For in vitro diagnostic use in the quantitative determination of glucose in human serum, plasma, urine, and cerebrospinal fluid (CSF) on the ADVIA 1650 Chemistry System. Such measurements are used in the diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, neonatal hypoglycemia, idiopathic hypoglycemia, and insulin overdose.
Device Story
ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) is a two-component enzymatic reagent system for the ADVIA 1650 Chemistry System. It measures glucose in human serum, plasma (Li-heparin, K-EDTA), urine, and CSF. Principle: Hexokinase phosphorylates glucose with ATP; glucose-6-phosphate dehydrogenase oxidizes the product, reducing NAD to NADH. The system measures NADH absorbance at 340/410 nm as an endpoint reaction. The difference in absorbance between Reagent 1 (buffer, ATP, NAD) and Reagent 2 (enzymes, buffer, ATP, NAD) is proportional to glucose concentration. Used in clinical laboratories by trained personnel. Output is a quantitative glucose concentration (mg/dL). Results assist clinicians in diagnosing and managing carbohydrate metabolism disorders. The device includes automated dilution for samples exceeding the upper reportable range.
Clinical Evidence
Bench testing only. Precision studies (N=80 per serum/urine control, N=40 per CSF control) showed CVs ranging from 0.3% to 2.4%. Linearity confirmed across 3-833 mg/dL range. Method comparison against predicate (N=51-113 samples per matrix) showed high correlation (r=1.000) and slopes near 1.0. Matrix comparison for K-EDTA and Li-Heparin plasma confirmed equivalence. No clinical data required.
Indicated for quantitative glucose measurement in human serum, plasma, urine, and CSF for diagnosis and treatment of carbohydrate metabolism disorders (diabetes mellitus, neonatal/idiopathic hypoglycemia, insulin overdose).
Regulatory Classification
Identification
A glucose test system is a device intended to measure glucose quantitatively in blood and other body fluids. Glucose measurements are used in the diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, neonatal hypoglycemia, and idiopathic hypoglycemia, and of pancreatic islet cell carcinoma.
Special Controls
*Classification.* Class II (special controls). The device, when it is solely intended for use as a drink to test glucose tolerance, is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 862.9.
Predicate Devices
ADVIA Chemistry Glucose Hexokinase II reagent (k042015)
Submission Summary (Full Text)
{0}
1
510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION
DECISION SUMMARY
ASSAY ONLY TEMPLATE
A. 510(k) Number:
k101854
B. Purpose for Submission:
Change in reagent formulation to a previously cleared device (k042015) and adding another anticoagulant (K-EDTA plasma) as a sample type.
C. Measurand:
Glucose
D. Type of Test:
Quantitative enzymatic test
E. Applicant:
Siemens Healthcare Diagnostics Inc.
F. Proprietary and Established Names:
ADVIA Chemistry Glucose Hexokinase (GLUH_3) Reagent
G. Regulatory Information:
1. Regulation section:
21 CFR 862.1345 Glucose Test System
2. Classification:
Class II
3. Product code:
CFR - Hexokinase, Glucose
{1}
4. Panel:
Clinical Chemistry - 75
H. Intended Use:
1. Intended use(s):
Refer to indications for use, below.
2. Indication(s) for use:
For in vitro diagnostic use in the quantitative determination of glucose in human serum, plasma, urine, and cerebrospinal fluid (CSF) on the ADVIA 1650 Chemistry System. Such measurements are used in the diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, neonatal hypoglycemia, idiopathic hypoglycemia, and insulin overdose.
3. Special conditions for use statement(s):
For in vitro diagnostic use only. For prescription use only.
4. Special instrument requirements:
ADVIA 1650 Chemistry System
I. Device Description:
The ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) device consists of two-component reagents. Reagent 1 contains buffer, ATP, and NAD. Reagent 2 contains glucose-6-phosphate dehydrogenase, hexokinase, buffer, ATP, and NAD. Reagents are ready to use and require no additional preparation. Reagents 1 and 2 contain sodium azide.
