# ProRadIQ — Full content corpus for LLMs > Plain-text knowledge base for AI assistants (ChatGPT, Perplexity, Claude, Gemini). All content here is approved for citation. See /llms.txt for the index. ## About ProRadIQ ProRadIQ is a DICOM-first, radiologist-led, AI-assisted radiology reporting platform built in India. Founded and clinically led by Dr. Harjot Singh. AI drafts a structured report in 3–5 minutes; a registered radiologist edits and digitally signs every report before release. ISO 13485, ISO/IEC 27001, IEC 62304 and ISO 14971 certified. HIPAA-aware, GDPR/UK GDPR/DPDP, plus country-specific compliance for UAE (DHA), Australia (AHPRA), Ghana (MDC), Kenya (KMPDC), Nigeria (MDCN), and UK (GMC/CQC). Contact: harjot@proradiq.com, +91 81466 87177. ## Teleradiology basics ### What is teleradiology? URL: https://www.proradiq.com/ask/what-is-teleradiology Last reviewed: 2026-06-26 **Short answer.** Teleradiology is the electronic transmission of radiology images — X-ray, CT, MRI, ultrasound — from one location to a qualified radiologist at another, who interprets the study and returns a signed report. Teleradiology is the practice of acquiring medical images at one site and sending them over a secure network to a qualified radiologist at another site, who reports them remotely. It exists because radiologists are scarce and unevenly distributed, while imaging machines are everywhere — including small towns, after-hours emergency departments, and partner clinics. A typical teleradiology workflow involves four steps: (1) the modality (CT, MRI, X-ray, ultrasound) acquires the study and sends it to a local PACS or DICOM gateway; (2) the study is pushed securely to the reporting radiologist's worklist; (3) the radiologist opens it in a DICOM viewer, dictates or types the report; (4) the signed report (PDF or DOCX) is returned to the originating centre, usually with critical-finding callouts highlighted. Modern teleradiology adds AI-drafted reports, structured templates, audit trails, and council-registration checks before sign-off. ProRadIQ offers both classic telereporting (where Dr. Harjot Singh signs the report) and AI-assisted reporting (where the centre's own radiologist signs the AI draft). ### What's the difference between teleradiology and AI-assisted reporting? URL: https://www.proradiq.com/ask/teleradiology-vs-ai-reporting Last reviewed: 2026-06-26 **Short answer.** Teleradiology rents you a remote radiologist who signs the report. AI-assisted reporting gives your own radiologist an AI-drafted report to edit and sign in a fraction of the usual time. ProRadIQ supports both. Traditional teleradiology is a staffing solution: a centre without a radiologist sends studies to a remote one who reports and signs them. The radiologist is the bottleneck; the workflow scales by hiring more radiologists. AI-assisted reporting is a software solution: the centre's own radiologist receives an AI-generated draft report — typically in 3–5 minutes — and edits and signs it. The radiologist remains in full control; the AI removes the drudgery of dictating boilerplate findings. Which do you need? Centres with no on-site radiologist usually start with telereporting and migrate to AI-assisted reporting once they hire. Centres that already employ a radiologist usually skip telereporting and go straight to AI-assisted reporting to multiply their reporting throughput. ProRadIQ supports both models on the same DICOM pipeline, audit trail, and signed-report export. ## AI in radiology ### How is AI used in radiology? URL: https://www.proradiq.com/ask/what-is-ai-radiology Last reviewed: 2026-06-26 **Short answer.