating androgen-related disorders (hypogonadism, hirsutism, virilization).
Clinical Evidence
Bench testing only. Precision study (n=80 per sample) showed total CVs 7.8-12.4%. Linearity confirmed across 0.11-60 pg/mL (R^2=0.9888). LoB 0.0295 pg/mL, LoD/LoQ 0.0519 pg/mL. Interference and cross-reactivity studies showed no significant bias (>10%) for tested substances. Method comparison with predicate (n=137) yielded y=1.017x-0.244, r=0.997. Reference range established using 261 healthy subjects.
Indicated for quantitative measurement of free testosterone in human serum for diagnosis and treatment of androgen-related disorders, including primary/secondary hypogonadism, impotence, hirsutism, and virilization in males and females.
Regulatory Classification
Identification
A testosterone test system is a device intended to measure testosterone (a male sex hormone) in serum, plasma, and urine. Measurement of testosterone are used in the diagnosis and treatment of disorders involving the male sex hormones (androgens), including primary and secondary hypogonadism, delayed or precocious puberty, impotence in males and, in females hirsutism (excessive hair) and virilization (masculinization) due to tumors, polycystic ovaries, and adrenogenital syndromes.
Predicate Devices
EiAsy Free Testosterone EIA (k030730)
Submission Summary (Full Text)
{0}
1
# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY ASSAY ONLY TEMPLATE
A. 510(k) Number:
k181017
B. Purpose for Submission:
New device
C. Measurand:
Testosterone (free)
D. Type of Test:
Quantitative, Enzyme Immunoassay (EIA)
E. Applicant:
Monobind, Inc.
F. Proprietary and Established Names:
Free Testosterone AccuBind ELISA Test System
G. Regulatory Information:
| Product Code | Classification | Regulation Section | Panel |
| --- | --- | --- | --- |
| CDZ | Class I, reserved | 21 CFR 862.1680
Testosterone test system | Clinical Chemistry (75) |
H. Intended Use:
1. Intended use(s):
See Indication(s) for use below.
{1}
2. Indication(s) for use:
The Free Testosterone AccuBind ELISA Test System is an Enzyme Immunoassay (EIA) for the quantitative measurement of free testosterone in human serum. Measurement of free testosterone is used in the diagnosis and treatment of disorders involving the male sex hormones (androgens), including primary and secondary hypogonadism, impotence in males and in females; hirsutism (excessive hair) and virilization (masculination) due to tumors, polycystic ovaries and adrogenital syndromes.
3. Special conditions for use statement(s):
For Prescription Use only.
4. Special instrument requirements:
Microplate Reader with 450nm and 620nm wavelength absorbance capability.
I. Device Description:
The Free Testosterone AccuBind ELISA Test System consists of the following:
- Seven vials of 1 mL serum reference calibrators for Free Testosterone containing testosterone in human serum with preservative at concentrations of 0, 0.2, 1.0, 2.5, 7.5, 20, and 60 pg/mL.
- Three vials of 1 mL controls (one low, one medium, and one high level) containing free testosterone in human serum with preservative.
- Free Testosterone Enzyme Reagent: One vial of 6 mL Testosterone (Analog)-horseradish peroxidase (HRP) conjugate in a protein stabilizing matrix.
- Substrate A: one vial of tetramethylbenzidine (TMB) in buffer.
- Substrate B: one vial of hydrogen peroxide in buffer.
- One 96-well microplate coated with testosterone antibody and packaged in an aluminum bag with a drying agent.
- One vial containing hydrogen peroxide in buffer.
- One vial of 20 mL concentrated wash solution.
- One vial of 8 mL stop reaction solution.