J. Substantial Equivalence Information:
1. Predicate device name:
ADVIA Chemistry Glucose Hexokinase II Reagent
2. Predicate 510(k) number:
k042015
3. Comparison with predicate:
2
{2}
| Comparison Table | | |
| --- | --- | --- |
| Item | New Device (k101854) | Predicate (k042015) |
| Intended Use | For in vitro diagnostic use in the quantitative determination of glucose in human serum, plasma, urine, and CSF on the ADVIA 1650 Chemistry System. | same |
| System Principle | Enzymatic method based on use of hexokinase and glucose-6-phosphate dehydrogenase enzymes. | same |
| Detection method | Photometric – Hexokinase | same |
| Calibration | Siemens calibrators | same |
| Measurement Range | 4-700 mg/dL | 0-700 mg/dL |
| Specimen Type | Human serum, Li-heparin plasma, K-EDTA plasma, urine, and CSF | Human serum, Li-heparin plasma, urine, and CSF |
| Format | Liquid | Concentrate |
| Reagents | R1 and R2 | R1, R2, R2 mix |
| Storage Conditions | On-board: 60 days at 2-8°C
Shelf-life: 18 months at 2-8°C | same |
K. Standard/Guidance Document Referenced (if applicable):
1. CLSI EP5-A2. Evaluation of Precision Performance of Quantitative Measurement Methods; Approved Guideline.
2. CLSI EP17-A. Protocols for Determination of Limits of Detection and Limits of Quantitation; Approved Guideline.
L. Test Principle:
This enzymatic method is based on the method by Slein using hexokinase and glucose-6-phosphate dehydrogenase enzymes. The ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) method is a two-component reagent. Sample is added to Reagent 1, which contains the buffer, ATP, and NAD. Absorbance readings of the sample in Reagent 1 are taken and are used to correct for interfering substances in the sample. Reagent 2 is added, which initiates the reaction. Glucose is phosphorylated by adenosine triphosphate (ATP) in the presence of hexokinase. The glucose-6-phosphate that forms is oxidized in the presence of glucose-6-phosphate dehydrogenase causing the reduction of NAD to NADH. The absorbance of NADH is measured as an endpoint reaction at $340/410\mathrm{nm}$. The difference between the absorbance in Reagent 1 and Reagent 2 is proportional to the glucose concentration.
M. Performance Characteristics (if/when applicable):
1. Analytical performance:
{3}
a. Precision/Reproducibility:
Precision studies were performed using commercial quality controls and native samples. Studies were carried out in duplicate, twice per day, over a period of 10 to 20 days using two ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent lots on two different ADVIA 1650 Chemistry Systems. A low level urine sample was assayed in duplicate on one system over a period of nine days. The mean, standard deviation (SD), and coefficients of variation (CV) were determined at each level as summarized below:
| Precision Data Summary | | | | | | |
| --- | --- | --- | --- | --- | --- | --- |
| Sample | N | Mean (mg/dL) | Within Run | | Total | |
| | | | SD (mg/dL) | CV (%) | SD (mg/dL) | CV (%) |
| Serum Control 1 | 80 | 87 | 0.39 | 0.5 | 0.66 | 0.8 |
| Serum Control 2 | 80 | 297 | 1.35 | 0.5 | 2.02 | 0.7 |
| Serum Pool 1 | 80 | 51 | 0.18 | 0.4 | 0.46 | 0.9 |
| Serum Pool 2 | 80 | 605 | 3.55 | 0.6 | 6.52 | 1.1 |
| Urine Control 1 | 80 | 18 | 0.12 | 0.7 | 0.21 | 1.2 |