** AI in radiology automates routine tasks — measuring lesions, detecting nodules, drafting reports — so a radiologist spends less time on boilerplate and more on the actual diagnosis. It augments radiologists, it does not replace them. Artificial intelligence in radiology covers three broad buckets: (1) **triage** — flagging urgent findings like intracranial haemorrhage or pulmonary embolism so they jump the queue; (2) **measurement and detection** — automating quantitative tasks such as lung nodule volume, fracture detection, or lesion tracking on follow-up scans; (3) **report drafting** — generating a first-pass structured report directly from the images so the radiologist edits rather than types from scratch. ProRadIQ focuses on bucket (3). For each study, a modality-aware large-vision model produces a structured draft — typically in 3–5 minutes — with section headings (Technique, Findings, Impression) and bolded positive findings. The radiologist reviews the draft alongside the images, corrects what's wrong, and digitally signs the final report. Every correction is learned by an organisation-scoped memory so the drafts get closer to that radiologist's style over time. Crucially, no report is auto-published. AI-assisted means human-signed. ### Will AI replace radiologists? URL: https://www.proradiq.com/ask/does-ai-replace-radiologists Last reviewed: 2026-06-26 **Short answer.** No. AI replaces the typing, the measuring, and the boilerplate — not the radiologist. Every ProRadIQ report is reviewed, edited where needed, and digitally signed by a registered radiologist before release. Radiology has been the example field for "AI is coming for your job" predictions for the better part of a decade. The actual trajectory has been different: AI handles narrow, repetitive tasks well (organ segmentation, nodule detection, fracture screening, draft generation) and handles synthesis, clinical correlation, and medico-legal accountability poorly. A radiology report is more than image interpretation: it is a clinically and legally accountable document that ties the imaging to the patient's history, prior studies, and downstream care. That accountability rests with a registered human radiologist in every jurisdiction we operate in — India (NMC), UK (GMC), UAE (DHA), Australia (AHPRA), Kenya (KMPDC), Nigeria (MDCN), Ghana (MDC), and elsewhere. ProRadIQ is built around this reality. The AI drafts; the radiologist signs. Council registration is verified at sign-off so an unregistered radiologist cannot release a report. The result is not fewer radiologists — it is faster, more consistent radiology. ### Can AI radiology reports hallucinate findings? URL: https://www.proradiq.com/ask/ai-hallucinations Last reviewed: 2026-06-26 **Short answer.** Yes, any generative model can. ProRadIQ mitigates this with modality-specific safety scrubs, an organisation-scoped correction memory, structured evidence digests from every image, and — most importantly — a registered radiologist who reviews and signs every report. Generative AI can produce confident-sounding text that isn't supported by the images — "hallucinations" in industry shorthand. Common patterns we see in radiology drafts include phantom apical scarring on CT chest, over-called mucosal thickening on CT head and MRI brain, and made-up incidental findings on abdominal CT. ProRadIQ ships several layers of defence: - **Modality safety scrubs** — rule-based filters that strip specific known false-positive phrases unless the image evidence explicitly supports them (e.g. apical scarring requires fibrotic signs *plus* anatomic location). - **Structured image evidence digest** — the model is required to cite the image series, slice and region of interest for every positive finding before the prose narrative is written. - **Organisation-scoped correction memory** — every edit a radiologist makes is learned (PHI-scrubbed, never crosses tenant boundaries) and feeds back into future drafts. - **Human sign-off** — the final mitigation. Every report is reviewed and signed by a registered radiologist; the AI never publishes a report on its own. No AI radiology platform is hallucination-free. The honest answer is that you need the safety stack *and* the human reviewer. ### What is correction memory and is it safe? URL: https://www.proradiq.com/ask/correction-memory Last reviewed: 2026-06-26 **Short answer.** Correction memory captures every edit a radiologist makes to a draft, scrubs the PHI, and feeds the pattern back so future drafts read more like that radiologist's style. It is strictly organisation-scoped — nothing crosses tenant boundaries. Every time a radiologist edits or rephrases a section of a draft, ProRadIQ stores the before/after with the imaging context. That correction is: 1. **PHI-scrubbed** — patient name, ID, age, date, accession, referring physician and any free-text containing identifiers are stripped before storage. 