How Is Osteomyelitis Diagnosed? From X-Rays to Biopsy

Diagnosing osteomyelitis relies on a layered approach that starts with plain X-rays and blood tests, then escalates to advanced imaging like MRI, and often requires a bone biopsy to confirm the infection and identify the responsible organism. No single test can reliably diagnose or rule out a bone infection on its own, which is why clinicians combine clinical findings, lab work, imaging, and sometimes tissue sampling into a composite picture. The path from suspicion to confirmed diagnosis varies depending on whether the infection is in a long bone, the spine, or beneath a diabetic foot ulcer, and the tools that work best shift accordingly.

Why X-Rays Come First

A plain radiograph is almost always the first imaging study ordered when osteomyelitis is suspected. It is cheap, fast, widely available, and provides a useful overview of the bone’s anatomy and any obvious problems like fractures, tumors, or foreign bodies that might explain the patient’s symptoms without invoking infection at all.¹ In spinal cases, for example, an X-ray may reveal collapsed disc spaces or endplate irregularities that raise a red flag even before advanced imaging is performed.

The catch is that X-rays are insensitive early on. Bone destruction generally needs to progress for ten days to two weeks before it becomes visible on a plain film, so a normal X-ray in the first week of symptoms means very little. Conventional radiography is better at confirming an established or chronic infection than catching one in its earliest stages.² Think of the X-ray as a screening tool that helps triage: if the film is clearly abnormal, the diagnosis moves forward quickly; if it is normal but clinical suspicion remains high, the next step is usually more sensitive imaging.

Blood Tests and Their Limits

Two inflammatory markers dominate the initial laboratory workup: the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Both rise when the body is fighting infection, and both are routinely drawn when osteomyelitis is on the table. But how much weight you can put on them depends heavily on the clinical context.

In patients with diabetes-related foot infections, an ESR above roughly 60 mm/h had a sensitivity of about 74% and a specificity of 56% for osteomyelitis, while a CRP above about 7.9 mg/dL was less sensitive (around 49%) but more specific (80%). When both markers were elevated beyond those thresholds, the likelihood of bone infection was high enough that clinicians were advised to strongly consider treatment.³ On the reassuring end, a very low ESR (below 30 mm/h) made osteomyelitis unlikely in that population.

In people without diabetes, the picture is less encouraging. One study of foot osteomyelitis in non-diabetic patients found that ESR and CRP had poor sensitivity and specificity, with the combined sensitivity of both markers together dropping to about a third. The researchers concluded these markers had little standalone diagnostic value in the non-diabetic foot.⁴ The practical takeaway: elevated inflammatory markers support a suspicion of osteomyelitis, and very low values can be somewhat reassuring, but normal blood work does not rule the diagnosis out. These tests are best used as one piece of a larger puzzle, not as a final answer.

Newer ratios derived from a routine blood count, like the neutrophil-to-lymphocyte ratio, have shown some promise in diabetic foot patients. Research has found that these ratios climb alongside the severity of the foot ulcer and are highest in patients who turn out to have osteomyelitis.⁵ Whether these ratios will eventually outperform traditional markers remains an open question, but they add another data point at essentially no extra cost.

MRI as the Imaging Workhorse

Magnetic resonance imaging is widely considered the best single imaging test for osteomyelitis. It is the most sensitive and specific imaging modality available for detecting bone infection, and it provides excellent detail about how far the infection has spread into surrounding soft tissues.² Unlike X-rays, MRI can detect the bone marrow edema that appears in the earliest stages of infection, well before the bone itself starts to break down visibly.

In lower-limb osteomyelitis, MRI has demonstrated a mean sensitivity around 89% and specificity around 88%.⁶ For vertebral osteomyelitis specifically, those numbers climb even higher: sensitivity, specificity, and accuracy have been estimated at roughly 97%, 92%, and 94%, respectively.¹ A hallmark finding is abnormal signal in the bone marrow on certain MRI sequences: areas that should appear bright become dark, reflecting the replacement of normal marrow by infectious and inflammatory tissue. Contrast-enhanced MRI adds further detail by highlighting areas of active inflammation and any abscess formation.⁷

MRI is not perfect. It can be hard to get in some settings (long wait times, high cost, claustrophobic patients, those with certain metal implants). And in post-surgical or post-traumatic bones, the marrow signal changes from surgery itself can mimic infection, making interpretation tricky. Still, when osteomyelitis is seriously suspected and the patient can undergo the scan, MRI is the imaging modality clinicians reach for first.

