Skip to content

Scientific Evidence & Source Legend

This file documents the scientific and technical sources used to shape IINTS-AF defaults, validation targets, report framing, local AI setup guidance, and best-effort device emulation notes.

Scope: - Pre-clinical simulation and retrospective forecasting only. - Not a treatment recommendation engine. - Not a medical device.

Use iints sources to print the packaged source manifest from the installed SDK.

For the full public source index, including dataset registry sources and documentation-only implementation references, see Complete Source Library.

How To Read This Legend

There are two source buckets in this project:

  1. Packaged evidence sources These are the core medical and dataset references shipped with the SDK and exposed through iints sources.

  2. Documentation-only implementation sources These are the additional references used in the guides for:

  3. Ollama setup
  4. local open Mistral model selection
  5. best-effort device emulation context

That split is intentional: - the packaged manifest stays focused on repeatable research evidence - this page stays readable as the full legend for humans

Category Legend

Category Meaning
guideline formal standards or standards-of-care style guidance
consensus expert consensus used for interpretation targets
trial clinical or comparative outcome study
pharmacology insulin PK/PD reference
sensor CGM behavior, lag, or sensor interpretation source
model mathematical or simulator foundation source
dataset public dataset provenance source
runtime technical runtime or installation reference for local AI
model_card model-family reference for local Mistral choices
regulatory manufacturer or regulator-facing system reference
technical_manual user guide or product documentation
clinical_trial named trial registry or trial program page
physiology human physiology or organ-system mechanism source
review peer-reviewed review used to summarize established mechanisms

How Sources Map to SDK Components

SDK area Why it exists Source IDs
Validation targets (TIR, TBR, TAR, hypoglycemia framing) Keep benchmark interpretation aligned with accepted diabetes metrics ada_2026_glycemic_goals, attd_2019_time_in_range
CGM + AID context in docs/reports Keep language aligned with current standards for technology use ada_2026_diabetes_technology
AID benchmark envelopes Compare algorithm behavior against realistic closed-loop outcomes nejm_2019_control_iq, adapt_2022_ahcl
Meal timing / pre-bolus scenarios Avoid unrealistic meal-response assumptions cobry_2010_meal_bolus_timing
Insulin action profiles (rapid and ultra-rapid) Parameterize onset/peak assumptions from pharmacology literature heise_2017_fiasp_pkpd, klaff_2020_urli_pkpd
Input validation and CGM lag rationale Keep signal handling biologically plausible wentholt_2004_cgm_lag
Sensor long-memory noise model Add colored, persistent CGM noise without claiming exact vendor reproduction mandelbrot_1968_fractional_brownian
Virtual patient meal dynamics Ground meal disturbance dynamics in established models dalla_man_2007_meal_model
Bergman/Hovorka patient model foundations Ground internal glucose, insulin, and insulin-action state variables in established mathematical physiology bergman_1979_minimal_model, hovorka_2004_nmpc_t1d, dalla_man_2007_meal_model
Circadian, exercise, and incretin physiology Document experimental dawn, GLUT4/NIMGU, and GLP-1 gastric-emptying modifiers campbell_1985_dawn_phenomenon, richter_2013_glut4_exercise, naslund_1999_glp1_gastric_emptying
Hypoglycemia counterregulation and HAAF experimental layer Make the body's rescue mechanisms, repeated-low memory, and hypo-awareness assumptions explicit gerich_1979_counterregulation, cryer_2013_haaf_mechanisms, cryer_2013_haaf_diabetes
Bi-hormonal glucagon research Separate delivered glucagon, subcutaneous absorption, plasma appearance, and closed-loop dual-hormone context lv_2013_exogenous_glucagon_pk, haidar_2013_insulin_glucagon_pk, haidar_2013_dual_hormone_ap
Renal glucose clearance experimental layer Document high-glucose renal threshold/splay behavior for whole-day mass-balance realism hummel_2018_renal_glucose_handling, defronzo_2013_renal_reabsorption_splay
Simulator realism/validation framing Align in-silico model evaluation with accepted simulator literature visentin_2018_uvapadova, mujahid_2024_generative_t1d_simulator
Exercise stress scenarios Use consensus guidance for exercise-related glucose behavior riddell_2017_exercise_consensus
Forecast training data provenance Use publicly documented dataset references marling_2020_ohiot1dm
AGP-style research reports Present dense glucose traces with time-in-ranges, modal-day percentile bands, and daily profiles idc_2025_agp_report_overview, attd_2019_time_in_range
Local AI setup and model selection docs Keep Ollama and open Mistral setup instructions grounded in official docs ollama_linux_install, mistral_2025_ministral_3_announcement, mistral_2025_ministral_3_3b, mistral_2025_ministral_3_8b, mistral_2025_ministral_3_14b
Device emulation notes Provide best-effort references behind 780G / Control-IQ / Omnipod 5 approximations bergenstal_2020_780g, fda_k193510_780g, medtronic_780g_user_guide, brown_2019_control_iq_dtt, fda_k191289_control_iq, idcl_nct03563313, control_iq_user_guide, assert_omnipod_5, onset_omnipod_5, fda_k203467_omnipod5, omnipod5_user_guide