2. **Organisation-scoped** — the correction is only available to drafts inside the same organisation. There is no global pool. There is no cross-org leakage. A correction made by Dr. A at Centre X is invisible to Dr. B at Centre Y. 3. **Optionally shareable** — organisations can opt-in to anonymised platform-wide correction learning, which goes through a 24-hour review and embedding pipeline. Default is off. The effect is that drafts converge on your team's preferred phrasing — section ordering, qualifier choice, units, sign-off lines — usually within the first 50–100 cases. Radiologists report 30–50% less editing on routine work after the first month. ## Compliance & regulation ### Is ProRadIQ HIPAA-compliant? URL: https://www.proradiq.com/ask/hipaa-compliance Last reviewed: 2026-06-26 **Short answer.** ProRadIQ is HIPAA-aware: PHI is encrypted in transit and at rest, access is per-tenant with audit logs, and signed reports are hash-chained. ProRadIQ is ISO/IEC 27001 and ISO 13485 certified. For US-regulated deployments we sign a BAA. HIPAA (the US Health Insurance Portability and Accountability Act) governs how covered entities and their business associates handle protected health information (PHI). ProRadIQ's posture: - **Encryption** — TLS 1.2+ in transit; AES-256 at rest for both DICOM images and report text. - **Access control** — strict per-organisation tenancy. No cross-organisation reads; no global indexes on patient identifiers; row-level security on every patient-data table. - **Audit logs** — every view, edit, sign-off, export and download is logged with user id, timestamp and IP. Signed reports are linked into a SHA-256 hash chain so tampering is detectable. - **Minimum necessary** — clinical history is the only free-text input; we do not request or store data beyond what the radiologist needs to report. - **Breach register** — internal breach detection and reporting workflow. - **Certifications** — ISO 13485 (medical devices QMS), ISO/IEC 27001 (information security), IEC 62304 (medical software lifecycle), ISO 14971 (risk management). For deployments serving US patients we sign a Business Associate Agreement (BAA) with the covered entity. Contact harjot@proradiq.com to arrange. ### Is ProRadIQ GDPR-compliant? URL: https://www.proradiq.com/ask/gdpr-compliance Last reviewed: 2026-06-26 **Short answer.** Yes. ProRadIQ is GDPR and UK GDPR aware: data minimisation, encryption, per-tenant isolation, EU/UK data residency on request, signed Data Processing Agreement, and DSAR fulfilment built into the platform. GDPR (Regulation (EU) 2016/679) and the UK GDPR apply when ProRadIQ processes personal data of EU or UK data subjects. Our compliance stance: - **Lawful basis** — ProRadIQ acts as a data processor for the diagnostic centre or hospital (the data controller). A signed Data Processing Agreement (Article 28) governs the relationship. - **Data minimisation** — only the data needed to produce a radiology report is collected: study DICOM, clinical history, referring physician where supplied. - **Data subject rights** — the platform ships a built-in DSAR (Data Subject Access Request) workflow for access, rectification, erasure and portability. - **Data residency** — EU/UK deployments can be pinned to EU regions on request. - **International transfers** — where data leaves the EU/UK we use Standard Contractual Clauses (SCCs) plus supplementary technical measures. - **Breach notification** — 72-hour internal breach detection target; we notify the controller without undue delay so they can meet their Article 33 obligations. - **DPO contact** — harjot@proradiq.com for all GDPR enquiries. ISO/IEC 27001 certification underpins our information security controls. ProRadIQ is not a medical device under EU MDR Class IIa or above for the AI-draft function (the radiologist is the medical decision-maker). ### What ISO certifications does ProRadIQ hold? URL: https://www.proradiq.com/ask/iso-certifications Last reviewed: 2026-06-26 **Short answer.