Tuberculosis CT
|
Tuberculosis Microchapters |
|
Diagnosis |
|---|
|
Treatment |
|
Drug-Susceptible and Drug-Resistant Tuberculosis(ATS 2025 Guidelines) |
|
Case Studies |
|
Tuberculosis CT On the Web |
|
American Roentgen Ray Society Images of Tuberculosis CT |
Editor-In-Chief: C. Michael Gibson, M.S., M.D. [1]; Associate Editor(s)-in-Chief: Mashal Awais, M.D.[2];Sophia Saad, Associate Editor - WikiDoc [3] Alejandro Lemor, M.D. [4]
Tuberculosis CT
Overview
Computed tomography is a second-line, problem-solving modality in tuberculosis, not a first-line diagnostic test. Chest radiography remains the initial imaging study; CT is added when the radiograph is normal, equivocal, or discordant with clinical suspicion, when complications or extrapulmonary disease are suspected, or when activity must be assessed. The principal advantage of CT over radiography is increased specificity: it resolves findings that radiography cannot, particularly endobronchial spread (tree-in-bud nodules), cavitation, and necrotic lymphadenopathy.[1] CT never establishes or excludes a diagnosis of TB; bacteriologic confirmation remains mandatory. This microchapter covers CT indications, thoracic and extrathoracic findings, activity assessment, and the correlation between CT features and infectiousness. Chest radiographic findings, MRI-preferred sites (CNS and spine), and PET-CT are covered in their own microchapters.
Indications for CT
ACR Appropriateness Criteria state that the major advantage of CT is increasing the specificity of the diagnosis of TB; CT better demonstrates cavitation and endobronchial spread with tree-in-bud nodules and is helpful when chest radiography does not show classic findings. CT findings help predict AFB smear positivity, and even in smear-negative patients CT may indicate the risk of culture positivity when features consistent with active TB are present. CT may be of value in a severely immunocompromised patient with a normal or near-normal radiograph by revealing abnormal lymph nodes or subtle parenchymal disease. CT is often not necessary in the acute setting when TB is already suspected and appropriate precautions and testing are under way; the decision to isolate and test should never await CT.[1]
Consolidated indications:
- Normal or equivocal chest radiograph with persisting clinical suspicion
- Severe immunosuppression (advanced HIV, anti-TNF therapy, transplant, hematologic malignancy), where CT may reveal abnormal nodes or subtle parenchymal disease despite a near-normal film[2]
- Assessment of disease activity when radiographic findings are indeterminate
- Suspected complications (empyema, bronchopleural fistula, aspergilloma, broncholithiasis, fibrosing mediastinitis, airway stenosis)
- Suspected extrapulmonary or disseminated disease
- Image-guided sampling (percutaneous nodal, pleural, bone, or abscess biopsy)
- Suspected drug-resistant or treatment-refractory disease requiring extent mapping before surgical consideration
- Emerging and unvalidated: ACR notes CT may have a role in identifying patients with latent TB at risk for reactivation, but this is not established practice and no randomised evidence supports it[1][3]
Radiation dose and protocol selection
Radiation exposure is a material consideration because TB patients frequently require repeated follow-up imaging; conventional follow-up chest CT carries an effective dose of approximately 7 mSv per examination.[4] In a prospective study of 59 patients with known or suspected pulmonary TB, ultralow-dose CT reconstructed with knowledge-based iterative model reconstruction achieved a mean effective dose of 0.28 ± 0.02 mSv — a 59% reduction versus low-dose CT (0.69 ± 0.15 mSv) — with significantly better subjective visualization of centrilobular nodules, consolidation, tree-in-bud, and cavity, at the cost of blurred edges.[5] Low-dose and ultralow-dose protocols should be the default where serial imaging is anticipated and in children and young adults.