Packaged Medical And Dataset Sources

  1. ada_2026_glycemic_goals ADA Professional Practice Committee. Glycemic Goals and Hypoglycemia: Standards of Care in Diabetes—2026. DOI: 10.2337/dc26-S006

  2. ada_2026_diabetes_technology ADA Professional Practice Committee. Diabetes Technology: Standards of Care in Diabetes—2026. DOI: 10.2337/dc26-S007

  3. attd_2019_time_in_range Battelino T, Danne T, Bergenstal RM, et al. Clinical Targets for Continuous Glucose Monitoring Data Interpretation. Diabetes Care. 2019. DOI: 10.2337/dci19-0028

  4. nejm_2019_control_iq Brown SA, et al. Six-Month Randomized, Multicenter Trial of Closed-Loop Control in Type 1 Diabetes. N Engl J Med. 2019. DOI: 10.1056/NEJMoa1907863

  5. adapt_2022_ahcl Benhamou PY, et al. Advanced hybrid closed loop therapy versus conventional treatment in adults with type 1 diabetes (ADAPT). Lancet Diabetes Endocrinol. 2022. DOI: 10.1016/S2213-8587(22)00212-1

  6. cobry_2010_meal_bolus_timing Cobry E, et al. Timing of Meal Insulin Boluses to Achieve Optimal Postprandial Glycemic Control. J Diabetes Sci Technol. 2010. DOI: 10.1177/193229681000400404

  7. heise_2017_fiasp_pkpd Heise T, et al. A Faster-Onset Formulation of Insulin Aspart. Clin Pharmacokinet. 2017. DOI: 10.1007/s40262-017-0510-8

  8. klaff_2020_urli_pkpd Klaff LJ, et al. Ultra Rapid Lispro Demonstrates Accelerated Pharmacokinetics and Pharmacodynamics. Diabetes Obes Metab. 2020. DOI: 10.1111/dom.14049

  9. wentholt_2004_cgm_lag Wentholt IME, et al. How glucose sensors can facilitate therapy in diabetes management. Diabetes Technol Ther. 2004. DOI: 10.1089/dia.2004.6.615

  10. dalla_man_2007_meal_model Dalla Man C, Rizza RA, Cobelli C. Meal simulation model of the glucose-insulin system. IEEE Trans Biomed Eng. 2007. DOI: 10.1109/TBME.2007.893506

  11. visentin_2018_uvapadova Visentin R, et al. The University of Virginia/Padova Type 1 Diabetes Simulator Matches the 2014 DMMS.R. J Diabetes Sci Technol. 2018. DOI: 10.1177/1932296818757747

  12. mujahid_2024_generative_t1d_simulator Mujahid O, Contreras I, Beneyto A, Vehi J. Generative deep learning for the development of a type 1 diabetes simulator. Communications Medicine. 2024. DOI: 10.1038/s43856-024-00476-0

