Ultrasound-guided ulnar-to-radial approach to the carpal tunnel — median nerve-sparing technique for CTS
Carpal tunnel syndrome (CTS) — compressive median neuropathy at the wrist — is the most common entrapment neuropathy encountered in clinical practice. Confirming the diagnosis with a combination of clinical examination, electrodiagnostic testing, and increasingly point-of-care ultrasound is essential before proceeding to image-guided corticosteroid injection.
Patients with CTS classically report nocturnal or activity-related paresthesia and numbness in the median nerve distribution — thumb, index, middle finger, and the radial half of the ring finger — often relieved by shaking the hand ("flick sign"). Advanced disease produces thenar muscle weakness and atrophy (abductor pollicis brevis) and permanent sensory loss.
Provocative clinical tests, while individually only moderately sensitive/specific, remain part of the standard evaluation: • Phalen's test: wrist flexion for 60 seconds reproducing paresthesia (sensitivity ~68%, specificity ~73%) • Tinel's sign: percussion over the median nerve at the wrist crease reproducing distal tingling (sensitivity ~50%, specificity ~77%) • Durkan's carpal compression test: direct manual compression over the carpal tunnel, generally more sensitive than Phalen/Tinel
Because clinical signs alone have imperfect accuracy, electrodiagnostic testing remains the reference standard for confirming median nerve dysfunction and grading severity before invasive treatment.
Nerve conduction studies (NCS), per American Association of Neuromuscular & Electrodiagnostic Medicine (AANEM) practice parameters, remain the gold-standard confirmatory test: prolonged distal motor latency (>4.2 ms) and/or distal sensory latency (>3.5 ms) across the wrist, with severity graded from mild (sensory latency prolongation only) to moderate (added motor latency prolongation) to severe (absent sensory/motor responses, denervation on needle EMG of the abductor pollicis brevis).
High-resolution ultrasound has emerged as a complementary, increasingly primary, diagnostic and pre-procedural tool: measurement of median nerve cross-sectional area (CSA) at the level of the pisiform bone, using the standardized technique described by Wilson, is now well validated, with a CSA threshold >10–12 mm² carrying a sensitivity of approximately 87–90% and specificity of 80–95% for CTS across multiple studies and meta-analyses. Ultrasound additionally screens for structural causes of secondary CTS (ganglion cyst, flexor tenosynovitis, persistent median artery, bifid median nerve, space-occupying lesions) that may change management.
Combining electrodiagnostic testing with ultrasound CSA measurement improves diagnostic accuracy over either modality alone, and ultrasound has the added advantage of directly informing the anatomic approach and safety margins for a subsequent image-guided injection.
The carpal tunnel is a rigid osteofibrous canal at the wrist bounded dorsally by the carpal bones and volarly by the flexor retinaculum (transverse carpal ligament), containing the median nerve and nine flexor tendons. Short-axis ultrasound imaging at the level of the pisiform provides the clearest, most reproducible view for both diagnosis and needle guidance.
A high-frequency linear array transducer (12–18 MHz) is used, given the superficial location (2–15 mm deep) and small caliber of the target structures. The wrist is positioned in slight extension, often supported on a rolled towel, with fingers relaxed.
The transducer is placed in short axis (transverse) across the volar wrist at the level of the pisiform bone (the most reliable, reproducible landmark, per Wilson's original description and subsequent consensus protocols). At this level, from superficial to deep, the operator identifies:
• Skin and subcutaneous fat • Palmaris longus tendon (present in ~85% of individuals), lying superficial and central • Flexor retinaculum (transverse carpal ligament), a hyperechoic band spanning from the pisiform/hook of hamate (ulnar) to the scaphoid/trapezium (radial), forming the volar roof of the tunnel • Median nerve, typically ovoid to flattened, located just deep to the flexor retinaculum, radial to the palmaris longus, showing the characteristic hyperechoic epineurium surrounding hypoechoic fascicles ("honeycomb" pattern) • Flexor tendons (4 flexor digitorum superficialis, 4 flexor digitorum profundus, 1 flexor pollicis longus) — hyperechoic, fibrillar, and mobile with finger flexion/extension, used to dynamically distinguish tendon from nerve • Carpal bones (pisiform ulnarly, scaphoid tubercle radially) forming the deep and lateral boundaries
CSA is measured by tracing the inner margin of the hyperechoic epineurium (the "ant-eater" or "honeycomb" nerve, excluding the surrounding hyperechoic rim itself) using the ellipse or free-trace tool on the ultrasound system, at the pisiform level. Reference values:
• Normal: <9–10 mm² • Mild CTS: 10–13 mm² • Moderate CTS: 13–17 mm² • Severe CTS: >17 mm², often with flattening and loss of normal fascicular architecture
Operators should also screen for anatomic variants relevant to needle safety and injection planning: a bifid median nerve (present in ~2–3% of the population, often with a persistent median artery running between the two nerve divisions) and an anomalous thenar motor branch, both of which alter the safest needle trajectory.