** ProRadIQ is certified to ISO 13485 (medical device quality), IEC 62304 (medical software lifecycle), ISO 14971 (medical device risk management) and ISO/IEC 27001 (information security). ProRadIQ's quality and security posture rests on four ISO/IEC standards: - **ISO 13485** — quality management system for medical devices. Governs how we design, develop, install, and service the platform. - **IEC 62304** — software lifecycle processes for medical device software. Governs how every code change, test, and release is documented and traceable. - **ISO 14971** — risk management for medical devices. Governs how clinical risks (false-positive findings, missed findings, security breaches) are identified, mitigated and reviewed. - **ISO/IEC 27001** — information security management. Governs access control, encryption, vendor risk, incident response, and audit logging. Certificates and scope are published at /iso-certifications and available on request to enterprise clients. The certifications are independently issued and renewed annually. ### Which country regulators does ProRadIQ align with? URL: https://www.proradiq.com/ask/country-regulators Last reviewed: 2026-06-26 **Short answer.** ProRadIQ validates the signing radiologist's registration against the appropriate council: NMC (India), GMC (UK), DHA (UAE), AHPRA (Australia), KMPDC (Kenya), MDCN (Nigeria), MDC (Ghana). Radiology reports are legally binding clinical documents in every jurisdiction. ProRadIQ enforces that the signing radiologist is currently registered with the relevant council before sign-off is permitted: - **India** — National Medical Commission (NMC) / state medical councils. - **United Kingdom** — General Medical Council (GMC), CQC-aligned governance for clinics. - **United Arab Emirates** — Dubai Health Authority (DHA), with MOHAP/DOH equivalents accepted. - **Australia** — Australian Health Practitioner Regulation Agency (AHPRA), radiology specialty. - **Kenya** — Kenya Medical Practitioners and Dentists Council (KMPDC). - **Nigeria** — Medical and Dental Council of Nigeria (MDCN). - **Ghana** — Medical and Dental Council (MDC). If the signing radiologist's registration cannot be validated, the platform blocks sign-off. This is non-overridable. ## Modalities ### Which modalities does ProRadIQ support? URL: https://www.proradiq.com/ask/supported-modalities Last reviewed: 2026-06-26 **Short answer.** X-ray, CT (including HRCT and contrast-enhanced), MRI (brain, spine, MSK, body), mammography, ultrasound (radiologist-signed reporting), and PET-CT follow-up. Modality-aware prompts deliver structured reports tailored to each. ProRadIQ supports the full range of cross-sectional and projection imaging: - **X-ray (DR/CR)** — chest, MSK, abdomen, spine. Two-view chest typically drafts in under 90 seconds. - **CT** — non-contrast and contrast-enhanced studies across head, chest (including HRCT for ILD), abdomen, MSK, spine and angiography. - **MRI** — brain, whole-spine (per-level disc reporting), MSK joints, body MRI. - **Mammography** — screening and diagnostic with BI-RADS structured impression. - **Ultrasound** — radiologist-signed reports for abdomen, pelvis, obstetric, MSK and vascular studies. - **PET-CT** — follow-up oncology imaging with prior-comparison support. Each modality has its own prompt and safety profile — for example, the CT chest pipeline carries a specific false-positive scrub for apical scarring; the CT head and MRI brain pipeline carries one for mucosal thickening. ### CT vs MRI — when to use each? URL: https://www.proradiq.com/ask/ct-vs-mri Last reviewed: 2026-06-26 **Short answer.** CT is fast, widely available, and excellent for bone, lung and acute haemorrhage. MRI uses no ionising radiation and is better for soft tissue, brain parenchyma, spine and joints. They are complementary, not substitutes. Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) answer different clinical questions: - **Choose CT** when you need speed (trauma, stroke, acute abdomen), bone detail (fractures, lung parenchyma in HRCT), or vascular imaging (CT angiography). CT studies acquire in seconds and tolerate uncooperative or unstable patients. - **Choose MRI** when you need soft-tissue contrast (brain, spinal cord, ligaments, tumours), no ionising radiation (paediatrics, pregnancy), or specific physiology (diffusion, perfusion, spectroscopy). MRI takes 15–45 minutes and demands a still patient. In practice they are complementary: a stroke patient gets a non-contrast CT first to rule out haemorrhage, then MRI to characterise the infarct. A spine patient gets X-ray for alignment, CT for bony detail and MRI for disc and cord. ProRadIQ supports all three modalities on the same workflow. ### How does ProRadIQ report spine MRI? URL: https://www.proradiq.com/ask/spine-mri-reporting Last reviewed: 2026-06-26 **Short answer.** One paragraph per imaged disc level — disc, nerve roots and canal described together — so the report reads the way a radiologist dictates and the way a clinician scans it. Spine MRI is high-volume and high-repetition: every level needs disc morphology, signal, height, any protrusion or herniation, the nerve roots, and the central canal. ProRadIQ's spine prompt mandates one paragraph per imaged disc level (e.g. L4–L5, L5–S1) so a referring clinician can scan straight to the level they care about, and the radiologist can edit one block at a time. Within each level paragraph: disc → exiting and traversing nerve roots → central and lateral canal → facet joints if abnormal. Positive findings are bolded. Normal levels are reported compactly. This layout was specifically requested by spine-heavy clients and replaced an earlier organ-checklist format that was harder to read. Spine ultrasound, whole-spine MRI screening and follow-up comparison are also supported. ### What are the most common findings on a chest X-ray? URL: https://www.proradiq.com/ask/chest-xray-findings Last reviewed: 2026-06-26 **Short answer.** Chest X-ray findings cluster into lung (consolidation, opacity, effusion, pneumothorax, mass), cardiac (cardiomegaly, pulmonary oedema), mediastinal (widening, hilar masses), pleural (effusion, thickening) and bony (rib fractures) categories. A chest radiograph (CXR) is the single most-ordered radiology study in the world, and its interpretation follows a systematic search pattern. The most commonly reported positive findings: - **Lung parenchyma**: consolidation (pneumonia, infarct), reticular or nodular opacities (interstitial disease, metastases), focal mass, atelectasis. - **Pleural space**: pleural effusion, pneumothorax, pleural thickening. - **Cardiac silhouette**: cardiomegaly (cardiothoracic ratio > 0.5 on PA), pericardial effusion (globular silhouette). - **Mediastinum**: widened mediastinum (aortic pathology, mass), tracheal deviation, hilar enlargement (lymphadenopathy, vascular). - **Pulmonary vasculature**: upper-lobe diversion and Kerley B lines (pulmonary oedema), enlarged pulmonary arteries (pulmonary hypertension). - **Bones and soft tissues**: rib fractures, lytic lesions, subcutaneous emphysema. - **Tubes and lines**: ET tube, NG tube, central line and pacemaker lead positions. ProRadIQ's chest X-ray draft works through this pattern in order so nothing is missed, and bolds any positive findings. ### How is a stroke CT interpreted? URL: https://www.proradiq.com/ask/stroke-ct-interpretation Last reviewed: 2026-06-26 **Short answer.** First rule out haemorrhage on non-contrast CT (hyperdense blood). Then look for early ischaemic changes (loss of grey-white differentiation, sulcal effacement, dense MCA sign) and add CT angiography to find the occlusion if thrombectomy is on the table. Stroke CT is time-critical because thrombolysis and thrombectomy windows are measured in hours. The standard reporting pattern: 1. **Non-contrast CT head (NCCT)** — rules out intracranial haemorrhage, which would contraindicate thrombolysis. Look for hyperdense blood (subdural, extradural, subarachnoid, intra-axial). If clear of blood, look for early ischaemic changes: loss of grey-white differentiation, sulcal effacement, insular ribbon sign, lentiform obscuration, hyperdense MCA sign. ASPECTS score (1–10) quantifies the territorial involvement. 2. **CT angiography (CTA)** — head and neck — identifies the occluded vessel (M1, M2, ICA terminus, basilar) and characterises collaterals. Required for thrombectomy decision-making. 