Where CT Scans and Dual-Energy CT Fit In

Computed tomography does not replace MRI for detecting osteomyelitis, but it fills important gaps. CT excels at evaluating cortical bone, detecting early bone destruction, and identifying a sequestrum, which is a piece of dead bone walled off inside the infected area. In assessing sequestrum formation specifically, CT is more accurate than both plain X-rays and MRI.⁸ It also provides helpful information about cortical destruction, the presence of a cloaca (a channel through which pus drains out of the bone), and the thickness of involucrum (new bone forming around the dead segment).

In the spine, CT can help when MRI findings are ambiguous. Degenerative disc changes sometimes produce MRI appearances that overlap with early infection. CT can clarify whether the vertebral endplates are actually eroded, which tilts the diagnosis toward osteomyelitis, or whether the changes are more consistent with normal wear and tear.¹

Dual-energy CT, a newer technology, has shown diagnostic performance approaching MRI for lower-limb osteomyelitis, with a sensitivity of about 89% and specificity around 73%. For detecting bone erosions specifically, dual-energy CT actually outperformed MRI in one comparative study.⁶ This makes it a viable alternative for patients who cannot undergo MRI, although it involves radiation exposure and still lags behind MRI for soft-tissue detail.

Nuclear Medicine Scans

When MRI is not available or its findings are equivocal, nuclear medicine imaging offers another route. These scans work by injecting a radioactive tracer that concentrates in areas of active infection or inflammation. FDG-PET, which uses a glucose-based tracer, is particularly useful: sensitivities above 95% and specificities ranging from about 75% to 99% have been reported for acute and subacute bone infections.⁹

For vertebral osteomyelitis, a meta-analysis comparing different imaging approaches found that PET/CT had a sensitivity of 93% and specificity of 80%, while standard technetium bone scans were sensitive (86%) but much less specific (39%). The combination of gallium and technetium scans together improved specificity considerably, reaching about 92%.¹⁰ Nuclear scans are especially valuable when metallic hardware is present, because metal creates artifacts on both MRI and CT that can obscure the very area you need to evaluate.

Ultrasound in Children

Ultrasound is not a go-to tool for osteomyelitis in adults, but it fills a useful niche in pediatric cases. In children with acute osteomyelitis, infection often spreads from the bone to the space beneath the periosteum (the membrane covering the bone), forming a subperiosteal abscess. Ultrasound can detect these fluid collections early and without radiation, which matters when the patient is a child. One small pediatric study found periosteal thickening or elevation with subperiosteal fluid in 80% of confirmed cases.¹¹

Beyond diagnosis, ultrasound can guide treatment. Case reports have demonstrated the feasibility of ultrasound-guided percutaneous aspiration of subperiosteal abscesses as a minimally invasive option in selected pediatric patients, potentially sparing them open surgery.¹² Earlier work confirmed that ultrasonography could detect these abscesses reliably enough that the findings were later confirmed surgically or by other imaging.¹³ Ultrasound does not image bone itself well, so it cannot characterize the infection within the marrow. But as a rapid, bedside-available, radiation-free first look in a febrile child with a swollen limb, it earns its place in the diagnostic lineup.

The Probe-to-Bone Test for Diabetic Feet

Diabetic foot ulcers present a diagnostic challenge all their own. The skin is already broken, soft tissue may be compromised, and the underlying bone can be difficult to image clearly due to chronic changes from neuropathy. Before ordering an MRI, clinicians often perform a deceptively simple bedside maneuver: pressing a sterile metal probe into the wound to see if it touches bone. If it does, the probability of osteomyelitis rises sharply.

A systematic review of the probe-to-bone test reported pooled sensitivity of 87% and specificity of 83%, concluding that the test can reliably rule in osteomyelitis in high-risk patients and rule it out in low-risk ones.¹⁴ Another study found similar sensitivity (87%) with even higher specificity (91%) and a negative predictive value of 98%, meaning that a negative probe-to-bone result was highly reassuring.¹⁵ This test costs nothing, takes seconds, and provides immediately actionable information. It does not replace imaging or biopsy, but it helps clinicians decide how aggressively to pursue them.

