# iTech Clinical Intelligence > iTech Clinical Intelligence is a physician-founded, radiologist-led company building the intelligence layer for medical imaging: deterministic radiology criteria engines with provenance, clinician-controlled reporting, and an API that software and AI agents can call. Legal entity: iTECH LLC. Founded by a practicing, board-certified radiologist with graduate training in computer science, data science and artificial intelligence. Canonical site: https://itecclinical.com. Key facts: - Category: imaging intelligence infrastructure; domain-specialized clinical intelligence for radiology. - Core idea: "The model is one component." Deterministic criteria decide; language models write inside those verdicts; a qualified clinician signs. - RadVector API is a radiology inference API. Validated criteria packs today: Fleischner (2017, pack fleischner-2017-r6, validated 2026-10-04), BI-RADS (ACR 5th ed., pack 2026.1, validated 2026-08-18), DEXA (WHO 1994 + ISCD 2019, pack 2026.2, validated 2026-10-01). Draft (advisory only): Lung-RADS v2022, TI-RADS, PI-RADS v2.1, O-RADS US, LI-RADS, Bosniak, C-RADS, NI-RADS. - Every API verdict carries provenance: criteria pack, pack version, validation status, validator and date, engine, trace id. - Built to need minimal PHI (no identifier fields, free text masked, request text never logged or stored); Business Associate Agreement available for clinical deployments, including AI platforms calling it for hospitals. Not FDA-cleared. Clinical decision support for qualified professionals, not autonomous diagnosis. - Reporting standards applied by one deterministic engine: Lung-RADS, BI-RADS, TI-RADS, LI-RADS, O-RADS, PI-RADS, Bosniak, Fleischner, DEXA. Plain-language guide per standard at https://itecclinical.com/standards/. - RadVector API returns every evaluation as JSON or a FHIR R4 Bundle, accepts other imaging AI systems' results as FHIR R4 Observations or DICOM SR TID 1500, codes findings with verified RadElement codes, offers a pre-read context service from prior reports, sandbox keys with synthetic fixtures, criteria-pack version pinning, closed-set deferral codes, batch evaluation and idempotent retries. Developer guide: https://radvector.ai/portal/developers. - RadVector Reporting keeps a per-report governance record (standards applied, pack versions, validation status; no report text) with CSV export, and a workflow value panel built only from measured data. - PACS/DICOMweb connectivity is in development (not yet available). RadVector reads RIS worklists over REST or FHIR R4 and accepts HL7 v2 orders (ORM); RadVector Reporting does not yet write reports back over FHIR or HL7 (reports leave by copy). - EHR / EMR integration: RadVector Reporting opens from the EHR as a SMART on FHIR R4 app (SMART App Launch, confidential client, PKCE; patient, order, prior reports and documents read over FHIR R4, read-only) and offers a CDS Hooks 2.0 advisory service that stays silent until validated. One vendor-neutral gateway with a profile per vendor serves Epic (registered in Epic's developer program for non-production use), Oracle Health (formerly Cerner) and other SMART-enabled EHRs. Nothing is written back to the EHR. iTech is not affiliated with or endorsed by Epic or Oracle. - Does not hold SOC 2, HITRUST, ISO 27001 or an independent penetration test. ## Products - [RadVector API](https://itecclinical.com/products/radvector-api/): deterministic radiology criteria verdicts with provenance, callable by software and AI agents. Docs: https://api.radvector.ai/v1/docs. Keys: https://radvector.ai/portal. - [RadVector Reporting](https://itecclinical.com/products/radvector-reporting/): radiologist-controlled reporting and decision support (formerly RadCloudAI). Product site: https://radcloud.net/ - [QLumis](https://itecclinical.com/products/qlumis/): private multimodal AI companion with persistent memory. Not a medical device. Product site: https://qlumis.com/ ## Reporting standards - [All reporting standards](https://itecclinical.com/standards/): Lung-RADS, BI-RADS, TI-RADS, LI-RADS, O-RADS, PI-RADS, Bosniak, Fleischner, DEXA. - [Lung-RADS (v2022)](https://itecclinical.com/standards/lung-rads/): Lung-RADS is the American College of Radiology system for reporting and managing findings on low-dose CT lung