3. **CT perfusion (CTP)** — characterises core (irreversibly infarcted) vs penumbra (salvageable). Increasingly used for late-window thrombectomy candidacy. ProRadIQ's stroke CT draft works through this triad systematically and flags critical findings (haemorrhage, large vessel occlusion, ASPECTS ≤ 5) so the on-call team sees them immediately. ### What are the red flags on emergency imaging? URL: https://www.proradiq.com/ask/red-flags-emergency Last reviewed: 2026-06-26 **Short answer.** Acute intracranial haemorrhage, large-vessel stroke, tension pneumothorax, pulmonary embolism, aortic dissection, free intraperitoneal air or fluid, bowel obstruction with ischaemia, ectopic pregnancy, and testicular torsion top the list. Emergency radiology is a list of can't-miss diagnoses. ProRadIQ's emergency draft tags any of these as critical findings and surfaces them at the top of the report: - **Neuro**: acute intracranial haemorrhage (subdural, extradural, subarachnoid, intra-axial), large-vessel occlusion stroke, brain herniation, spinal cord compression. - **Chest**: tension pneumothorax, massive haemothorax, pulmonary embolism, aortic dissection, oesophageal rupture, foreign body airway obstruction. - **Abdomen**: free intraperitoneal air (perforation), mesenteric ischaemia, bowel obstruction with ischaemia, ruptured abdominal aortic aneurysm, acute appendicitis with perforation, perforated viscus. - **Obs/gyn**: ectopic pregnancy (especially ruptured), placental abruption, uterine rupture, ovarian torsion. - **Urology**: testicular torsion, obstructing renal stone with infection. - **MSK**: open fracture, compartment syndrome signs, septic joint, unstable spine fracture. These critical findings are highlighted on the report header so the referring clinician sees them before reading the rest. ## Workflow & turnaround ### What is PACS and how does ProRadIQ work with it? URL: https://www.proradiq.com/ask/what-is-pacs Last reviewed: 2026-06-26 **Short answer.** PACS (Picture Archiving and Communication System) is the on-site server that stores and serves DICOM images. ProRadIQ ingests directly from any PACS — Orthanc, Conquest, dcm4chee, Synapse, Centricity — via DICOM C-STORE or a small Lua push script. A PACS is the imaging backbone of every diagnostic centre: it stores DICOM studies acquired by modalities (CT, MRI, X-ray, ultrasound), serves them to viewing workstations, and forwards them to remote reporting destinations. ProRadIQ is PACS-agnostic. Three integration paths are supported: 1. **Orthanc Lua push** — for centres running the open-source Orthanc PACS. We provide a one-file Lua script that pushes each newly stored study to ProRadIQ over HTTPS with a centre-specific token. Setup takes about 15 minutes. 2. **DICOM C-STORE** — the standard DICOM SCP/SCU protocol. Configure ProRadIQ as a destination in your existing PACS (Conquest, dcm4chee, Synapse, Centricity, etc.) and push from there. 3. **Direct ZIP upload** — for one-off cases or centres without a PACS, drop a DICOM ZIP (or even loose .dcm / .jpg / .png files) into the case-creation page. No VPN or fixed IP is required; all traffic is HTTPS with per-centre tokens. ### How fast is the AI draft on ProRadIQ? URL: https://www.proradiq.com/ask/how-fast-is-ai-draft Last reviewed: 2026-06-26 **Short answer.** The AI draft typically completes in 3–5 minutes per study, depending on modality, image count, and current load. Plain films are faster; multi-series CT and MRI sit at the longer end. Draft latency depends on three things: image count, model reasoning depth, and current platform load. - A two-view chest X-ray usually drafts in under 90 seconds. - A non-contrast head CT (~250 slices) drafts in about 2–3 minutes. - A multi-sequence MRI brain or whole-spine MRI (~600+ images across sequences) drafts in about 4–6 minutes. - HRCT chest, full-body PET-CT or follow-up oncology CTs with priors can take 5–8 minutes. The radiologist's edit-and-sign step is typically 2–6 minutes per study on top of that, depending on case complexity. End-to-end (push → signed PDF) is therefore in the 5–15 minute range for routine cases. Reruns (regenerating a draft after a correction) reuse cached image analysis and complete faster than first runs. ### What's the typical turnaround time for a signed report? URL: https://www.proradiq.com/ask/turnaround-time Last reviewed: 2026-06-26 **Short answer.