Thoracic CT findings
Findings of active disease
High-resolution CT features most characteristic of activity are centrilobular nodules and the tree-in-bud pattern, reflecting caseous material and inflammatory exudate filling terminal and respiratory bronchioles — the direct imaging correlate of endobronchial spread. In an HRCT series of 32 patients with active and 34 with inactive pulmonary TB, centrilobular lesions (n=29), tree-in-bud appearance (n=23), and 5–8 mm macronodules (n=22) predominated in active disease, whereas fibrotic lesions, bronchovascular distortion, emphysema, and bronchiectasis characterized inactive disease. HRCT clearly separated old fibrotic lesions from new active lesions and demonstrated early bronchogenic spread.[6]
Core active features:
- Centrilobular micronodules and tree-in-bud opacities — endobronchial spread; the single most useful activity marker
- Cavitation — thick-walled, irregular margins, upper lobe and superior segment predominance; a marker of high bacillary burden
- Consolidation — often patchy and ill-defined, may be lobar
- Ground-glass and ill-defined nodules
- Necrotic lymphadenopathy — low-attenuation centers with rim enhancement on contrast-enhanced CT; highly suggestive in the correct clinical context
Findings of inactive disease and sequelae
Fibrotic bands, architectural distortion, traction bronchiectasis, calcified nodules and nodes, emphysematous and cicatricial change, and pleural thickening predominate. These findings appear as fibronodular apical and upper-zone opacity. Radiographic stability over 6 months is a conventional, though not formally validated, marker of inactivity, and no CT-specific stability interval has been established against a bacteriologic reference standard.[7] Residual cavities may persist after microbiologic cure and do not by themselves indicate active disease or warrant re-treatment.
Miliary tuberculosis
Innumerable 1–2 mm nodules in a random distribution throughout both lungs, reflecting hematogenous dissemination; often accompanied by multiorgan involvement and more common in immunocompromised hosts.[8] Miliary nodules may be detectable on CT before they are visible radiographically.
CT features and infectiousness
CT features correlate with smear positivity and therefore with transmission risk, though CT does not replace microbiology. Among 108 patients with active pulmonary TB, centrilobular micronodules (63% vs 38%), tree-in-bud opacities (63% vs 33%), consolidation (98% vs 81%), and cavitation (86% vs 33%; P<0.001) were all significantly more frequent in AFB smear-positive than smear-negative patients, and both frequency and extent rose in a graded fashion with increasing smear grade. These features did not differ significantly between culture-positive and culture-negative patients.[9] In a cohort of 189 hospitalized adults, multivariate predictors of smear positivity were consolidation (OR 2.52; 95% CI 1.18–5.41; P=.02), lymphadenopathy (OR 1.95; 95% CI 1.03–3.70; P=.04), and multilobar involvement (OR 2.80; 95% CI 1.08–7.21; P=.03); the resulting predictive model achieved an area under the ROC curve of only 0.724, so CT should be used to modify rather than determine pretest probability of infectiousness.[10] Even in smear-negative patients, CT findings consistent with active TB raise the probability of culture positivity.
CT activity grading and negative predictive value
Structured CT activity categorization performs well as a rule-out tool. Among 650 patients with suspected pulmonary TB stratified as definitely active, probably active, indeterminate, or probably inactive, culture-confirmation rates were 61.6%, 60.7%, 4.3%, and 0% respectively. Treating definitely plus probably active as a positive result gave a sensitivity of 97.1% (95% CI 94.6–98.5) and a negative predictive value of 92.7% (95% CI 86.6–96.2) for definite pulmonary TB. Alternative diagnoses among radiographically active cases were predominantly bacterial pneumonia and nontuberculous mycobacterial infection, so imaging activity does not confirm microbiological TB.[11]
An HRCT scoring model developed in the emergency department (8,245 patients) identified cavitation, consolidation, and clusters or nodules in the upper lobes excluding the anterior segment, plus consolidation of the superior segment of a lower lobe, as independent positive predictors of culture-positive TB, with sensitivity 98.5% and specificity 99.7% in the derivation cohort; these figures have not been externally validated and should not be treated as transferable operating characteristics. Post-test probability was strongly prevalence-dependent — 94.5% in high-prevalence, 91.0% in moderate, and 76.8% in moderate-to-low prevalence settings — so CT-based prediction must be interpreted against local epidemiology.[12]
