The management of thyroid nodules was relatively simple before 1980: only palpable nodules were evaluated; scintigraphy allowed differentiation between hot nodules which are virtually always benign and cold nodules which are benign in 80% of cases even though most cancers are cold. As such, nearly all palpable cold nodules were resected, especially when isolated. Cancer was diagnosed at histology in about 10% of cases .
Cytology allowed a better selection of patients requiring surgery. Colloid lesions (benign) could be diagnosed. Patients with suspicious or malignant results at cytology would undergo surgery in order to obtain an histological diagnosis.
In the mean time, the wider use of US has created some problems due to the high incidence of non-palpable nodules that are often present in patients with clinically isolated palpable thyroid nodule (40% of cases) . Should all these nodules be resected (some of which being actual cancers)? Should all these thyroids be classified as multinodular goiters for which surgery is seldom indicated? Can US be used to identify thyroid nodules that should be biopsied or resected? What role should be given to scintigraphy?
The ANDEM has issued, in 1997, guidelines regarding the management of thyroid nodules .
This paper will review the role of imaging studies in the work-up of thyroid nodules based on these guidelines. Biopsy of thyroid nodules, which is the best technique to differentiate benign from malignant thyroid nodules, will only be discussed because it may be performed under US guidance. The relevance of imaging studies for daily clinical practice is based on an understanding of the dilemmas arising from the diagnostic work-up of thyroid nodules.
The high incidence of thyroid nodules has been described in autopsy studies performed in the 50's: about 50% of individuals (without clinically palpable thyroid nodule) have thyroid nodules .
This was confirmed by several US studies: 41% of patients have incidental thyroid nodules . Fifty percent of adults over 50 years have at least 1 thyroid nodule . Thirty-four percent of healthy adults referred for unrelated reasons have thyroid nodules .
The incidence of palpable nodules is about 5% (somewhat similar to the 12% incidence of thyroid nodules > 1 cm on autopsy series).
The incidence increases with age, and thyroid nodules are more common in women . This natural incidence may be further increased by environmental factors: ionizing radiation, iodine deficiency; the role of tobacco is debated [9, 10].
In France, the incidence of thyroid cancer is estimated at about 1.9/100000 for men, and 3.1/100000 for women . If tiny (occult) papillary carcinomas are included, the incidence increases to about 36% of individuals based on autopsy series [12, 13]. The natural history of these occult carcinomas is unknown. Some never evolve to become clinical carcinomas, whereas some may evolve to become invasive carcinomas. It is not currently possible to identify which ones will become invasive carcinomas. The frequent occurrence of such lesions on autopsy series or on surgical specimen significantly complicates the management of patients with thyroid nodules [14, 15]. One must remember that the mortality due to thyroid carcinoma is very low.
The percentage of cancers is higher in men. It also is more common at both extremes of life.
Thyroid nodules are generally detected at palpation, sometimes by patients themselves. Nodules are increasingly being detected incidentally at CT and US performed for other reasons. Rarely, nodules will be discovered in patients being screened because of a family history of thyroid carcinoma or MEN (multiple endocrine neoplasia) syndrome, or in patients with metastatic disease from unknown primary origin.
Irrespective of the clinical context, all patients should undergo clinical and biological (T4 and TSH) work-up prior to imaging.
High-resolution US using a 7.5 MHz or higher frequency transducer is required to image the thyroid gland and lymph node bearing areas. A wide or curvilinear transducer (may be of lower frequency) is useful to measure the height of the thyroid lobes. A smaller or sector transducer may also be needed to better evaluate patients with intra-thoracic extension of thyroid tissue. A stand-off pad or other coupling device may be needed to evaluate the isthmus and superficial portion of the thyroid gland .
US units equipped with pulsed Doppler (PDI), color Doppler (CDI), and power Doppler are quite helpful.
PDI allows quantitative evaluation of flow: resistive index (RI), systolic velocities.
CDI and power Doppler allow real-time angiographic evaluation. Power Doppler is more sensitive than CDI. The lack of flow direction encoding is not a pitfall for thyroid gland imaging.