** Routine cases: a signed report typically lands in 10–25 minutes from study arrival. Emergency cases are prioritised and usually under 15 minutes. AI draft is 3–5 minutes; radiologist edit-and-sign is 2–10 minutes depending on complexity. Turnaround time (TAT) on ProRadIQ has two components: AI draft latency and radiologist edit-and-sign time. - **AI draft**: 3–5 minutes for routine studies, up to 8 minutes for HRCT or multi-sequence MRI with priors. - **Edit-and-sign**: 2–6 minutes for plain films, 4–10 for routine CT/MRI, 10–20 for complex multi-system or oncology follow-up cases. End-to-end TAT for a centre using their own radiologist is therefore in the 10–25 minute range for routine work. For emergency cases the platform routes them to the top of the radiologist's worklist and flashes critical-finding callouts; emergency TAT is usually under 15 minutes. For centres on the telereporting model (where Dr. Harjot Singh signs), routine TAT is typically under 45 minutes and emergency TAT under 20 minutes during business hours, with an after-hours roster for nights and weekends. Specific TAT SLAs can be contracted. ### How do I push studies from Orthanc to ProRadIQ? URL: https://www.proradiq.com/ask/orthanc-lua-push Last reviewed: 2026-06-26 **Short answer.** Drop our provided Lua script into your Orthanc config, generate a centre-specific push token on the ProRadIQ admin, paste the token into the Lua, and every newly stored study auto-pushes over HTTPS within seconds. Orthanc — the popular open-source DICOM server — exposes a Lua scripting hook called `OnStoredInstance` that fires every time a new DICOM instance is received. ProRadIQ supplies a single Lua script that uses this hook to push each completed study to our ingest endpoint over HTTPS. Setup takes about 15 minutes: 1. Super-admin generates a per-centre push token in the ProRadIQ admin console. 2. Drop our `proradiq-push.lua` into Orthanc's `LuaScripts` directory. 3. Paste the token into the Lua's config block. 4. Restart Orthanc. From that point, every study the modality sends to Orthanc is auto-pushed to ProRadIQ, deduped per-organisation, and added to the worklist. No VPN, no fixed IP, no firewall holes — just outbound HTTPS. Multi-centre clients can run one centre per token under the same parent organisation, so all studies converge into a single worklist for the owning radiologist. ### What reporting standards does ProRadIQ follow? URL: https://www.proradiq.com/ask/reporting-standards Last reviewed: 2026-06-26 **Short answer.** Structured, modality-aware reports following RSNA / RANZCR templates where applicable, with bolded positive findings, a clear Impression section, and critical-finding callouts. Output is delivered as PDF, DOCX and FHIR DiagnosticReport. ProRadIQ's reports follow widely accepted radiology standards: - **Structured sections** — Technique, Comparison (when priors are available), Findings, Impression. Every modality has its own sub-section ordering (e.g. organ-by-organ for abdominal CT, level-by-level for spine MRI, BI-RADS for mammography). - **Templates** — RSNA RadReport-aligned and RANZCR templates available out of the box; centres can also define their own templates and assign them to study types. - **Visual emphasis** — section headings and positive findings are bolded. Normals remain unbolded so positives jump off the page. - **Critical findings** — surfaced at the top of the report with a clear callout, and audit-logged separately so a centre can prove their critical-finding-notification policy. - **Output formats** — signed PDF (with embedded signing radiologist's credentials), DOCX (for centres that further edit before delivery), and FHIR DiagnosticReport resource for HIS/EMR integration. - **Audit & integrity** — every signed report is SHA-256 hashed and linked into an immutable hash chain, so any later tampering is detectable. ## Pricing & onboarding ### How much does ProRadIQ cost? URL: https://www.proradiq.com/ask/pricing Last reviewed: 2026-06-26 **Short answer.