Serial CT during and after treatment
For drug-susceptible disease, treatment response is monitored bacteriologically and serial CT is not indicated. For drug-resistant TB, the 2025 ATS/CDC/ERS/IDSA guideline directs baseline imaging with monitoring approximately every 3 months until end of treatment, together with clinical assessment, sputum collection, and chest imaging at 3, 6, 12, 18, and 24 months after treatment completion (joint panel opinion).[13]
The expected CT trajectory provides a benchmark. In a prospective series of 41 patients, centrilobular lesions were the most common feature of early active disease (95%) and most resolved within 5 months of starting therapy; CT clearly differentiated old fibrotic from new active lesions in 11 of 12 patients with reactivation.[14]
Residual CT abnormality at treatment completion should not be assumed to represent inert scar. Among 75 patients declared cured on negative sputum culture who underwent FDG PET/CT within 2 weeks of completing therapy, 54.7% had residual metabolic activity, and four of five CT features of active disease — lung nodule, consolidation, tree-in-bud micronodules, and FDG-avid chest nodes — were significantly associated with it (P<0.05 for each); pleural effusion was not. Residual metabolic activity in this setting has been associated with relapse.[15]
Worsening CT during treatment: paradoxical reactions and TB-IRIS
Radiologic deterioration during therapy is not synonymous with treatment failure. Paradoxical TB-associated immune reconstitution inflammatory syndrome (TB-IRIS) occurs in people with HIV who start antiretroviral therapy while on TB treatment, with a pooled incidence of 18% across 40 studies and onset typically 1 to 4 weeks after ART initiation. Radiologic manifestations include new or enlarging lymphadenopathy, new or worsening pulmonary infiltrates, new or enlarging pleural and pericardial effusions, enlarging cerebral tuberculomas, and intra-abdominal or retroperitoneal abscesses; nodal enlargement may cause airway obstruction and pericardial enlargement may cause tamponade.[16][17] Risk factors are CD4 count <100 cells/mm³ at ART start, high pre-ART viral load, disseminated or extrapulmonary TB, and a short interval between starting TB treatment and ART.[16]
No imaging finding confirms TB-IRIS. Attribution requires exclusion of alternative causes of deterioration — particularly undetected drug resistance, treatment failure, and another opportunistic infection or malignancy such as non-Hodgkin lymphoma.[16][17] Transient paradoxical worsening also occurs in HIV-negative patients on antituberculous therapy.[17] Radiologic worsening during treatment should therefore prompt repeat microbiology and drug-susceptibility testing before the imaging change is ascribed to an inflammatory reaction.
CT in subclinical tuberculosis
Radiography substantially under-detects disease in subclinical TB. Among 296 patients with subclinical pulmonary TB in a 16-year Canadian cohort, cavitation was 4.77 times (95% CI 1.95–11.66), endobronchial spread 19.36 times (95% CI 8.05–46.52), and moderate or far-advanced parenchymal disease 3.23 times (95% CI 1.66–6.30) more commonly identified on CT than on chest radiograph.[18] The public health implication is that individuals classified as minimally abnormal on radiography may harbor cavitary or bronchogenically spreading disease. This does not currently justify CT-based screening, which has not been evaluated for that purpose. CT has also been evaluated for detecting latent infection. A systematic review of 16 studies found CT consistently more sensitive than chest radiography for identifying latent TB, with low-dose CT showing promise in four studies limited by small sample size; the authors concluded that a randomised trial is needed before low-dose CT can be recommended for this indication.[3]
CT in immunocompromised hosts
In people with HIV and CD4 <200 cells/mm³, infiltrates lose upper-lobe predilection, cavitation becomes uncommon, and lower-zone noncavitating nodular opacities or consolidation with hilar and mediastinal adenopathy predominate; 8–29% of patients with culture-positive pulmonary TB have a normal chest radiograph, and thoracic CT may demonstrate reticulonodular infiltrates in this setting.[16] Extrapulmonary and disseminated disease become progressively more common with worsening immunosuppression, most often involving lymph nodes, liver, spleen, serosal surfaces, and the CNS.