  13. riddell_2017_exercise_consensus Riddell MC, et al. Exercise management in type 1 diabetes: a consensus statement. Lancet Diabetes Endocrinol. 2017. DOI: 10.1016/S2213-8587(17)30014-1

  14. marling_2020_ohiot1dm Marling C, Bunescu R. The OhioT1DM Dataset for Blood Glucose Level Prediction: Update 2020. CEUR Workshop Proceedings. Paper: ceur-ws.org/Vol-2675/paper2.pdf

  15. idc_2025_agp_report_overview HealthPartners Institute / International Diabetes Center. Guide to Understanding the Ambulatory Glucose Profile (AGP) Report. 2025. PDF: healthpartners.com

  16. bergman_1979_minimal_model Bergman RN, Ider YZ, Bowden CR, Cobelli C. Quantitative estimation of insulin sensitivity. Am J Physiol. 1979. DOI: 10.1152/ajpendo.1979.236.6.E667

  17. hovorka_2004_nmpc_t1d Hovorka R, et al. Nonlinear model predictive control of glucose concentration in subjects with type 1 diabetes. Physiol Meas. 2004. DOI: 10.1088/0967-3334/25/4/010

  18. gerich_1979_counterregulation Gerich J, Davis J, Lorenzi M. Hormonal mechanisms of recovery from insulin-induced hypoglycemia in man. Am J Physiol. 1979. DOI: 10.1152/ajpendo.1979.236.4.E380

  19. cryer_2013_haaf_mechanisms Cryer PE. Mechanisms of Hypoglycemia-Associated Autonomic Failure in Diabetes. N Engl J Med. 2013. DOI: 10.1056/NEJMra1215228

  20. cryer_2013_haaf_diabetes Cryer PE. Hypoglycemia-associated autonomic failure in diabetes. Handb Clin Neurol. 2013. DOI: 10.1016/B978-0-444-53491-0.00023-7

  21. lv_2013_exogenous_glucagon_pk Lv D, Breton MD, Farhy LS. Pharmacokinetics modeling of exogenous glucagon in type 1 diabetes mellitus patients. Diabetes Technol Ther. 2013. DOI: 10.1089/dia.2013.0150

  22. haidar_2013_dual_hormone_ap Haidar A, et al. Glucose-responsive insulin and glucagon delivery in adults with type 1 diabetes. CMAJ. 2013. DOI: 10.1503/cmaj.121265

  23. haidar_2013_insulin_glucagon_pk Haidar A, Duval C, Legault L, Rabasa-Lhoret R. Pharmacokinetics of Insulin Aspart and Glucagon in Type 1 Diabetes during Closed-Loop Operation. J Diabetes Sci Technol. 2013. DOI: 10.1177/193229681300700610

  24. hummel_2018_renal_glucose_handling Hummel CS, Lu C, Loo DDF, Hirayama BA, Voss AA, Wright EM. Physiology of renal glucose handling via SGLT1, SGLT2 and GLUT2. Diabetologia. 2018. DOI: 10.1007/s00125-018-4656-5

  25. defronzo_2013_renal_reabsorption_splay DeFronzo RA, et al. Characterization of Renal Glucose Reabsorption in Response to Dapagliflozin. Diabetes Care. 2013. DOI: 10.2337/dc13-0387

  26. mandelbrot_1968_fractional_brownian Mandelbrot BB, Van Ness JW. Fractional Brownian Motions, Fractional Noises and Applications. SIAM Review. 1968. DOI: 10.1137/1010093

  27. campbell_1985_dawn_phenomenon Campbell PJ, Bolli GB, Cryer PE, Gerich JE. Pathogenesis of the Dawn Phenomenon in Patients with Insulin-Dependent Diabetes Mellitus. N Engl J Med. 1985. DOI: 10.1056/NEJM198506063122302

  28. richter_2013_glut4_exercise Richter EA, Hargreaves M. Exercise, GLUT4, and skeletal muscle glucose uptake. Physiol Rev. 2013. DOI: 10.1152/physrev.00038.2012

  29. naslund_1999_glp1_gastric_emptying Naslund E, et al. GLP-1 slows solid gastric emptying and inhibits insulin, glucagon, and PYY release in humans. Am J Physiol Regul Integr Comp Physiol. 1999. DOI: 10.1152/ajpregu.1999.277.3.R910

Documentation-Only Local AI Setup Sources

These are the official references used in the guides for installing Ollama, understanding the local Ministral 3 family, and explaining why the SDK recommends different model sizes for different hardware.