After the neck is palpated, the US evaluation is performed with patients in the supine position with neck hyperextension (a pillow can be placed under the shoulders). US evaluation of the thyroid should always be meticulous and include an evaluation of the thyroid tissue, lymph node bearing areas, and supra- and infra-thyroid regions.
1. Evaluation of thyroid gland volume. Each thyroid lobe is measured:
V = (H x L x W) x 0.5
The volume of the isthmus is negligible except when markedly enlarged. Normal thyroid volume varies between 10-20 g.
2. Evaluation of thyroid echogenicity:
Thyroid hypoechogenicity suggests a diagnosis of thyroiditis or Graves' disease. It should always be described in the report. Thyroid echogenicity is compared to that of the strap muscles: normal thyroid tissue always is more echogenic than muscle.
3. Presence of nodules
- diagnosis: all focal areas of echotexture abnormality are not nodules: a nodule corresponds to a rounded or oval-shaped focus developing within an organ. Therefore a thyroid nodule corresponds to a rounded or oval-shaped structure that can be identified at US on two orthogonal planes due to its different echotexture or due to displacement of surrounding vessels, as seen on CDI. In some cases, US will show that no nodule is present where a palpable nodule is suspected clinically;
- differential diagnosis: it includes hypoechoic pseudonodules seen in patients with thyroiditis Fig. 1 (these pseudonodules should not be described as nodules in the report as it would lead to unnecessary work-up) and false nodules related to septations. The only pitfall is the intra-thyroid parathyroid nodule: this diagnosis is suggested in patients with hyperparathyroidism and only confirmed at biopsy (PTH level in the sample) or at parathyroid scintigraphy (subtraction or Tc99m-sestamibi).
Each nodule must be carefully described
- size: 3 diameters are required to correctly describe the size of a nodule (and eventually its volume). Micronodules (< 5 mm) are frequently present and do not necessarily have any clinical implication;
- echotexture and echogenicity relative to normal thyroid tissue. Three types are described: cysts, solid nodules, and mixed nodules Figs. 2 , 3 , and 4 . Table I is a summary of all different types of nodules with their relative risk of malignancy. Most cancers are in the group of hypoechoic solid nodules (about 50%) ; the risk of malignancy is very low for purely cystic nodules and hyperechoic solid nodules ;
- contours: ill-defined or irregular margins are suspicious. A complete and regular perinodular halo is reassuring though its presence does not completely exclude the possibility of malignancy Fig. 5 ;
- presence of calcifications: especially diffuse tiny calcifications seen at high frequency (13 MHz) . Large chunk calcifications have been described in association with anaplastic carcinomas ;
- evaluation of nodule vascularity: it is mainly qualitative (CDI and power Doppler). Nodule vascularity is compared with the vascularity of normal thyroid parenchyma. The parameters must be standardized and should not be modified during the examination (PRF about 1 kHz for CDI, and about 4 kHz for power Doppler). The color gain should be set at the highest sensitivity while avoiding artefacts. Most authors have described 4 types of vascular patterns
- Type I: no vascularity;
- Type II: perinodular vascularity;
- Type III: intranodular vascularity;
- Type IV: diffuse hypervascularity Figs. 6 and 7 .
Contradictory results have been published regarding the significance of vascularity for characterization. Color mapping appears to be more useful than spectral analysis. Intranodular vascularity would be more suspicious than perinodular vascularity [22, 23, 24, 25].
It is useful to provide clinicians with a diagram of nodule location. This diagram is useful for follow-up and at the time of biopsy Fig. 8 .
4. Evaluation of cervical lymphadenopathy.
Normal nodes are visible at US when using a high frequency transducer . Abnormal nodes are larger, have a more rounded appearance and loose the normal architecture, especially the fatty hilum Fig. 8 . The ratio of Solbiati (height/width) is useful to identify abnormal nodes: nodes are usually normal when > 2, whereas they are usually abnormal when < 1.5.