In hematology and transplant patients, HRCT is the technique of choice to detect, characterize, and quantify pulmonary involvement; findings include centrilobular tree-in-bud opacities, ill-defined ground-glass nodules, consolidation, and cavitation, with lower-lobe localization more typical than in immunocompetent adults.[2]
CT in children
Mediastinal and hilar lymphadenopathy is the most characteristic CT finding of pediatric pulmonary TB. In 41 consecutive children with confirmed TB (1997 series), adenopathy was present in 83%, with low-attenuation centers and rim enhancement in 29 of 34, and calcification in 5. Segmental or lobar consolidation occurred in 49%, bronchogenic-spread nodules in 29%, and miliary nodules in 17%. In 20% of children the diagnosis was suggested only by CT, and CT altered clinical management in 37%.[19] Chest radiography has limited accuracy for mediastinal lymphadenopathy compared with CT.[20] Chest radiography nonetheless remains the accepted initial modality; cross-sectional imaging should be considered in a symptomatic child when available, weighed against radiation exposure and the need for sedation in young children.[21][22]
Pleural tuberculosis
CT features that distinguish tuberculous from non-tuberculous pleural infection are interlobular septal thickening (P=0.022) and micronodules in a perilymphatic distribution — subpleural, peribronchovascular, and septal (P<0.001). Subpleural abscess with loss of overlying pleural integrity and peripheral bronchopleural fistula favors bacterial empyema. Consolidation, ground-glass opacity, cavitation, centrilobular nodules, and lymphadenopathy do not discriminate between the two.[23] Consistent with this, subpleural nodules (69% vs 14%; P<0.001) and interlobular septal thickening (81% vs 64%; P=0.009) are significantly more common in pulmonary TB with pleural effusion than without, while tree-in-bud is less common (29% vs 48%; P=0.007).[24]
CT also guides the diagnostic workup. Among 80 patients with TB pleuritis, 80.6% had pleural thickening or nodularity and 82.1% had CT features of active pulmonary TB; sputum culture yielded MTB complex in 33.3% of those with active parenchymal CT features versus 0% of those without, and pleural biopsy yield was markedly higher when pleural thickening or nodularity was present (84.0% vs 25.0%; P=0.010). Pleural fluid culture yield did not differ.[25]
Pericardial tuberculosis
CT may show pericardial thickening (>3 mm), pericardial effusion, and mediastinal lymphadenopathy, frequently with necrotic nodes.[26] Echocardiography, not CT, is the initial test for suspected tuberculous pericarditis; CT is complementary for characterizing pericardial thickening, calcification, and constrictive physiology, and for identifying an extracardiac biopsy target.
Extrathoracic CT findings
The lymphatic system is the most frequently affected extrapulmonary site. Necrotic lymph nodes with low-attenuation centers and rim enhancement are the unifying CT signature across sites and substantially increase diagnostic probability.[27]
| Site | Characteristic CT findings |
|---|---|
| Abdominal / nodal | Lymphadenopathy in 57.1% of abdominal TB, necrotic in 81.7%, at mesenteric, peripancreatic, periportal, and upper para-aortic sites; complex (high-attenuation, septated) ascites; adenopathy adjacent to bowel involvement |
| Gastrointestinal | Ileocecal wall thickening with adjacent adenopathy; strictures; less commonly mass-like lesions |
| Solid organ | Splenic microabscesses; hepatic and splenic hypodense lesions; nonspecific pancreatic masses that can mimic malignancy |
| Peritoneal | Smooth peritoneal thickening with enhancement, omental caking, loculated ascites |
| Genitourinary | Renal parenchymal scarring, calyceal distortion, ureteric strictures, autonephrectomy with calcification |
| Musculoskeletal | Contiguous vertebral body destruction with relative disc preservation, paravertebral and psoas abscess with rim enhancement; monoarticular synovitis |
| CNS | Basal hyperdense exudate on non-contrast images; basal meningeal enhancement, hydrocephalus, infarcts, and tuberculomas on contrast-enhanced images |
Compiled from extrapulmonary imaging reviews and CT series.[28][29] Quantitatively, in 105 CT-evaluated cases of abdominal TB, peritoneal involvement was most common (77.1%), comprising ascites (49.4%), peritoneal (28.4%), and omental involvement (27.2%); the ileocecal region was the commonest bowel site (cumulative 62.9%), typically as circumferential wall thickening without stratification and mild luminal narrowing; hepatic (13.3%) and splenic (16.2%) disease presented predominantly as multiple microabscesses. 38.1% of patients had concomitant pulmonary and extrapulmonary TB, so abdominal TB on CT should prompt chest imaging and respiratory sampling.[30] Abdominal lymph nodes are best evaluated on CT; MRI offers no established added advantage for hepatobiliary disease.[31]