cancer screening, assigning each screening exam a category from 0 to 4X that sets the next step in screening. - [BI-RADS (ACR BI-RADS Atlas, 5th edition)](https://itecclinical.com/standards/bi-rads/): BI-RADS is the American College of Radiology system for breast imaging that assigns every mammogram, breast ultrasound or breast MRI a final assessment category from 0 to 6 tied to a management recommendation, plus a breast density category from a to d. - [TI-RADS (ACR TI-RADS white paper (2017))](https://itecclinical.com/standards/ti-rads/): ACR TI-RADS is a point-based system for thyroid nodules on ultrasound: points for composition, echogenicity, shape, margin and echogenic foci add up to a level from TR1 to TR5, and the level plus the nodule's size decides whether fine-needle aspiration or follow-up is recommended. - [LI-RADS (CT/MRI LI-RADS v2018)](https://itecclinical.com/standards/li-rads/): LI-RADS is the American College of Radiology system for liver observations on CT and MRI in patients at risk for hepatocellular carcinoma, assigning categories from LR-1 (definitely benign) to LR-5 (definitely HCC), plus LR-M, LR-TIV and LR-NC. - [O-RADS (O-RADS US v2022)](https://itecclinical.com/standards/o-rads/): O-RADS US is the American College of Radiology risk-stratification and management system for ovarian and adnexal lesions on ultrasound, assigning a score from 0 to 5 that estimates the risk of malignancy and guides management. - [PI-RADS (PI-RADS v2.1 (2019))](https://itecclinical.com/standards/pi-rads/): PI-RADS v2.1 is the scoring system for prostate MRI that assigns each lesion a category from 1 (clinically significant cancer highly unlikely) to 5 (highly likely), using diffusion-weighted imaging as the dominant sequence in the peripheral zone and T2-weighted imaging in the transition zone. - [Bosniak (Version 2019)](https://itecclinical.com/standards/bosniak/): The Bosniak classification, updated in version 2019, sorts cystic renal masses on CT and MRI into classes I, II, IIF, III and IV by wall, septa, enhancement, mural nodules and attenuation, from simple benign cysts to masses likely to be malignant. - [Fleischner (Fleischner Society 2017)](https://itecclinical.com/standards/fleischner/): The Fleischner Society 2017 guidelines give follow-up recommendations for pulmonary nodules found incidentally on CT in adults, based on nodule type (solid or subsolid), size, number and the patient's risk for lung cancer. - [DEXA (WHO 1994 T-score criteria with ISCD 2019 Official Positions)](https://itecclinical.com/standards/dexa/): DEXA bone density classification applies the World Health Organization T-score thresholds (normal at -1.0 or above, low bone mass between -1.0 and -2.5, osteoporosis at -2.5 or below) at the diagnostic sites the International Society for Clinical Densitometry specifies. ## Platform - [Platform overview](https://itecclinical.com/platform/): four layers (clinical intelligence, clinical applications, embedded intelligence, human-AI continuity) and five technology principles. - [Developers](https://itecclinical.com/developers/): real request/response example and the /v1 route list. ## Solutions - [PACS, RIS and VNA / enterprise imaging](https://itecclinical.com/solutions/pacs-ris-vna/): add intelligence without replacing the imaging stack. - [EHR and EMR integration](https://itecclinical.com/solutions/ehr-emr-integration/): open RadVector from Epic, Oracle Health and other EHRs via SMART on FHIR R4 and CDS Hooks. - [AI and foundation-model companies](https://itecclinical.com/solutions/ai-models/): a callable, maintained radiology layer for general-purpose models and agents. - [Modality manufacturers](https://itecclinical.com/solutions/modalities/): integration direction for acquisition context (no current integration). - [All solutions](https://itecclinical.com/solutions/) ## Company - [Company](https://itecclinical.com/company/): physician-founded, radiologist-led; values and lineage. - [Trust](https://itecclinical.com/trust/): clinical governance, PHI handling per product, provenance, security, regulatory posture. - [FAQ](https://itecclinical.com/faq/): direct answers about the company and products. - [Contact](https://itecclinical.com/contact/) ## Optional - [Full text for AI systems](https://itecclinical.com/llms-full.txt)