5. Evaluation of the trachea.
Tracheal deviation by a large nodule can easily be detected. Tracheal compression is more difficult to detect at US; it commonly is an indication for surgery. In some cases, frontal and lateral radiographs of the trachea must be obtained.
6. Evaluation of thyroid capsule.
It is always visible in normal patients when using high frequency US imaging. It may no longer be visible in patients with thyroiditis or invasive cancers.
It must be clear and systematic, especially when describing nodules. The US examination and terminology used in the report must be standardized in order to reduce the operator dependent nature of this imaging technique.
The radiology report must help the clinician in assessing the likelihood of malignancy and should answer the following questions
- Is the thyroid of normal size?
- Are nodules present? How many are present? What is their size? Are some suspicious at US?
- Are abnormal cervical lymph nodes present? Is there evidence of local invasion?
The likelihood of malignancy is highest for isolated nodules that are hypoechoic, have ill-defined margins, with presence of abnormal lymph nodes or evidence of mass effect or local invasion.
The US appearance is more suggestive of benign lesions when the nodules are multiple (> 3), cystic or solid hyperechoic, and without cervical lymph nodes or evidence of local invasion. However, US alone cannot exclude that some nodules may be malignant. The clinician must correlate these US findings with other findings: clinical history, physical examination, lab results, scintigraphy, and cytology.
The solid hypoechoic nature of a nodule is the most suspicious US appearance for malignancy [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26] (superior to the "cold nodule" at scintigraphy ). However, it may be difficult to characterize small nodules as being hypoechoic. Small nodules that are partly cystic may be incorrectly classified as solid hypoechoic, hence decreasing the specificity of this US finding. Therefore, it may be reasonable to exclude small nodules (< 6-8 mm in diameter, or < 100 mm3) from this category, especially since it seems generally accepted that such small nodules should not be further evaluated . In other words, one must be confident that a nodule is solid hypoechoic to benefit from the predictive value of this US appearance, which may be difficult to achieve for smaller nodules. Table II is a summary of the diagnostic value of US based on 3 published series as reviewed by Leenhardt.
On the other hand, the echogenicity of very large nodules is less useful since it is well accepted that these nodules should be surgically removed, especially when the loco-regional work-up is abnormal.
Currently, cytology is the technique used by several groups to determine if a nodule should be surgically removed. It is sometimes performed before US evaluation. Relatively safe and inexpensive, cytology can be used to separate benign colloid nodules from suspicious or malignant nodules that should be removed [30, 31, 32]. Cytology should be performed by experienced groups to insure the validity of its results. Palpable nodules can be biopsied directly by the clinician or cytologist. US can be used to biopsy non-palpable or heterogeneous nodules when a specific area of the nodule must be biopsied (solid component) [33, 34]. The value of cytology is assessed by its rate of non-contributive samples. A rate between 10-15% seems satisfactory. This rate is difficult to achieve if one performs only a few biopsies per month. Collaboration between the individual performing the biopsy and the cytologist is needed to achieve good results.
US guided biopsy is performed under standard sterile technique. The equipment must be sterilized between patients. Single-use sterile plastic covers are available for the transducer. Sterile water should be used as a coupling agent since sonographic gel, even sterile, may interfere with the cytological evaluation. In patients with several nodules, a complete US evaluation must first be performed in order to select the nodules that need to be biopsied. Using a diagram, the suspicious nodules should be easy to identify. Two to three samples are obtained in different areas for homogeneous nodules. For heterogeneous nodules, the samples should be obtained from the solid components and the walls. US can also be used to empty a fluid filled nodule and biopsy its walls.
Samples are placed of glass slides and may be air-dried or fixed. The samples are stained prior to being read. Several techniques of biopsy have been described, with or without aspiration. It takes about 30 seconds to obtain each sample, a time during which the patient should refrain from swallowing. The biopsy is not painful and local anesthesia is not needed. However, it is difficult to biopsy more than 3 nodules per session. After the biopsy, a compressive bandage is applied for 6 hours.