For CNS disease, contrast-enhanced CT can detect hydrocephalus, basal exudates, large infarcts, and tuberculomas and is often the accessible modality, but MRI is more sensitive for small and evolving infarcts, particularly in the brainstem, and is the preferred modality where available.[32][33] Importantly, approximately 30% of children with early tuberculous meningitis have a normal brain CT, and basal meningeal enhancement with or without tuberculoma has been reported as 89% sensitive and 100% specific for TB meningitis in a pediatric comparison against pyogenic meningitis.[34] Consistent neuroimaging features — hydrocephalus, basal exudates, infarcts, or tuberculomas — increase the probability of tuberculous meningitis. Obtaining neuroimaging before lumbar puncture can delay treatment initiation, and empirical therapy should be strongly considered in patients who are critically unwell, alongside treatment for alternative causes.[32]
In a series of 559 patients with CNS TB, initial CT showed hydrocephalus in 40%, infarction in 25%, basal meningeal enhancement in 14%, and tuberculoma in 12% — each less frequent than on MRI, where tuberculoma (40%) and basal enhancement (41%) were considerably more often detected.[35]
Differential diagnosis on CT
Tree-in-bud opacity is not specific to TB and occurs in nontuberculous mycobacterial (NTM) disease, aspiration, bronchiolitis, cystic fibrosis, and diffuse panbronchiolitis. NTM lung disease is the most important and most frequent mimic, and laboratory confirmation is required to distinguish them. Several CT features nonetheless shift probability. In 128 HIV-negative patients with NTM cavities matched to 128 with TB cavities, NTM cavity walls were thinner (6.9 ± 4 mm vs 10.9 ± 6 mm; P<0.001) and more even (thickness ratio 2.6 ± 1 vs 3.7 ± 2; P<0.001). On multivariate analysis, thickening of the pleura adjacent to the cavity (OR 6.49; P<0.001), ill-defined satellite tree-in-bud nodules (OR 8.82; P<0.001), and right-upper-lobe bronchiectasis (OR 5.3; P=0.002) favored NTM, whereas a greater number of non-cavitary nodules ≥10 mm favored TB (OR 0.72; P=0.003). Interobserver agreement for these cavity characteristics was excellent (κ 0.853–0.938).[36] In a series of 200 patients per group, NTM disease more often showed a nodule–bronchiectasis pattern, thin-walled cavities, and lingular and right-middle-lobe predominance, whereas active TB and MDR-TB more often showed thick-walled cavities, consolidation, atelectasis, calcification, pleural thickening, and pleural effusion.[37] Other CT mimics include chronic pulmonary aspergillosis, lung cancer (cavitating), endemic fungal infection, sarcoidosis, granulomatosis with polyangiitis, and septic emboli. Necrotic adenopathy has its own differential including lymphoma and metastatic squamous carcinoma.[1][7]
Clinically actionable recommendations
- Do not delay airborne isolation, sputum collection, or molecular testing to obtain a CT. CT increases specificity; it does not confirm or exclude TB.[1]
- Obtain chest CT when the radiograph is normal or equivocal and clinical suspicion persists, particularly in advanced HIV, anti-TNF therapy, transplant, or hematologic malignancy.[1][2]
- Report tree-in-bud and centrilobular nodularity explicitly; they are the most useful CT markers of active endobronchial spread and correlate with smear positivity.[6][9]
- Document cavitation and its extent — it correlates with bacillary burden and smear grade.[9]
- Use contrast when nodal or extrapulmonary disease is in question; low-attenuation centers with rim enhancement are the key diagnostic sign and are missed on non-contrast studies.[27][30]
- In suspected TB pleuritis, use CT to select the sampling strategy: pleural thickening or nodularity substantially increases pleural biopsy yield, whereas pleural fluid culture yield is unaffected.[25]
- Do not use CT to declare cure. Residual cavities, fibrosis, and bronchiectasis persist after microbiologic cure; treatment response is monitored bacteriologically.[15]
- Before attributing radiologic worsening on treatment to a paradoxical reaction or TB-IRIS, repeat microbiology and drug-susceptibility testing and exclude an alternative opportunistic process.[16][17]
- Prefer MRI over CT for suspected CNS and spinal TB where available, but do not delay empirical therapy for imaging in a critically ill patient.[32][31]
- In children, weigh the diagnostic gain — CT altered management in 37% in one 1997 series — against radiation and sedation risk; radiography remains the initial study.[19][22]
- In drug-resistant TB, obtain baseline imaging and monitor approximately every 3 months until end of treatment, with imaging at 3, 6, 12, 18, and 24 months post-completion.[13]
- Use low-dose or ultralow-dose protocols when serial CT is anticipated and in children and young adults; ultralow-dose CT with iterative model reconstruction preserves depiction of tree-in-bud, cavitation, and consolidation at approximately 0.28 mSv.[5]
References
- ↑ 1.0 1.1 1.2 1.3 1.4 1.5 Ravenel JG, Chung JH, Ackman JB; et al. (2017). "ACR Appropriateness Criteria: Imaging of Possible Tuberculosis". American College of Radiology.
