Home / Clinical Guideline Hubs / Chapter 5.9
Section 5 — General NICU Care / Supportive Care Pending expert review v1.0 · July 2026

Chapter 5.9 — Neonatal Diagnostic Imaging: Modality Selection, Safety, and Reading the Films

Board-review synthesis · pending expert review

Educational guideline — verify locally. A reasoning/reference guide (no images); not a substitute for a radiologist's report. Cross-references: RDS/air leak (Section 6), NEC/obstruction (Section 11), IVH/HIE/PVL (Section 9), ROP (02-07).

1. Clinical Overview

Clinical Overview

Imaging is where much of neonatal reasoning becomes visible, but the newborn is small, radiosensitive, and often too unstable to travel, so the modality chosen must balance information gained against the cost of obtaining it. The discipline reduces to one habit: ask what question the image must answer, then choose the least costly modality that can answer it — and if no plausible result would change management, the right image is often none.

2. Key Clinical Points

Key Clinical Points
  • Modalities: radiography (portable, air/bone/line position; ionising radiation, poor soft-tissue detail); ultrasound (portable, real-time, radiation-free, no sedation, uses fontanelle/thin-wall windows; operator-dependent, blocked by bone/gas); MRI (best soft-tissue/brain detail, prognostic, no radiation; needs transport/time/sedation); CT (fast/detailed but large dose — use sparingly); plus fluoroscopic contrast (malrotation, distal obstruction) and echocardiography (definitive cardiac test).
  • Radiation is uniquely costly to the newborn: dividing (radiosensitive) tissues, long remaining lifespan, small body (a chest field may hit the abdomen/gonads), and many repeated films. ALARA in practice: image only when the result changes management, prefer ultrasound, collimate, position carefully to avoid repeats, and question each film against observation.
  • Non-radiation costs: thermal instability, transport risk (dislodged lines/tubes, hypothermia, instability), handling/pain (instability precipitates IVH — cluster care), and sedation/contrast risks. Choose the least costly modality that answers the question.
  • Chest film: read systematically (technique → lines/tubes → volumes → parenchyma → pleura → heart/vessels → bones/upper abdomen). Patterns: diffuse granular/ground-glass + air bronchograms + low volume = RDS (but pneumonia, esp. GBS, is indistinguishable → cover for infection); streaky perihilar + fluid in fissures = TTN; patchy coarse infiltrates + hyperinflation = meconium aspiration. Pitfalls: normal thymus mimics cardiomegaly; underinflated film magnifies the heart. In a deteriorating ventilated infant, read for air leak first (decompress a suspected tension pneumothorax before imaging).
  • Abdominal film: neonatal small and large bowel cannot be reliably distinguished (a contrast study localises distal obstruction). Count dilated loops: few + no distal gas = proximal (double bubble = duodenal atresia + polyhydramnios → trisomy 21); many loops = distal → contrast enema (distinguishes and can treat meconium ileus/plug).
  • Bilious vomiting is the most time-critical abdominal decision: a NORMAL plain film does not exclude malrotation with midgut volvulus (often normal early) — obtain an urgent upper GI contrast study (abnormal duodenojejunal junction ± corkscrew) and involve surgery, not observation.
  • NEC/perforation: findings in ascending severity — nonspecific dilatation/disordered gas, a fixed loop across serial films, pneumatosis intestinalis (specific sign), portal venous gas, and free air (perforation). Detect free air with a left lateral decubitus or cross-table lateral film in an unstable infant. Spontaneous intestinal perforation gives free air WITHOUT pneumatosis/inflammation.
  • Neurologic imaging: cranial ultrasound screens the preterm brain at the cot side (IVH, ventricular size, cystic white matter injury) but is blind to the periphery, posterior fossa/cerebellum, and diffuse non-cystic white matter injury — a normal scan does not exclude injury or predict good outcome. MRI defines injury pattern/extent (deep grey/perirolandic in acute profound HIE vs watershed in prolonged partial; stroke, malformation, diffuse injury) and is the best imaging predictor of outcome (timed ~end of first week).
  • Lines/tubes (the imaging task that changes management fastest — read first): ETT mid-trachea above the carina (too far → right main bronchus → left lung collapse/right hyperinflation → withdraw; neck flexion advances it). UVC tip at the IVC–right atrial junction OUTSIDE the heart (intracardiac → arrhythmia/perforation/tamponade; hepatic/portal → necrosis with hypertonic fluids). UAC tip high (descending aorta above coeliac/mesenteric/renal origins) or low (below them, above bifurcation), avoiding the intermediate position (thrombosis/embolisation). PICC tip in a central vein outside the cardiac silhouette. Feeding tube coiled in the upper oesophagus = oesophageal atresia.
  • Skeletal: birth fractures (clavicle commonest, ± brachial plexus injury; heal well); metabolic bone disease of prematurity (under-mineralisation, fractures from handling); skeletal dysplasias (formal survey; lethal vs non-lethal often turns on thoracic-cage adequacy); congenital syphilis periostitis. Concerning fracture patterns (differing ages, metaphyseal/posterior rib, inconsistent history) obligate a protocol-driven child-protection process — but metabolic bone disease/genetic fragility can mimic inflicted injury; never act on impression alone.

3. References to Verify

Back to Clinical Guideline Hubs