Possible complications include intrathyroid (intracapsular) hematomas. They spontaneously resolve over a few weeks. Patients with coagulation disorders or on anti-coagulants should not undergo thyroid biopsy. Also, patients with severe respiratory problems may not be able to tolerate the supine position with neck hyperextension.
If patients are not operated, US can be used for follow-up of nodules whether or not they are placed on hormonal replacement. Each nodule will be carefully assessed (value of diagram). Interval variations in size, echogenicity, or vascular pattern are noted. In patients with multinodular goiter, enlarging or suspicious nodules are detected and US can be used to biopsy these nodules. The interval appearance of intranodular vascularity must also be noted.
Scintigraphy detects uptake of a radiotracer by a nodule, and hormonogenesis when iodine is used. Scintigraphy provides functional and morphological data. This technique detects nodules 5 mm or larger. It is reproducible and only minimally operator dependent.
Scintigraphy using I123 or Tc99m pertechnetate was for many years the main imaging technique used for thyroid imaging. The uptake of nodules at scintigraphy lead to the classification of nodules as "hot", "cold", or isointense
- Hot nodules correspond to about 15-30% of thyroid nodules. They are virtually always benign. They may evolve to become adenomas or undergo necrosis Fig. 9 .
- Cold nodules are more common: 70-80% of nodules. About 15-20% are malignant Fig. 10 .
- Isointense nodules are not visible at scintigraphy: 10-15% of nodules.
Other radionuclides have been proposed to improve characterization of nodules as benign or malignant. Tl201 provided interesting results [35, 36]. Tc99m sestamibi and Tc99m tetrofosmin  also provided interesting results. These tracers are not routinely used.
Diagnostic work-up and role of scintigraphy|
In some cases, the management of patients is straightforward. Surgery is performed early when a nodule is very suspicious and scintigraphy may not be necessary. However, scintigraphy should be performed when a nodule is associated with a low TSH level to confirm the presence of a toxic nodule. On the other hand, if the TSH level is elevated, in the setting of auto-immune thyroiditis, patients must be evaluated for the presence of hot or cold pseudonodules that should be evaluated as such.
In most cases, the type of diagnostic work-up can be customized. The ANDEM , in their study, have proposed three strategies, based on results at three examinations: biopsy, scintigraphy, and US.
1. Cytology first, followed by scintigraphy when results at cytology are intermediate or doubtful, or when they are non-contributive, even after being repeated. If cytology shows malignancy, surgery is performed. If cytology confirms the benign nature of a nodule, only follow-up is performed.
2. US first, followed by scintigraphy and/or cytology when the nodule is hypo- or isoechoic. Hyperechoic nodules are simply followed-up.
3. Scintigraphy first, followed by US and/or cytology if the nodule is cold or isointense. TSH levels are followed-up in patients with a hot nodule.
It must be noted for groups 1 and 2 that a diagnosis has not been achieved for nodules that are being followed-up. The follow-up must be based on what: TSH level, volume of the nodule, or cytology? If only about 10% of nodules are hot, then, about 90% of patients with cold or isointense nodules will undergo serial TSH level measurements without clear indication. If US is performed first, then, hyperechoic cancers may be overlooked.
A French study on cost-effectiveness in the evaluation of thyroid nodules  has shown that the immediate cost was higher but that long-term costs were lower when a precise diagnosis was achieved at initial work-up. As such, it is common to perform 2 or 3 of these evaluation techniques at initial work-up, based on local expertise.
Interpretation of scintigraphy|
1. Nodule with low TSH
- The nodule may be a toxic adenoma, hot at scintigraphy, with preferential T3 secretion.
- Other possibilities include: Graves disease with a cold nodule, painless thyroiditis, low TSH for unrelated reasons...
2. Nodule with normal TSH:
Scintigraphy may be reassuring in 15-30% of cases by showing a hot nodule which is virtually always benign. Biopsy may be avoided. Follow-up may be performed using serial TSH level measurements. If the nodule is cold, biopsy is performed.