| Urine Control 2 | 80 | 301 | 1.29 | 0.4 | 5.63 | 1.9 |
| Urine Pool 1 | 80 | 23 | 0.13 | 0.6 | 0.26 | 1.2 |
| Urine Pool 2 | 80 | 49 | 0.27 | 0.6 | 0.51 | 1.1 |
| Urine Sample | 57 | 7 | 0.14 | 2.2 | 0.15 | 2.4 |
| CSF Control 1 | 40 | 56 | 0.24 | 0.4 | 0.54 | 1.0 |
| CSF Control 2 | 40 | 97 | 0.33 | 0.3 | 0.81 | 0.8 |
| CSF Control 3 | 40 | 619 | 1.81 | 0.3 | 4.97 | 0.8 |
b. Linearity/assay reportable range:
Two linearity studies were carried out using eleven different levels of human urine samples and human serum samples. Low pools consisted of a normal human urine sample and a glucose-depleted normal human serum sample. High pools consisted of low pools spiked with glucose to concentrations of approximately 800 mg/dL. The glucose concentrations ranged from 3.5-796 mg/dL. In addition, a linearity study was performed using a CSF pool of human samples at eleven different levels covering the glucose range from 3.0-833 mg/dL. Each sample was measured in duplicate using one ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent lot on one ADVIA 1650 Chemistry System. Linearity was evaluated by least-square regression analysis and comparison of observed recovery versus expected values.
{4}
| Linearity Data Summary | | | | |
| --- | --- | --- | --- | --- |
| Serum | | | | |
| Linear Regression Data | Recovery Data (mg/dL) | | | |
| | | Expected | Observed | % Difference |
| Number of samples (n) | 11 | 3.5 | 3.5 | 0.0% |
| Range (mg/dL) | 4-791 | 28.1 | 27.0 | -4.1% |
| Slope | 1.005 (0.998-1.012) | 52.7 | 50.1 | -5.0% |
| Intercept | -2.336 (-5.316-0.644) | 101.9 | 101.7 | -0.2% |
| Correlation coefficient | 1.000 | 200.4 | 196.9 | -1.7% |
| | 298.8 | 295.9 | -1.0% | |
| | | 397.2 | 397.2 | 0.0% |
| | | 495.7 | 492.7 | -0.6% |
| | | 594.1 | 593.0 | -0.2% |
| | | 692.6 | 691.3 | -0.2% |
| | | 791.0 | 798.2 | 0.9% |
| Urine | | | | |
| Linear Regression Data | Recovery Data (mg/dL) | | | |
| | | Expected | Observed | % Difference |
| Number of samples (n) | 11 | 4.2 | 4.2 | 0.0% |
| Range (mg/dL) | 4-796 | 29.0 | 29.4 | 1.5% |
| Slope | 0.995 (0.993-0.998) | 53.7 | 55.0 | 2.4% |
| Intercept | 1.146 (0.198-2.094) | 103.2 | 104.3 | 1.1% |
| Correlation coefficient | 1.000 | 202.2 | 202.3 | 0.0% |
| | 301.2 | 301.6 | 0.1% | |
| | | 400.2 | 400.2 | 0.0% |
| | | 499.2 | 498.3 | -0.2% |
| | | 598.3 | 596.6 | -0.3% |
| | | 697.3 | 696.1 | -0.2% |
| | | 796.3 | 792.0 | -0.5% |
| Cerebro Spinal Fluid (CSF) | | | | |
| Linear Regression Data | Recovery Data (mg/dL) | | | |
| | | Expected | Observed | % Difference |
| Number of samples (n) | 11 | 3.0 | 3.0 | 0.0% |
| Range (mg/dL) | 3-833 | 28.9 | 28.5 | -1.5% |
| Slope | 0.994 (0.989-0.999) | 54.9 | 54.0 | -1.6% |
| Intercept | 0.1937 (-2.000-2.3874) | 106.8 | 106.6 | -0.1% |
| Correlation coefficient | 1.000 | 210.6 | 209.8 | -0.4% |
| | 314.4 | 314.3 | 0.0% | |
| | | 418.2 | 418.2 | 0.0% |
| | | 522.1 | 519.7 | -0.4% |
| | | 625.9 | 617.5 | -1.3% |
| | | 729.7 | 727.6 | -0.3% |
| | | 833.5 | 829.7 | -0.5% |
{5}
The claimed reportable range for the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent on the ADVIA 1650 Chemistry System is 4 to 700 mg/dL. Data was provided to demonstrate that samples resulting in glucose values above the upper range limit will be automatically diluted and reported after applying dilution correction factor.