- ↑ 2.0 2.1 2.2 Bergeron A, Mikulska M, De Greef J; et al. (2022). "Mycobacterial Infections in Adults With Haematological Malignancies and Haematopoietic Stem Cell Transplants: Guidelines From the 8th European Conference on Infections in Leukaemia". Lancet Infect Dis.
- ↑ 3.0 3.1 Moore N, Maher M, Murphy G; et al. (2023). "CT in the Detection of Latent Tuberculosis: A Systematic Review". Clin Radiol.
- ↑ Yuan C, Yang J, Huang Y; et al. (2026). "Comparative Assessment of Image Quality and Radiation Dose: Spectral Purification Computed Tomography (Sn100kV) With Adaptive Iterative Reconstruction for Tuberculosis Follow-Up". J Thorac Dis.
- ↑ 5.0 5.1 Yan C, Liang C, Xu J; et al. (2019). "Ultralow-Dose CT With Knowledge-Based Iterative Model Reconstruction (IMR) in Evaluation of Pulmonary Tuberculosis: Comparison of Radiation Dose and Image Quality". Eur Radiol.
- ↑ 6.0 6.1 Hatipoğlu ON, Osma E, Manisali M; et al. (1996). "High Resolution Computed Tomographic Findings in Pulmonary Tuberculosis". Thorax.
- ↑ 7.0 7.1 Nachiappan AC, Rahbar K, Shi X; et al. (2017). "Pulmonary Tuberculosis: Role of Radiology in Diagnosis and Management". Radiographics.
- ↑ Meintjes G, Maartens G (2024). "HIV-Associated Tuberculosis". N Engl J Med.
- ↑ 9.0 9.1 9.2 Ko JM, Park HJ, Kim CH, Song SW (2015). "The Relation Between CT Findings and Sputum Microbiology Studies in Active Pulmonary Tuberculosis". Eur J Radiol.
- ↑ Kim JH, Kim MJ, Ham SY (2019). "Clinical Characteristics and Chest Computed Tomography Findings of Smear-Positive and Smear-Negative Pulmonary Tuberculosis in Hospitalized Adult Patients". Medicine.
- ↑ Ko Y, Lee HY, Park YB; et al. (2018). "Correlation of Microbiological Yield With Radiographic Activity on Chest Computed Tomography in Cases of Suspected Pulmonary Tuberculosis". PLoS One.
- ↑ Yeh JJ, Neoh CA, Chen CR, Chou CY, Wu MT (2014). "A High Resolution Computed Tomography Scoring System to Predict Culture-Positive Pulmonary Tuberculosis in the Emergency Department". PLoS One.
- ↑ 13.0 13.1 Saukkonen JJ, Duarte R, Munsiff SS; et al. (2025). "Updates on the Treatment of Drug-Susceptible and Drug-Resistant Tuberculosis: An Official ATS/CDC/ERS/IDSA Clinical Practice Guideline". Am J Respir Crit Care Med.
- ↑ Im JG, Itoh H, Shim YS; et al. (1993). "Pulmonary Tuberculosis: CT Findings — Early Active Disease and Sequential Change With Antituberculous Therapy". Radiology.
- ↑ 15.0 15.1 Lawal IO, Mokoala KMG, Mathebula M; et al. (2022). "Correlation Between CT Features of Active Tuberculosis and Residual Metabolic Activity on End-of-Treatment FDG PET/CT in Patients Treated for Pulmonary Tuberculosis". Front Med.
- ↑ 16.0 16.1 16.2 16.3 16.4 Benson C, Brooks J, Dhanireddy S; et al. (2026). "Guidelines for the Prevention and Treatment of Opportunistic Infections in Adults and Adolescents With HIV". Office of AIDS Research Advisory Council.