3. Nodule with high TSH:
It most commonly corresponds to chronic auto-immune thyroiditis. If the nodule is hot, it most commonly is benign. If it is cold, it must be biopsied.
In daily practice, the main contra-indication to scintigraphy is pregnancy. Scintigraphy can be performed in patients receiving hormonal replacement therapy as long as the TSH is not too low. It can also be performed in patients with iodine overload (increased dose of radiotracer, longer acquisition). In patients with thyroid dysfunction due to iodine overload, scintigraphy remains the best imaging technique.
Scintigraphy remains an important part of the diagnostic work-up of thyroid nodules, especially when cytology is not readily available.
CT and MRI play no role in the first line evaluation of thyroid nodules. They may be helpful for preoperative evaluation, especially in patients with large goiter extending into the mediastinum. The administration of iodinated contrast material may interfere with scintigraphy, especially following complete thyroidectomy (evaluation for metastases). As such, MRI is preferred over CT. If MRI is not available, then CT should be scheduled after scintigraphy [41, 42].
Helical acquisitions are preferable. Contiguous postcontrast images should be reconstructed every 2 to 5 mm Fig. 11 .
Enhancement is not helpful for characterization. Cystic and calcified areas are readily identified. Tracheal compression and extracapsular extension can be diagnosed. The main value of CT is to evaluate the relationship of thyroid lesions with surrounding structures, including vessels in the mediastinum [41, 42].
Even though some groups have reported that MRI was somewhat specific for benign nodules , the limited availability of MR units and its higher cost have lead some groups to use MRI mainly for the preoperative evaluation of some patients, especially when it may be preferable to avoid the use of iodinated contrast material.
MRI provides an excellent evaluation of the relationship of a goiter with surrounding vessels, and can detect extracapsular extension or cervical lymphadenopathy Fig. 12 . MRI is helpful to detect recurrent disease, especially medullary carcinoma which does not show tracer uptake at scintigraphy.
Imaging plays an important role in the management of patients with thyroid nodules. Morphological and functional findings are helpful to identify suspicious nodules that should be biopsied or surgically removed, while preventing unnecessary biopsy of probably benign nodules.
Scintigraphy is helpful to detect hot nodules which are managed differently.
US has become the baseline imaging technique because it is highly sensitive, relatively inexpensive, non-invasive, and widely available. Because US remains operator dependent, the examination technique and reporting of findings must be standardized to provide clinicians with useful information for patient management. US can detect non-palpable nodules and US guidance can be used for biopsy of non-palpable or heterogeneous nodules.
(Les tableaux sont exclusivement disponibles en format PDF).
(Les tableaux sont exclusivement disponibles en format PDF).
Thyroïdite avec pseudo-nodules. Thyroiditis with pseudonodules
Nodule hyperéchogène. Hyperechoic nodule.
Nodule isoéchogène. Isoechoic nodule.
Nodule hypoéchogène à contours réguliers. Hypoechoic nodule with well-defined margins.
Nodule hypoéchogène suspect, à contours flous. Suspicious hypoechoic nodule with ill-defined margins.
Vascularisation périnodulaire régulière en écho-Doppler couleur. Well-defined perinodular vascularization at CDI.
PDI of a perinodular artery.
(Les tableaux sont exclusivement disponibles en format PDF).
Adénopathie jugulo-carotidienne. Carotid sheath lymph node.
Scintigraphie à l'Iode : nodule chaud. Iodine scintigraphy: hot nodule.
Scintigraphie à l'Iode : nodule froid. Iodine scintigraphy: cold nodule.
Thyroïde normale en TDM. Normal thyroid at CT.
Nodule préthyroïdien gauche en IRM en T2. T2W MR image. Lesion anterior to the left thyroid lobe.
 Gharib H. Fine-needle aspiration biopsy of thyroid nodules: advantages, limitations and effects. Mayo Clin Proc 1994;69:44-9.
 Scheible W, Leoplold GR, Woo VL, Gosink BB. High resolution real-time ultrasonography of thyroid nodules. Radiology 1979;133:413-7.