J. Substantial Equivalence Information:
1. Predicate device name(s):
EiAsy Free Testosterone EIA
2. Predicate 510(k) number(s):
k030730
{2}
3
3. Comparison with predicate
Similarities and Differences
| Items | Candidate Device
Free Testosterone AccuBind
ELISA Test System
(k181017) | Predicate Device
EiAsy Free Testosterone EIA
(k030730) |
| --- | --- | --- |
| Intended Use | The direct quantitative determination of free testosterone by enzyme immunoassay in human serum. | Same |
| Antibody | Utilizes a highly specific rabbit polyclonal antibody at a low binding capacity. | Same |
| Sample Type | Human serum | Same |
| Test Principle | Competitive Enzyme Immunoassay | Same |
| Detection Instrument | Microplate Colorimeter Reader | Same |
| Microplate coating | Antibody coated microwell plate | Same |
| Calibrators | Seven vials containing testosterone in human serum with preservative. | Six vials containing testosterone in human serum with preservative. |
| Controls | Three vials containing testosterone in human serum. | Two vials containing testosterone in human serum. |
| Measuring range | 0.11 - 60 pg/mL | 0.018 - 60 pg/mL |
K. Standard/Guidance Document Referenced (if applicable):
CLSI EP05-A3 Evaluation of Precision Performance of Quantitative Measurement Procedures (October 2014).
CLSI EP06-A Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach (April 2003).
CLSI EP17-A2 Evaluation of Detection Capacity for Clinical Laboratory Measurements Procedures (June 2012).
CLSI EP28-A3C Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory (October 2010).
CLSI EP07-A2 Interference Testing in Clinical Chemistry (November 2005).
CLSI EP25-A Evaluation of Stability of In-Vitro Diagnostic Reagents (September 2009).
{3}
4
L. Test Principle:
The Free Testosterone AccuBind ELISA test system uses a competitive enzyme immunoassay technology. The essential reagent includes an immobilized antibody, enzyme-antigen conjugate and a native antigen. Upon mixing the immobilized antibody, enzyme-antigen conjugate and a serum containing the free native antigen, a competitive reaction results between the native free antigen and the enzyme-antigen conjugate for a limited number of insolubilized binding sites. After equilibrium is attained, the antibody-coated fraction is separated from unbound antigen by decantation or aspiration. By utilizing several different serum references of known antigen concentration, a dose response curve can be generated from which the antigen concentration of an unknown sample can be ascertained. The serum calibrators are prepared in human serum matrix. The enzyme-antigen conjugate is labelled with horseradish peroxidase (HRP) and the substrate reagent contains tetramethylbenzidine (TMB), a blue color is produced. The reaction is stopped with addition of an acid and a yellow color is developed. The plate is read in a microtiter plate reader at 450nm.
M. Performance Characteristics (if/when applicable):
1. Analytical performance:
a. Precision/Reproducibility:
A study was performed by testing three levels of human serum pools and three levels of control material using three reagent lots. The samples were tested in duplicate, two times a day for a period of 20 days for a total of 80 measurements per sample. Each of the three lots of reagents produced similar precision results. The precision results from one representative lot are summarized in the table below:
| | | Within-Run | | Total | |
| --- | --- | --- | --- | --- | --- |
| Sample | Mean (pg/mL) | SD | %CV | SD | %CV |
| Control 1 | 2.51 | 0.09 | 3.7% | 0.20 | 7.8% |
| Control 2 | 10.98 | 0.40 | 3.6% | 0.96 | 8.7% |
| Control 3 | 22.72 | 0.83 | 3.6% | 2.18 | 9.6% |
| Serum 1 | 0.98 | 0.06 | 5.9% | 0.12 | 12.4% |
| Serum 2 | 4.53 | 0.26 | 5.7% | 0.36 | 8.0% |
| Serum 3 | 53.62 | 4.24 | 7.9% | 4.32 | 8.1% |
b. Linearity/assay reportable range:
A study was performed to evaluate the linearity of the Free Testosterone AccuBind ELISA Test System. Test samples were prepared by performing serial dilutions of a high human serum free testosterone pool with a low human serum free testosterone pool. The ten free testosterone concentrations tested were as follows: 0.11, 6.58, 13.05, 19.52, 25.99, 32.46, 45.40, 51.87, 58.34, and 64.81 pg/mL.
{4}
The samples were tested in replicates of 4. The following linearity regression equation was obtained:
$$
y = 1.0149x - 0.6028, R^2 = 0.9888
$$
The results of the linearity study support a measuring range of 0.11 - 60 pg/mL.
c. Traceability, Stability, Expected values (controls, calibrators, or methods):
The test system is traceable to a certified reference material, Cerrlian testosterone, T-037.
d. Detection limit:
Detection limit studies were performed in accordance with CLSI EP17 guideline.