Color Doppler interrogation for a persistent median artery — which, when present, frequently runs directly adjacent to or between a bifid median nerve — should be performed before selecting the final needle trajectory, since this vessel is not reliably predicted by surface anatomy alone.
The ulnar-to-radial in-plane approach under short-axis visualization is the most widely validated technique for carpal tunnel injection, allowing the needle to approach the median nerve tangentially from its ulnar (deep-to-tendon) aspect rather than head-on, minimizing the risk of direct nerve puncture.
1. Skin antisepsis and sterile probe cover/gel are applied as for any image-guided procedure.
2. Needle entry point: marked on the ulnar side of the wrist, at the level of the transducer's ulnar edge, typically just radial to the flexor carpi ulnaris tendon and ulnar to the neurovascular bundle of the ulnar nerve/artery (which lie further ulnarly and are avoided by staying close to the pisiform).
3. Needle advancement: a 25–27G, 1.5 inch needle is advanced in-plane with the transducer, traveling from ulnar to radial, in the same fascial plane immediately deep to the flexor retinaculum and superficial to the flexor tendons — this "safe corridor" is wider and technically easier to enter from the ulnar side than attempting a direct dorsal-to-volar approach onto the nerve.
4. Tangential approach to the nerve: rather than aiming directly at the nerve, the needle tip is directed to pass just deep to the retinaculum and adjacent to — not into — the median nerve, using the flexor tendons as the deep safety boundary.
5. Confirmation of extraneural position: before injecting any volume, a small test bolus (0.5 mL) is given to confirm fluid spreads in the correct plane around, not within, the nerve.
The technique's central safety principle is maintaining the needle tip in the extraneural, subretinacular plane throughout the injection, avoiding several nearby vulnerable structures:
• Median nerve: the primary structure to avoid puncturing; any high resistance to injection, or patient-reported sharp paresthesia/electric shock sensation on needle advancement or injection, mandates immediate withdrawal and repositioning — a cardinal rule shared with all peripheral nerve block/injection techniques • Ulnar nerve and artery: located more ulnarly within Guyon's canal; avoided by keeping the entry point and trajectory closer to the radial aspect of the ulnar-sided structures • Persistent median artery / bifid median nerve: when present, requires trajectory modification, often approaching more superficially or choosing an alternative radial-to-ulnar approach depending on individual anatomy • Flexor tendons: puncture is generally inconsequential but should be avoided by staying in the plane just deep to the retinaculum rather than advancing all the way to the carpal bones
Unlike joint injections where firm bony contact confirms an endpoint, in carpal tunnel injection any resistance or paresthesia signals a potential intraneural or intratendinous position — the needle must never be forced, and real-time visualization of the tip relative to the nerve margin is maintained continuously rather than relying on a tactile endpoint.
Correct needle placement is confirmed by visualizing an anechoic (fluid) plane expanding around the median nerve and lifting it away from the overlying flexor retinaculum — a technique termed hydrodissection, which is both a real-time confirmation of extraneural placement and, increasingly, a therapeutic maneuver in its own right.
As the test bolus and subsequent injectate are delivered, real-time ultrasound should show:
• A progressively expanding anechoic (black) fluid collarette forming around the circumference of the median nerve, most visible on its superficial (retinacular) and deep (tendon-facing) aspects • The nerve appearing to "float" and lift slightly away from the undersurface of the flexor retinaculum as fluid dissects the plane • No focal ballooning or swelling within the nerve's own fascicular substance, which would indicate inadvertent intrafascicular injection — an absolute stop signal requiring immediate cessation and needle repositioning
Injection proceeds slowly and incrementally, with the sonographer continuously confirming the fluid plane remains perineural (around the nerve) rather than intraneural, adjusting the needle tip position by small increments as needed to redirect the leading edge of fluid spread circumferentially around the nerve.