Bone Biopsy and Why Cultures Matter

Imaging can tell you that osteomyelitis is probably present, but it cannot tell you which organism is causing it. That distinction matters enormously for treatment, because the antibiotic that clears a staphylococcal infection will not touch a tuberculosis-related bone infection, and a fungal osteomyelitis requires a completely different drug class. Surface swabs from ulcers or draining sinuses are often misleading, growing bacteria that colonize the wound surface rather than the ones actually infecting the bone. Bone cultures are necessary to identify the true pathogens.¹⁶

Bone biopsy can be performed surgically or percutaneously with image guidance (usually CT or fluoroscopy). One study of imaging-guided biopsies found that among confirmed osteomyelitis cases, only about a third yielded positive cultures. The single factor that significantly improved the odds of a positive culture was aspiration of a meaningful volume of purulent fluid (at least 2 mL), where five of six such cases grew an organism.¹⁷ Neither prior antibiotic use, fever, elevated white blood cell count, nor biopsy needle size significantly affected the culture yield in that study.

That last finding is worth emphasizing. A common clinical dilemma is whether to delay a biopsy until antibiotics have been stopped, hoping to improve the chance of growing the organism. Research on vertebral osteomyelitis found that receiving antibiotics before biopsy did not negatively affect pathogen recovery, and the authors cautioned that antibiotic exposure alone should not be a reason to skip the biopsy.¹⁸ When a patient is septic or deteriorating, waiting days for antibiotics to wash out before performing a biopsy is not always practical or safe, and the evidence suggests it is not necessary.

Diagnosing Infection Around Metal Implants

Osteomyelitis around orthopedic hardware, whether it is a plate, a screw, or a joint replacement, is a separate diagnostic headache. These infections are often caused by low-virulence organisms like Staphylococcus epidermidis that form biofilms on the metal surface and cause a smoldering, hard-to-detect infection. The clinical presentation can be nothing more than persistent pain and early loosening of the implant, without fever, redness, or other classic signs of infection.¹⁹ Imaging is hampered by metal artifacts, and standard tissue cultures may miss the organisms hiding in the biofilm.

Sonication, where the removed implant is placed in fluid and subjected to ultrasound waves to shake bacteria loose from the surface, has emerged as a way to improve detection. One study found that sonication fluid culture detected the infecting organism in about 90% of cases, compared with roughly 57% for standard tissue cultures taken from around the implant. Sonication was also better at identifying mixed infections and recovered organisms more often in patients who had already been on antibiotics.²⁰ That said, sonication is a complement to biopsy cultures, not a replacement. In one animal model of chronic implant infection, very few bacteria were found on the implant surface despite pronounced infection in the surrounding tissue, underscoring that both tissue sampling and implant sonication may be needed for a complete picture.²¹

When MRI Looks Like Something Else

One of the trickiest parts of diagnosing osteomyelitis is that its imaging appearance can overlap with conditions that are not infections at all. Two mimics come up repeatedly in clinical practice.

Ewing sarcoma, a bone tumor that affects children and young adults, can produce MRI findings that look alarmingly similar to osteomyelitis: bone marrow signal changes, periosteal reaction, and soft-tissue involvement. The most reliable differentiating feature is the margin of the bone lesion. In one study, every patient with Ewing sarcoma had a sharp, well-defined lesion margin, while none of the osteomyelitis patients did. Cystic or necrotic areas within the lesion were also much more common in the tumor. These distinctions were best seen on specific MRI sequences and required careful radiologic interpretation.²²

In the diabetic foot, the main imaging confounder is Charcot neuroarthropathy (neuropathic joint disease). Both conditions cause bone marrow edema, bone destruction, and soft tissue swelling on MRI. Some patterns help distinguish them: osteomyelitis tends to involve a single bone directly beneath an ulcer, whereas Charcot changes tend to affect multiple bones of the midfoot. Findings like thin rim enhancement of a joint effusion, subchondral cysts, and loose bodies within the joint point toward neuropathic disease rather than infection.²³ Even with these clues, the overlap can be so substantial that biopsy becomes the only way to settle the question.

Pediatric Risk Scoring

Children develop osteomyelitis differently from adults. The infection is usually blood-borne (hematogenous), tends to seed in the fast-growing ends of long bones, and can progress rapidly. Clinicians evaluating a child with a possible bone infection face an urgent triage question: does this child need advanced imaging and potentially surgery, or is the risk low enough to watch and wait?