A limit of blank (LoB) study was performed using three different blank samples that were measured using 3 reagent lots over 6 days to yield 144 measurements. The LoB was determined nonparametrically using the following equation:
$[\mathrm{N_B(p / 100)} + 0.5] =$ result at position $[0.95*\mathrm{N_B} = 0.5] = \mathrm{P}(1 - \alpha)$. The LoB was determined to be largest result of the 3 reagent lots, $0.0295~\mathrm{pg / mL}$.
The limit of detection (LoD) study was performed by obtaining 10 measurements of 10 low-level samples across 3 different reagent lots and 2 analyzers. LoD was calculated using the following equation: $\mathrm{LoD = LoB + Cv \times SDs}$. The LoD was determined to be $0.0519~\mathrm{pg / mL}$.
Limit of Quantitation (LoQ): The test results from the LoD study were used to calculate total error. The goal for the total error was set as $0.05~\mathrm{pg / mL}$. The TE was lower than $0.05~\mathrm{pg / mL}$, therefore the sponsor claims that the LoQ is equal to the LoD, $0.0519~\mathrm{pg / mL}$.
| Limit of Blank | Limit of Detection | Limit of Quantitation |
| --- | --- | --- |
| 0.0295 pg/mL | 0.0519 pg/mL | 0.0519 pg/mL |
The sponsor's claimed measuring range is 0.11 - 60 pg/mL.
e. Analytical specificity:
Interference:
An interference study was performed following CLSI EP07-A2 guideline. Aliquots from pools of human serum with a free testosterone concentration of $7.916~\mathrm{pg / mL}$ and $38.5~\mathrm{pg / mL}$ were spiked with potentially interfering substances at one or more concentrations. The sponsor defines significant interference as $>10\%$ bias.
{5}
The results are summarized in the following table:
| Substance | Highest concentration at which no significant interference was observed |
| --- | --- |
| Acetaminophen | 20 mg/dL |
| Acetylcysteine | 150 mg/dL |
| Ascorbic Acid | 6 mg/dL |
| Bilirubin, conjugated | 15 mg/dL |
| Bilirubin, unconjugated | 20 mg/dL |
| Biotin | 100 ng/mL |
| Caffeine | 6 mg/dL |
| Cholesterol | 503 mg/dL |
| Creatine | 30 mg/dL |
| Dextran | 5000 mg/dL |
| Digoxin | 6.1 ng/ mL |
| Doxycycline | 50 mg/dL |
| Erythromycin | 6 mg/dL |
| Gentamicin | 1 mg/dL |
| Human mouse antibodies | 440 ng/mL |
| Hemoglobin | 500 mg/dL |
| Heparin | 3 U/mL |
| Human serum albumin | 2.5 g/dL |
| Ibuprofen | 50 mg/dL |
| Immunoglobulin G | 4 g/dL |
| Levodopa | 20 mg/L |
| Lidocaine | 1.2 mg/dL |
| Lipemia (glycerides) | 1000 mg/dL |
| Methyldopa | 20 mg/dL |
| Nicotine | 0.1 mg/dL |
| Phenobarbital | 15 mg/dL |
| Protein, total | 10.5 g/dL |
| Rheumatoid factor | 1110 IU/mL |
| Salicylic Acid | 60 mg/dL |
| Sex hormone binding globulin | 200 μg/mL |
| Triglycerides | 900 mg/dL |
| Urea | 500 mg/dL |
Cross-Reactivity:
A cross-reactivity study was performed in accordance with CLSI EP07-A2 to evaluate whether various analytes or substances cross-react with the quantitation of free testosterone using the device. Aliquots from a pool of human serum with a free testosterone concentration of 7.408 pg/mL were spiked with substances at the concentrations listed in the table below. Cross-reactivity was determined using the following equation: (observed value - unspiked value)/ concentration of cross-reactant x 100%. The sponsor defines significant cross-reactivity as >10% difference.