 La prise en charge diagnostique du nodule thyroïdien. Recommandations pour la pratique clinique. ANDEM. Service des études. Décembre 1995. Éditions Norbert Attali.
 Mortensen JD, Woolner LB, Bennet WA. Gross and microscopic findings in clinically normal thyroid glands. J Clin Endocrinol Metab 1955;15:1270-80.
 Horlocker TT, Hay ID, James EM, Reading CC. Prevalence of incidental nodular thyroid disease detected during high resolution, parathyroid ultrasonography. In: Frontiers of thyroidology. Plenum Press, Medeiros, Neto G, édit., New York, 1986; vol 2:1309-12.
 Funari M, Campos Z, Gooding GAW, Higgins CB. MRI and ultrasound detection of asymptomatic thyroid nodules in hyperparathyroidism. J Comput Assist Tomogr 1992;16:615-9.
 Bruneton JN, Marcy PY, Ballu-Maestro C et al. Les nodules thyroïdiens. Apport de l'imagerie. Feuillets de Radiologie 1993;33:145-55.
 Vander JB, Gaston EA, Dawber TR. The significance of nontoxic thyroid nodules: final report of a 15-year study of the incidence of thyroid malignancy. Ann Intern Med 1968;69:537-40.
 Georgiadis E et al. The influence of smocking habits on thyroid gland volume: an ultrasonic approach. JR Soc Health. 1997;117:355-8.
 Gaitan E. Goitrigens. J Clin Endocrinol Metab 1988;2:683-702.
 Sancho-Garnier H. L'épidémiologie des cancers de la thyroïde. Ann Radiol 1977;20:715-21.
 Sampson RJ, Woolner LB, Bahn RC, Kurland LT. Occult thyroid carcinoma in Olmsted country, Minnesota: prevalence at autopsy compared with that in Hiroshima and Nagasaki. Japan. Cancer 1974;34:2072-6.
 Harach HR, Franssila KO, Wasenius VM. Occult papillary carcinoma of the thyroid. A "normal" finding in Finland: asystematic autopsy study. Cancer 1985;56:531-8.
 Leenhardt L. Prise en charge du nodule thyroïdien isolé: accords et controverses, évaluation échographique. Communication au Congrès de la Société de Chirurgie Endocrinienne. Marseille. Juin 1993.
 Tramalloni J, Leenhardt L. Échographie des nodules thyroïdiens: ce qu'en attend le clinicien. J Radiol 1994;75;3:187-90.
 Tramalloni J, Beylouné-Mainardi C, Séné V et al. Imagerie normale et pathologique de la thyroïde. Encycl Med Chir Radiodiagnostic Cœur Poumon. Éditions Techniques, Paris, 1994;32.700.A.30:10 p.
 Poncin-Viateau J, Hassan M. Échographie thyroïdienne (2e éd). Vigot, édit., Paris 1992.
 Solbiati L, Volterrani L, Rizzatto G et al. The thyroid gland with low up. take lesions: evaluation by ultrasound. Radiology 1985;155:187-91.
 Hayashi N, Tamaki N, Yamamoto K et al. Real time ultrasonography of thyroid nodules. Acta Radiologica Diagnosis 1986;27:403-8.
 Solbiati L, Cioffi V, Ballarati E. Ultrasonography of the neck. Radiol Clin North AM 1992;5:941-51.
 Hatabu H, Kasagi K, Yamamoto K. Undifferentiated carcinoma of the thyroid gland. Clin Radiol 1992;45:307-10.
 Fobbe F. Thyroid Gland. In color duplex sonography. Wolf KJ, Fobbe F. Thieme Medical Publishers, édit., Stuttgart, New York, 1995:217-24.
 Shimamoto K, Endo T, Ishigaki T, Sakuma S, Makino N. Thyroid nodules: evaluation with color duplex ultrasonography. J Ultrasound Med 1993;12:673-8.