A recently developed four-point risk score aims to help with that decision. It assigns one point each for illness lasting more than three days, a history of fever or a temperature of at least 38°C in the emergency department, a CRP above 2.0 mg/dL, and an ESR above 25 mm/h. When three or more of those factors are present, the score maximizes diagnostic accuracy at about 84%. Children with zero points are highly unlikely to have acute hematogenous osteomyelitis, with a sensitivity of 99% for identifying those who do.²⁴ Scoring systems like this are not diagnostic endpoints; they are decision aids that help clinicians allocate scarce resources (MRI time, sedation, operating room availability) to the children who need them most while safely observing those at low risk.

Vertebral Osteomyelitis and the Biopsy Dilemma

Spine infections deserve separate mention because the diagnostic approach has some unique wrinkles. Patients with vertebral osteomyelitis often present with weeks or months of back pain and no obvious source. By the time the infection is suspected, many have already been started on empiric antibiotics. The question of whether to biopsy the vertebra to identify the organism is a real clinical tension, because CT-guided spine biopsies are technically demanding and have a meaningful false-negative rate.

If blood cultures are already positive, which happens in a substantial fraction of vertebral osteomyelitis cases, biopsy may be unnecessary because the bloodstream isolate is usually the same organism infecting the bone. When blood cultures are negative, image-guided biopsy becomes more important, but the yield varies. As noted above, prior antibiotics do not appear to reduce the chance of growing an organism from a biopsy specimen, which means clinicians should not feel compelled to create a dangerous antibiotic-free window just to improve culture odds.¹⁸ MRI remains the preferred imaging modality for confirming the diagnosis itself, with sensitivity and specificity estimates that exceed most other imaging options for spinal infection.¹

1
PubMed Central. Diagnosis of vertebral osteomyelitis
2
Europe PMC. Radiographic imaging in osteomyelitis: the role of plain radiography, computed tomography, ultrasonography, magnetic resonance imaging, and scintigraphy
3
PubMed Central. What are the Optimal Cutoff Values for ESR and CRP to Diagnose Osteomyelitis in Patients with Diabetes-related Foot Infections?
4
PubMed. Diagnostic Utility of Erythrocyte Sedimentation Rate and C-Reactive Protein in Osteomyelitis of the Foot in Persons Without Diabetes
5
PubMed Central. Comparison of neutrophil to lymphocyte ratio, platelet to lymphocyte ratio, and C-reactive protein in predicting osteomyelitis in patients with diabetic foot ulcer
6
PubMed Central. Osteomyelitis of the Lower Limb: Diagnostic Accuracy of Dual-Energy CT versus MRI
7
PubMed Central. An ominous cause of back pain: Vertebral Osteomyelitis
8
Insights into Imaging. Osteomyelitis: what radiologists should know
9
Clinical Radiology. Nuclear medicine imaging of bone infections
10
PubMed. The use of MRI, PET/CT, and nuclear scintigraphy in the imaging of pyogenic native vertebral osteomyelitis: a systematic review and meta-analysis
11
PubMed Central. Role of ultrasound in the diagnosis of paediatric acute osteomyelitis
12
PubMed. Subperiosteal abscess in pediatric acute osteomyelitis: ultrasound-guided percutaneous aspiration in three cases
13
PubMed. Early detection of subperiosteal abscesses by ultrasonography. A means for further successful treatment in pediatric osteomyelitis
14
Clinical Infectious Diseases. Diagnostic Accuracy of Probe to Bone to Detect Osteomyelitis in the Diabetic Foot: A Systematic Review
15
PubMed. Probe-to-bone test for diagnosing diabetic foot osteomyelitis: reliable or relic?
16
Europe PMC. Osteomyelitis: approach to diagnosis and treatment
17
American Journal of Roentgenology. Imaging-guided bone biopsy for osteomyelitis: are there factors associated with positive or negative cultures?
18
PubMed Central. The Impact of Prebiopsy Antibiotics on Pathogen Recovery in Hematogenous Vertebral Osteomyelitis
19
PubMed. Diagnosis and treatment of implant-associated septic arthritis and osteomyelitis
20
PubMed Central. Improved diagnosis of infection associated with osteosynthesis by use of sonication of fracture fixation implants
21
PubMed Central. Biofilm and the effect of sonication in a chronic Staphylococcus epidermidis orthopedic in vivo implant infection model
22
PubMed. Ewing sarcoma versus osteomyelitis: differential diagnosis with magnetic resonance imaging
23
PubMed. Osteomyelitis of the lower extremity: pathophysiology, imaging, and classification, with an emphasis on diabetic foot infection
24
Pediatrics. A Novel Risk Score to Guide the Evaluation of Acute Hematogenous Osteomyelitis in Children