{6}
| Substance | Concentration. of substance (ng/mL) | % Cross Reactivity |
| --- | --- | --- |
| 11-Deoxycortisol | 1000 | 0.000% |
| 11-KetoTestosterone | 10 | 0.647% |
| 11β-Hydroxytestosterone | 100 | 0.065% |
| 17α-ethynyl estradiol | 1000 | 0.000% |
| 17α-Estradiol | 1000 | 0.000% |
| 17β-Estradiol | 100 | 0.000% |
| 17-Hydroxypregnenolone | 1000 | 0.000% |
| 17-Hydroxprogesterone | 10 | 0.000% |
| 3-EstriolGluc | 1000 | 0.000% |
| 3-EstriolSul | 1000 | 0.000% |
| 3β-Androstanediol | 500 | 0.000% |
| 5α-Dihydrotestosterone | 100 | 0.054% |
| Aldosterone | 8000 | 0.000% |
| Amitriptyl HCI | 1000 | 0.000% |
| Androsterone | 1000 | 0.000% |
| Andronstenedione | 1000 | 0.004% |
| Clomiphene Citrate | 1000 | 0.000% |
| Corticosterone | 1000 | 0.000% |
| Cortisone | 1000 | 0.000% |
| Cortisol | 1000 | 0.000% |
| Cyproterone acetate | 1000 | 0.000% |
| D-5-Androstene-3β,17β-diol | 1000 | 0.000% |
| Danazol | 1000 | 0.000% |
| Dehydroepiandrosterone | 100000 | 0.000% |
| Dehydroepiandrosterone Sulfate | 1000 | 0.000% |
| Desogestrel | 100 | 0.000% |
| Dexamethasone | 1000 | 0.000% |
| Epitestosterone | 1000 | 0.001% |
| Estriol | 1000 | 0.000% |
| Estrone | 1000 | 0.000% |
| Ethisterone | 1000 | 0.000% |
| Ethynediol | 1000 | 0.000% |
| Ethynediol diacetate | 50 | 0.000% |
| Flunisolide | 1000 | 0.000% |
| Fluoxymesterone | 1000 | 0.000% |
| Lynestrol | 1000 | 0.000% |
| Medroxyprogesterone acetate | 1000 | 0.000% |
| Methyl Testosterone | 100 | 0.000% |
| Mestranol | 1000 | 0.000% |
| Norethindrone | 50 | 0.000% |
| Norethindrone acetate | 50 | 0.000% |
| Norgestimate | 1000 | 0.000% |
{7}
| Substance | Concentration of substance (ng/mL) | % Cross Reactivity |
| --- | --- | --- |
| Norgestrel (Levonorgestrel) | 50 | 0.000% |
| Norethynodrel | 50 | 0.000% |
| Oxymetholone | 100 | 0.000% |
| Prednisolone | 1000 | 0.000% |
| Prednisone | 800 | 0.000% |
| Progesterone | 1000 | 0.000% |
| Salbutamol | 1000 | 0.000% |
| Spironolactone | 1000 | 0.000% |
| Stanozolol | 1000 | 0.000% |
| Testosterone enanthate | 10 | 0.000% |
| Testosterone SO4 | 1000 | 0.004% |
| Testosterone Propionate | 1000 | 0.000% |
| Triamcinolone | 50 | 0.000% |
An additional study was performed to evaluate the cross-reactivity effects of testosterone cypionate and testosterone undecanoate. Aliquots from pool of human serum with a free testosterone concentration of 38.4 pg/mL were spiked with 12 ng/mL of testosterone cypionate and testosterone undecanoate. Cross-reactivity was determined using the following equation: observed value - unspiked value/concentration of cross-reactant x100%. The sponsor defines significant cross-reactivity as >10% difference. The results are summarized in the chart below:
| Substance | Concentration of Substance (ng/mL) | % Cross Reactivity |
| --- | --- | --- |
| Testosterone cypionate | 12 | 0.000% |
| Testosterone undecanoate | 12 | 0.000% |
f. Assay cut-off:
Not applicable.