  1. ollama_linux_install Ollama. Linux Installation Documentation. URL: docs.ollama.com/linux

  2. mistral_2025_ministral_3_announcement Mistral AI. Introducing Mistral 3. URL: mistral.ai/news/mistral-3

  3. mistral_2025_ministral_3_3b Mistral AI Docs. Ministral 3 3B. URL: docs.mistral.ai/models/ministral-3-3b-25-12

  4. mistral_2025_ministral_3_8b Mistral AI Docs. Ministral 3 8B. URL: docs.mistral.ai/models/ministral-3-8b-25-12

  5. mistral_2025_ministral_3_14b Mistral AI Docs. Ministral 3 14B. URL: docs.mistral.ai/models/ministral-3-14b-25-12

  6. mistral_2026_adjustable_reasoning Mistral AI Docs. Adjustable Reasoning. URL: docs.mistral.ai/studio-api/conversations/reasoning/adjustable

  7. mistral_2026_small_4 Mistral AI Docs. Mistral Small 4. URL: docs.mistral.ai/models/model-cards/mistral-small-4-0-26-03

  8. mistral_2026_medium_35 Mistral AI Docs. Mistral Medium 3.5. URL: docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04

Documentation-Only Device Emulation References

These references are used for the best-effort emulator notes and are not claims of exact proprietary algorithm reproduction.

Medtronic MiniMed 780G

  1. bergenstal_2020_780g Bergenstal RM, et al. Safety of a Hybrid Closed-Loop Insulin Delivery System in Patients With Type 1 Diabetes. DOI: 10.1056/NEJMoa2003479

  2. fda_k193510_780g U.S. FDA. 510(k) K193510 - MiniMed 780G System. URL: accessdata.fda.gov

  3. medtronic_780g_user_guide Medtronic Diabetes. MiniMed 780G User Guide / Product Documentation. URL: medtronicdiabetes.com

Tandem Control-IQ

  1. brown_2019_control_iq_dtt Brown SA, et al. Performance of the Tandem t:slim X2 insulin pump with Control-IQ technology in the International Diabetes Closed-Loop trial. DOI: 10.1089/dia.2019.0226

  2. fda_k191289_control_iq U.S. FDA. 510(k) K191289 - Control-IQ System. URL: accessdata.fda.gov

  3. idcl_nct03563313 ClinicalTrials.gov. International Diabetes Closed Loop Trial. URL: clinicaltrials.gov/ct2/show/NCT03563313

  4. control_iq_user_guide Tandem Diabetes Care. Control-IQ User Guide / Product Documentation. URL: tandemdiabetes.com

Omnipod 5

  1. assert_omnipod_5 Insulet / Omnipod. ASSERT Trial - Omnipod 5. URL: omnipod.com/assert-trial

  2. onset_omnipod_5 Insulet / Omnipod. ONSET Trial - Omnipod 5 in Type 2 Diabetes. URL: omnipod.com/onset-trial

  3. fda_k203467_omnipod5 U.S. FDA. 510(k) K203467 - Omnipod 5 System. URL: accessdata.fda.gov

  4. omnipod5_user_guide Insulet / Omnipod. Omnipod 5 User Guide / Product Documentation. URL: omnipod.com

Reproducibility Note

For report reproducibility, pair this file with: - run metadata (run_metadata.json) - manifest hashes (run_manifest.json) - dataset lineage fields in research outputs (source_file_sha256, split metadata).

If you want the machine-readable packaged subset, export it with:

iints sources --output-json results/source_manifest.json