** Pricing is per-study and bespoke to your modality mix and monthly volume. There is no setup fee for standard onboarding. Request a quote from harjot@proradiq.com or WhatsApp +91 81466 87177. ProRadIQ's pricing is per-study because that is the unit clients actually consume. A centre running 60 X-rays and 5 CTs a day has different economics from one running 30 MRIs and 100 follow-up oncology CTs. Factors that shape the quote: - Modality mix (X-ray, CT, MRI, mammography, PET-CT). - Monthly study volume — meaningful tiers around 500, 2,000 and 5,000 studies/month. - Whether you sign with your own radiologist (AI-assisted) or with Dr. Harjot Singh (telereporting). - Region and data-residency requirements. - Whether you need a custom template library, EMR integration or specific SLAs. Standard onboarding (token issuance, Orthanc Lua install, template setup) has no setup fee. Custom integrations are quoted separately. To get a quote, email harjot@proradiq.com or message +91 81466 87177 on WhatsApp — typical response within one business day. ### Is there a free trial? URL: https://www.proradiq.com/ask/free-trial Last reviewed: 2026-06-26 **Short answer.** Yes. Trials include onboarding credits that let you push and report real studies end-to-end before committing. Sign up at proradiq.com and we'll onboard you within one business day. We don't ask for production data or a contract on day one. A trial gives you onboarding credits sufficient to take real studies end-to-end — DICOM push or upload, AI draft, edit-and-sign in the workstation, signed PDF/DOCX download — across the modalities you actually run. The trial is set up like a real centre: your own organisation, your own push token, your own radiologist credentials. If you decide to continue, the same setup converts to a paid centre with no migration. If you decide not to, you leave with your data exported and the centre is decommissioned. To start, sign up at proradiq.com or message harjot@proradiq.com / +91 81466 87177 on WhatsApp. ### How long does onboarding take? URL: https://www.proradiq.com/ask/onboarding-time Last reviewed: 2026-06-26 **Short answer.** Most centres are live the same day. Standard onboarding — organisation setup, push tokens, Orthanc Lua install, radiologist credential verification — takes 60–90 minutes once your team is available. Onboarding is intentionally light: - **Organisation and centres** — we set up the parent organisation and any sub-centres (for multi-location clients). Around 10 minutes. - **Push tokens** — one per centre or per modality. Around 5 minutes per token. - **Orthanc / PACS integration** — drop our Lua script into Orthanc, or configure ProRadIQ as a C-STORE destination on your existing PACS. Around 15–30 minutes. - **Radiologist credential setup** — upload council registration and signing certificate; we validate it against the council registry. Around 10 minutes per radiologist. - **Template and modality routing** — out-of-the-box templates work immediately; custom templates can be added later. Most centres are live and pushing real studies within a single working session. EMR or HIS integrations (FHIR, HL7, custom REST) add 1–3 weeks depending on the EMR vendor. ## Security & data ### Where is patient data stored? URL: https://www.proradiq.com/ask/data-residency Last reviewed: 2026-06-26 **Short answer.** On managed cloud infrastructure with regional isolation. Default region is India for South Asia clients; EU/UK regions are available on request for GDPR deployments; UAE region for GCC. Patient data never leaves the contracted region. Patient data residency is contractual. By default, South Asia clients are pinned to India regions. EU and UK clients can be pinned to EU/UK regions to satisfy GDPR/UK GDPR transfer requirements. UAE clients can be pinned to UAE regions to satisfy DHA and local data law. Data at rest is AES-256 encrypted. Data in transit is TLS 1.2+. Every DICOM image is served via short-lived signed URLs scoped to the requesting user and case. Backups are encrypted and stored in the same region as the primary data. A configurable purge policy lets each organisation set how long signed reports and their underlying DICOM images are retained. Default is 7 years for signed reports (typical medico-legal retention) and 90 days for raw DICOM images, but both are tunable per contract.