2. Comparison studies:
a. Method comparison with predicate device:
A method comparison study was performed by testing 137 (133 native and 4 spiked) human serum samples with the candidate and predicate devices. The test results on the candidate device ranged from 0.11-59.63 pg/mL. The following regression equation was obtained using Passing-Bablok analysis: y= 1.017x - 0.244, r= 0.997
b. Matrix comparison:
Not applicable. Serum is the only claimed sample type.
{8}
9
3. Clinical studies:
a. Clinical Sensitivity:
Not applicable.
b. Clinical specificity:
Not applicable.
c. Other clinical supportive data (when a. and b. are not applicable):
None.
4. Clinical cut-off:
Not applicable.
5. Expected values/Reference range:
A reference range study was performed by testing a total of 261 serum samples obtained from 130 females (aged 22-93 years), and 131 males (aged 22-89 years) on the candidate device. The subjects were not pregnant, not on hormone therapy, not taking contraceptives, glucocorticoids or corticosteroids; and had no history of thyroid, autoimmune, Cushing's, or Addison's disease. The resulting reference interval is summarized in the following table:
| Cohort | N | 95% Confidence Range (pg/ml) |
| --- | --- | --- |
| Male, 20-39 yrs. | 45 | 9.2-34.6 |
| Male, 40-59 yrs. | 43 | 6.1-30.3 |
| Male, ≥60 yrs. | 43 | 6.1-27.9 |
| Female, 20-39 yrs. | 44 | 0.2-6.1 |
| Female, 40-59 yrs. | 42 | 0.3-4.4 |
| Female, ≥60 yrs. | 44 | 0.5-3.4 |
N. Proposed Labeling:
The labeling is sufficient and it satisfies the requirements of 21 CFR Parts 801 and 809, as applicable.
O. Conclusion:
The submitted information in this premarket notification is complete and supports a substantial equivalence decision.
Two short videos show you everything — or skip straight to the written tutorial if you'd rather read. You can reopen this any time from the Tutorial button in the top bar.
Part 1 — Search, results, and everyday workflows 16 min
Part 2 — Embeddings: the galaxy map 3 min
1. Search: exact and fuzzy
Type a phrase like "coronary artery calcification" into the search box. You get two kinds of results. Exact results match the literal phrase — prefix searches work ("coronary artery calcificati") but suffix searches do not. Fuzzy results match on the meaning and intent of your phrase rather than the exact words, and are sorted by relevance score. Hover over the Exact or Fuzzy badge on any row to see exactly why it matched.
Use the checkboxes above the results to narrow: SaMD keeps only software-only devices, AI / ML keeps only devices with AI.
Exact vs. fuzzy search: what's the difference?
Exact matches on the literal phrase (prefix search works, suffix does not). Fuzzy matches on the meaning and intent of the phrase rather than the exact words. Hover over the badge on any row to see why it matched.
You search "coronary artery calcification" and want only software devices with AI. What two filters do you apply?
Narrow by SaMD (software-only devices), then narrow by AI/ML (devices with AI).
2. The results table
Scroll right in the results table. The intended use is extracted for you — no need to open the PDF. The device story gives a high-level snapshot of what the device does and how it's used. The AI Performance sub-table shows each output name, acceptance criteria, observed values, and development/test dataset descriptions — the same format Innolitics uses for regulatory strategy outputs, and the fastest high-level fingerprint of an AI device. It is AI-generated but has been very reliable in practice.
Where do you find a device's intended use without opening the PDF?
Scroll right in the search results table. The intended use column is extracted for you; no need to dig into the 510(k) summary PDF.
What does the AI Performance sub-table show, and why is it useful?
Output name, acceptance criteria, observed values, development dataset description, and test dataset description. It's the same format we use for regulatory strategy output and Fast 510(k) input, and the fastest high-level fingerprint of an AI device. AI-generated but reliable in practice.
3. Judging fuzzy relevance
Fuzzy results trail off in relevance as you scroll. Use three signals to decide how far down to go: the fuzzy badge explanations, the intended use column, and whether your target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, you're past the relevant zone. A top hit with a low score (~0.4) and a stretched explanation is a hint the closest predicates are far away — the project may be headed for De Novo. Note the fuzzy search is a pattern match: it doesn't handle negation ("not") well, and hardware devices can appear — filter by SaMD/AI ML to cut them.