A typical therapeutic injectate for CTS combines:
• Corticosteroid: triamcinolone acetonide 20–40 mg (or equivalent dose of methylprednisolone/betamethasone), for local anti-inflammatory effect on the inflamed/thickened synovium and nerve • Local anesthetic: 1% lidocaine, 0.5–1.5 mL, both to confirm correct plane by immediate symptom relief and to reduce injection-related discomfort • Total volume: typically 2–4 mL combined
An increasingly used variant is saline hydrodissection, in which a larger volume (5–10 mL) of normal saline (with or without a small dose of corticosteroid or platelet-rich plasma in some protocols) is used specifically to mechanically separate the nerve from the surrounding fibrotic or thickened flexor retinaculum and tendons — a technique with growing evidentiary support as a standalone or adjunctive treatment, distinct from simple corticosteroid delivery, aiming to reduce adhesions and improve nerve gliding.
Comparative trials (e.g., Wu et al. and subsequent meta-analyses) suggest ultrasound-guided perineural hydrodissection — with or without corticosteroid — produces symptom and electrophysiological improvement comparable to or exceeding blind corticosteroid injection, with a lower theoretical risk profile since larger fluid volumes further displace the nerve from the needle tip as the procedure progresses.
Ultrasound guidance has substantially improved the accuracy and safety of carpal tunnel corticosteroid injection compared to traditional landmark (blind) technique, which carries a meaningful risk of inadvertent intraneural injection given the small, crowded anatomy of the tunnel — translating into both better symptom outcomes and a lower complication rate.
Multiple studies comparing ultrasound-guided to landmark-guided (blind) carpal tunnel injection consistently favor image guidance:
• Ultrasound-guided injections achieve correct perineural (extraneural, subretinacular) placement in approximately 97–100% of attempts across published series, compared to considerably more variable and generally lower accuracy (roughly 70–90%, with some cadaveric studies reporting intraneural or extra-tunnel misplacement in up to 15–20% of blind attempts) for landmark-guided technique • Randomized comparative trials (e.g., Smith et al. and subsequent meta-analyses in journals such as Muscle & Nerve and Archives of Physical Medicine and Rehabilitation) generally demonstrate ultrasound-guided injection produces superior or at least equivalent short-term (4–12 week) symptom relief and functional outcome scores (Boston Carpal Tunnel Questionnaire) compared to blind injection • The American Institute of Ultrasound in Medicine (AIUM) and American College of Radiology (ACR) appropriateness criteria support ultrasound guidance for carpal tunnel injection given the proximity of the median nerve to the needle path and the well-documented accuracy advantage
The American Academy of Orthopaedic Surgeons (AAOS) clinical practice guideline on carpal tunnel syndrome includes corticosteroid injection as a recommended non-surgical treatment option, particularly for mild-to-moderate disease or as a bridge/diagnostic step before considering surgical release.
A single ultrasound-guided corticosteroid injection typically produces symptom improvement within days, with peak effect at 2–6 weeks. Reported rates of clinically meaningful relief (typically defined as significant reduction in the Boston Carpal Tunnel Symptom Severity Scale or equivalent) at 3 months range from roughly 70–80% for mild-to-moderate CTS, though relief often wanes over 6–12 months, and a substantial proportion of patients (particularly those with moderate-severe electrodiagnostic findings or long symptom duration) eventually proceed to surgical carpal tunnel release.
Injection is generally considered most durable and appropriate for mild-to-moderate CTS without significant thenar atrophy or fixed sensory loss; severe CTS with denervation on EMG is a relative indication favoring earlier surgical referral rather than repeated injection, given the risk of irreversible axonal loss.
Ultrasound-guided carpal tunnel injection has a favorable overall safety profile, though specific risks require attention given the confined anatomy:
• Transient paresthesia/dysesthesia: uncommon but can occur from needle proximity to the nerve even without frank puncture; usually resolves within minutes to hours • Intraneural injection: rare with ultrasound guidance (in contrast to blind technique) given continuous real-time visualization and the cardinal rule of stopping immediately if resistance or sharp pain occurs • Flexor tendon injury/partial tenotomy: rare, more of a theoretical risk with repeated injections directly through tendon substance • Infection: rare (<0.1%) with sterile technique • Post-injection flare and local corticosteroid-related skin/soft-tissue changes (fat atrophy, depigmentation): occasional, more common with superficial extravasation • Incomplete or transient relief necessitating surgical release: common in moderate-severe disease, representing treatment failure rather than a procedural complication per se
A structured decision pathway — clinical suspicion, confirmatory NCS/EMG and ultrasound CSA measurement, trial of ultrasound-guided injection for mild-moderate disease, and early surgical referral for severe/denervated cases — reflects the current evidence-based approach endorsed across AANEM electrodiagnostic guidelines and AAOS clinical practice recommendations for carpal tunnel syndrome.