- ↑ 17.0 17.1 17.2 17.3 Nahid P, Dorman SE, Alipanah N; et al. (2016). "Official American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America Clinical Practice Guidelines: Treatment of Drug-Susceptible Tuberculosis". Clin Infect Dis.
- ↑ Lau A, Lin C, Barrie J; et al. (2022). "A Comparison of the Chest Radiographic and Computed Tomographic Features of Subclinical Pulmonary Tuberculosis". Sci Rep.
- ↑ 19.0 19.1 Kim WS, Moon WK, Kim IO; et al. (1997). "Pulmonary Tuberculosis in Children: Evaluation With CT". AJR Am J Roentgenol.
- ↑ Vonasek B, Ness T, Takwoingi Y; et al. (2021). "Screening Tests for Active Pulmonary Tuberculosis in Children". Cochrane Database Syst Rev.
- ↑ Jain SK, Andronikou S, Goussard P; et al. (2020). "Advanced Imaging Tools for Childhood Tuberculosis: Potential Applications and Research Needs". Lancet Infect Dis.
- ↑ 22.0 22.1 Concepcion NDP, Laya BF, Andronikou S; et al. (2023). "Imaging Recommendations and Algorithms for Pediatric Tuberculosis: Part 1 — Thoracic Tuberculosis". Pediatr Radiol.
- ↑ Ko JM, Park HJ, Cho DG, Kim CH (2015). "CT Differentiation of Tuberculous and Non-Tuberculous Pleural Infection, With Emphasis on Pulmonary Changes". Int J Tuberc Lung Dis.
- ↑ Jung MK, Lee SY, Min EJ, Ko JM (2023). "CT Scan Differences of Pulmonary TB According to Presence of Pleural Effusion". Chest.
- ↑ 25.0 25.1 Young SL, Chua BLW, Tan QL; et al. (2025). "Pleural and Parenchymal Radiological Characteristics of Tuberculous Pleuritis and Correlation With Microbiological and Molecular Diagnostic Yield". BMC Pulm Med.
- ↑ Burrill J, Williams CJ, Bain G; et al. (2007). "Tuberculosis: A Radiologic Review". Radiographics.
- ↑ 27.0 27.1 Rodriguez-Takeuchi SY, Renjifo ME, Medina FJ (2019). "Extrapulmonary Tuberculosis: Pathophysiology and Imaging Findings". Radiographics.
- ↑ Hulnick DH, Megibow AJ, Naidich DP; et al. (1985). "Abdominal Tuberculosis: CT Evaluation". Radiology.
- ↑ Engin G, Acunaş B, Acunaş G, Tunaci M (2000). "Imaging of Extrapulmonary Tuberculosis". Radiographics.
- ↑ 30.0 30.1 Deshpande SS, Joshi AR, Deshpande SS, Phajlani SA (2019). "Computed Tomographic Features of Abdominal Tuberculosis: Unmask the Impersonator!". Abdom Radiol (NY).
- ↑ 31.0 31.1 Skoura E, Zumla A, Bomanji J (2015). "Imaging in Tuberculosis". Int J Infect Dis.
- ↑ 32.0 32.1 32.2 Donovan J, Cresswell FV, Tucker EW; et al. (2026). "A Clinical Practice Guideline for Tuberculous Meningitis". Lancet Infect Dis.
- ↑ Thwaites GE, van Toorn R, Schoeman J (2013). "Tuberculous Meningitis: More Questions, Still Too Few Answers". Lancet Neurol.
- ↑ Thwaites GE, Tran TH (2005). "Tuberculous Meningitis: Many Questions, Too Few Answers". Lancet Neurol.
- ↑ Azeemuddin M, Alvi A, Sayani R; et al. (2019). "Neuroimaging Findings in Tuberculosis: A Single-Center Experience in 559 Cases". J Neuroimaging.
- ↑ Kim C, Park SH, Oh SY; et al. (2017). "Comparison of Chest CT Findings in Nontuberculous Mycobacterial Diseases vs. Mycobacterium Tuberculosis Lung Disease in HIV-Negative Patients With Cavities". PLoS One.
- ↑ Xu L, Xu S (2021). "CT Imaging Characteristics of Nontuberculous Mycobacteria Lung Disease, Active Tuberculosis and Multi-Drug Resistant Tuberculosis". Sarcoidosis Vasc Diffuse Lung Dis.