 Solbiati L, Livraghi T, Ballarati E et al. Thyroid gland. In: Solbiati L, Rizatto G. Ultrasound of superficials structures. Churchill Livingstone, édit., Edinburgh, 1995:49-86.
 Spezia S Cerbone G, Colao A et al. Usefulness of power Doppler in the diagnostic management of hypoechoic thyroid nodules. Eur J Ultrasound 1997;6:167-70.
 Watters DA, Ahuja AT, Evans RM et al. Role of ultrasound in the management of thyroid nodules. A J Surg 1992;164:654-7.
 Léger A. La Pathologie thyroïdienne. Diagnostic et traitement. Médecine. Science Flammarion, édit., Paris, 1990.
 Rodier M, Dupre D, Adam B, Gouzes C, Bourgeois JM. Intérêt de l'échographie dans l'évaluation pré-opératoire d'un nodule isolé de la thyroïde Litter Ultra Sonore 1992;8:79-82.
 Espinasse P, Liaras A, Labadie M, Vassy V. Intérêt de la cytoponction thyroïdienne à l'aiguille fine, couplée à l'échographie et à la scintigraphie. Rev Fr Endocrinol Clin 1993;34:663-72.
 Grant CS, Hay ID, Gough IR, MacCarthy PM, Goellner JR. Long term follow-up of patients with benign thyroid fine-needle aspiration cytologic diagnoses. Surgery 1989;106:980-6.
 Hamburger JI. Consistancy of sequential needle biopsy findings for thyroid nodules. Arch Intern Med 1987;147:97-9.
 Hamburger JI. Diagnosis of thyroid nodules by fine needle biopsy: use and abuse. J Clin Endocrinol Metab 1994:79;335-9.
 Tramalloni J, Merceron RE, Voillemot N et al. Cytoponction thyroïdienne échoguidée à l'aiguille fine des nodules thyroïdiens non palpables. JEMU 1989;10, 6:270-4.
 Cordray JP, Merceron RE, Tramalloni J et al. Intérêt de la cytoponction échoguidée dans la pathologie thyroïdienne nodulaire non palpable. Rev Franç Endocrinol Clin 1990;31:327-35.
 Hardoff R, Baron E, Sheinfeld M. Early and late lesion-to-non lesion ratio of thallium 201 chloride uptake in the evaluation of "cold" thyroid nodule. J Nucl Med 1991;32:1873-6.
 Hermans J, Schmitz A, Merlo P et al. Le thallium 201 permet-il de différencier le nodule thyroïdien bénin du nodule malin? Ann Endocrinol 1993;54:248-54.
 Klain M, Maureas S, Cuocolo A et al. Technetium-99m tetrofosmin imaging in thyroid diseases: comparitison with Tc-99m-pertechnetate, thallium-201 and Tc-99m-methoxyisobutylisonitrile scans. Eur J Nucl Med 1996;23:1568-74.
 Eeckhoudt L, Hermans J, Merlo P et al. Analyse économique des approches diagnostiques et thérapeutiques du nodule froid de la thyroïde. Ann Endocrinol 1993, 54:293-6.
 Galli G. The characterization of thyroid nodules: rationale, cost and benefits. In: Thyroid diseases. Boca Raton. CRC Press. Troncone L, Shapiro B, Satta MA, Monaco F, édit., 1994:369-87.
 Van Herlé AJ, Rich P, Ljung BME et al. The thyroid nodule. Ann Intern Med 1982;96:221-32.
 Radecki PD et al. Thyroid imaging: comparison of high-resolution real-time ultrasound and computed tomography. Radiology 1984;153:145-7.
 Bruneton JN, Padovani B, Marcy PY, Darcourt J, Maestro C, Raffaeli C. Exploration de la thyroïde par tomodensitométrie et imagerie par résonnance magnétique. In: Imagerie en endocrinologie. Masson, édit., Paris, 1996:70-3.
 Renard E, Jaffiol C, Rouanet JP et al. Nodules thyroïdiens et goitres non fonctionnels. Apport diagnostique de l'imagerie par résonnance magnétique. Presse Med 1991;20:294-8.