How do you judge how far down fuzzy search results to go?
Use the relevancy signals: the fuzzy badge explanations, the intended use column, and whether the target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, results are trailing off in relevancy.
4. Device detail page: chat and citations
Click a device name to open its detail page: device facts on the left, a chat window on the right. Ask something like "Describe the training data". The answer carries little citation bubbles — click one to jump to the highlighted passage in the source PDF, so you can verify every AI answer against the document. There's also a Download PDF button for sharing.
How do you verify an AI chat answer on the device detail page?
Click the citation bubbles to jump to the relevant highlight in the source document.
Reading rule for every project: how many summaries do you read in full?
At least the three most relevant 510(k) or De Novo summaries, in full. After that, use targeted chat questions to confirm your memory quickly. The tool supports this professional habit — it doesn't replace it.
5. Side-by-side comparison
Select multiple rows in the results table (aim for under ~10), then open the PDF Viewer tab. Ask one question — it goes to all selected devices in parallel, each with citations. This is the fastest way to compare and contrast devices: training data, PCCP scope, how they handled adding new scanners, and so on.
What does the side-by-side PDF viewer mode do?
Select multiple devices, open the PDF viewer tab, and ask one question (e.g., "Describe the training data"). It queries all selected devices simultaneously with citations, so you can compare and contrast quickly.
6. Collections
With rows selected, go to the Collections tab and create a labeled collection (e.g., "Cobb Angle Project"). Reload that selection any time — before a client call, pull up the collection and ask questions across all of its devices at once.
How do you save a set of selected devices for later use?
Select the rows, go to the Collections tab, and create a labeled collection (e.g., "Cobb Angle Project"). You can reload the selection anytime and carry it into the PDF viewer and other tabs that support selections.
7. Product codes and the regulations tree
Click a product code in the results to jump to it in the regulations tree — identification text, sibling product codes, and devices you can open in a PDF viewer on the right. Click a regulation number to see its identification, special controls, and related product codes. You can also search by product code or regulation number at the top of the tree. Always read the special controls if any exist for your device — it broadens your search and sharpens pre-kickoff research.
What can you do from the regulations tree view?
Browse product codes and regulation numbers, read the identification text and special controls, browse sibling product codes, open device PDFs on the right, and search by product code or regulation number at the top of the tree.
8. Chart view
Click Show Chart and segment by regulation number (or product code) to see which regulations dominate your result set. Clicking a regulation takes you into the regulations tree. Great for spotting that most matches are, say, hardware laparoscopic devices — a cue to go back and filter.
How do you see which regulations dominate a search result set?
Click "Show Chart" and segment by Regulation Number. Clicking a regulation takes you to the regulations tree.
9. The predicate graph
Open the Predicates tab for a family-tree view of predicate relationships. Click a node to trace its parents and children; selections from search carry over pre-selected. Commonly predicated devices are worth reading — a lot of people predicated them for a reason. The visual lineage is also handy on client calls, e.g. to show how a predicate family evolved and justify why your predicate still holds.
In the predicate graph, why are commonly predicated devices worth reading?
A lot of people predicated them for a reason. Clicking a node traces parents and children, and selections from search carry over pre-selected.
10. Embeddings: the galaxy map
The Embeddings tab plots every matching document in a 2-D "galaxy map" where semantically similar devices cluster together. Hover or click clusters to explore, and let AI label the clusters for you. Embeddings beat product codes for grouping: two devices can carry different product codes (LLZ vs. QIH) yet do the same thing — the embedding captures the meaning of the intended use and device story. This is also exactly how retrieval-augmented generation (RAG) works under the hood, and it makes a great visual on client calls.
Try it yourself
Head to the search page and work through a few of these AI/ML fuzzy searches to build intuition: perivascular fat on CT · aortic valve calcification opportunistic screening on noncontrast CT · breast cancer prediction on digital pathology slides · autism detection · gestational age prediction · a hearing aid that can also detect a pulse · foundation model based analysis of ECG · large language models · penetration test. Watch how the relevance scores, intended use, and AI Performance tables tell you when results stop being meaningful.