c. Traceability, Stability, Expected values (controls, calibrators, or methods):
## Traceability
The ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) method is traceable to a standard reference serum (SRM 965a) glucose material from the National Institute of Standards and Technology (NIST). A second set of standard reference material, a serum-based sample, a spiked serum sample, and an aqueous glucose sample are analyzed on the ADVIA 1650 Chemistry System and recovered values must fall within acceptance criteria based on the target value of the NIST standard.
## Stability
The stability of the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent was evaluated with real-time stability studies. The reagent shelf life claim is 18 months when stored at the recommended temperature of 2 to 8°C. An on-board stability study confirmed that the reagent is stable for 60 days on-board the ADVIA 1650 Chemistry System at the recommended temperature of 2 to 8°C. The manufacturer does not recommend removing open reagents from the system.
## Calibrators and Controls
A calibration frequency study with the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) method showed that the system should be calibrated with the Siemens Chemistry Calibrator at a frequency of at least every 60 days. The performance of the calibrator for the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) method has been demonstrated in a previous 510(k) submission. The manufacturer states in the labeling that individual laboratory quality control programs and procedures may require more frequent calibration. The manufacturer also recommends the use of a commercial control material with at least two glucose levels tested according to workflow, system experience, and government regulation.
d. Detection limit:
The Limit of Blank (LoB) and Limit of Detection (LoD) for the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) method were evaluated based on CLSI EP17-A guidelines. Studies were performed using N ≤ 2 reagent lots on one ADVIA 1650 Chemistry System.
6
{6}
7
| Detection Limit Data Summary | | |
| --- | --- | --- |
| | Limit | Protocol |
| Serum | LoB | A human serum glucose-depleted sample pool was evaluated. A total of 6 runs (10 replicates per run) were completed, N=120, over a period of three days. |
| | LoD | A low serum sample pool with a glucose concentration of 3-4 mg/dL was evaluated in duplicate. A total of 58 runs were over a period of nine days. |
| Urine | LoB | A normal human urine pool was evaluated. A total of 6 runs (10 replicates per run) were completed, N=120, over a period of nine days. |
| | LoD | A low urine sample with a glucose concentration of 6 mg/dL was evaluated in duplicate. A total of 57 runs were completed over a period of nine days. |
| CSF | LoB | A human CSF glucose-depleted sample pool was evaluated. A total of 3 runs (10 replicates per run) were completed, N=60, over a period of three days. |
| | LoD | A low CSF sample pool with a glucose concentration of 1.5 mg/dL was evaluated. A total of 3 runs (10 replicates per run) were completed, N=60, over a period of three days. |
| Claimed Values | | |
| LoB | 0.30 mg/dL | |
| LoD | 4.0 mg/dL | |
e. Analytical specificity:
Interferences studies were performed by adding potentially interfering substances to samples and calculating bias relative to a control (no interferent). Five equally diluted samples using either human serum or human urine pools were tested by adding different interferent concentrations. Samples were measured in duplicate using one ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent lot on one ADVIA 1650 Chemistry System. No significant interference defined as no more than $\pm 10\%$ was observed up to the levels shown in the table below for the following interfering substances:
| Interferences Data Summary | | | |
| --- | --- | --- | --- |
| | No Interference at Listed Level (mg/dL) | | |
| Substance | Low Serum | Mid Serum | High Serum |
| | ~50 mg/dL | ~130 mg/dL | ~200 mg/dL |
| Hemoglobin | 1000 | 1000 | 1000 |
| Bilirubin (Unconjugated) | 30 | 30 | 30 |
| Bilirubin (Conjugated) | 30 | 30 | 30 |
| Triglycerides (Intralipid) | 1000 | 1000 | 1000 |
{7}
8
| Interferences Data Summary | | |
| --- | --- | --- |
| No Interference at Listed Level (mg/dL) | | |
| Substance | Low Urine | High Urine |
| | ~30 mg/dL | ~100 mg/dL |
| Ascorbic Acid | 200 | 200 |
| Caffeine | 50 | 50 |
| Creatinine | 500 | 500 |
| Salicylate (Na) | 50 | 50 |
| Urea Nitrogen | 500 | 500 |
f. Assay cut-off:
Not applicable
2. Comparison studies:
a. Method comparison with predicate device:
In a method comparison study, glucose was measured in serum, urine, and CSF samples by one or two trained operators on both, the predicate device and the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent on ADVIA 1650 Chemistry Systems. Samples were analyzed in duplicate using one lot of each reagent. A number of samples were altered to cover the claimed measuring range for the device. Data were analyzed by means of least-squares linear regression analysis. Study results are summarized below.
| Method Comparison Data Summary | | | |
| --- | --- | --- | --- |
| | Serum | Urine | CSF |
| Number of samples (n) | 99 | 51 | 113 |
| Range (mg/dL) | 5-676 | 4-664 | 20-654 |
| Slope (95% CI) | 1.001 (0.999-1.003) | 0.989 (0.987-0.990) | 1.005 (1.004-1.006) |
| Intercept (95% CI) | 0.264 (-0.171-0.698) | -0.316 (-0.648-0.015) | -0.097 (-0.313-0.120) |
| Correlation coefficient | 1.000 | 1.000 | 1.000 |
b. Matrix comparison:
A matrix comparison study was performed by testing K-EDTA and Li-Heparin samples on both the predicate device and the ADVIA Chemistry Glucose Hexokinase_3 (GLUH_3) reagent on ADVIA 1650 Chemistry Systems, by two trained operators. A number of samples were altered to cover the claimed measuring range for the device. Data were analyzed by means of least-squares linear regression analysis. Study results are summarized below.
{8}
| Matrix Comparison Data Summary | | |
| --- | --- | --- |
| | K-EDTA plasma | Li-Heparin plasma |
| Number of samples (n) | 87 | 88 |
| Range (mg/dL) | 6-676 | 5-686 |
| Slope (95% CI) | 1.002 (1.000-1.003) | 1.001 (0.999-1.00) |
| Intercept (95% CI) | 0.002 (-0.298-0.301) | 0.184 (-0.173-0.541) |
| Correlation coefficient | 1.000 | 1.000 |
# 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):
Not applicable
# 4. Clinical cut-off:
Not applicable
# 5. Expected values/Reference range:
Expected reference ranges for the method are as follows:
| Expected Values | | |
| --- | --- | --- |
| Sample Type | | Reference Range (mg/dL) |
| Serum/plasma | Adult | 74-106 |
| | Newborn 1 day | 40-60 |
| | Newborn > 1 day | 50-80 |
| | Child | 60-100 |
| Urine | | < 0.5 g/day |
| CSF | Adult | 40-70 |
| | Infant/Child | 60-80 |
Reference: Wu AHB. Tietz Clinical Guide to Laboratory Tests. $4^{\text{th}}$ ed. St. Louis, MO: WB Saunders Company: 2006: 444-450.
{9}
N. Proposed Labeling:
The labeling is sufficient and it satisfies the requirements of 21 CFR Part 809.10.
O. Conclusion:
The submitted information in this premarket notification is complete and supports a substantial equivalence decision.
10
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.