Advancements in Diagnosing Pulmonary Embolism

=Introduction=

Pulmonary embolism (PE) is one of the most fatal illnesses around the globe (Lambrini et al 2018). Global statistics classify PE as being third most common, and highly fatal, cardiovascular condition after stroke and coronary artery disease, otherwise known as acute coronary syndrome (Dogan et al 2015).

PE occurs as a result of a blockage in the blood vessels of the lungs (Pulmonary embolism). The blockage is due to a blood clot which is then referred to as the emboli which having travelled from other parts of the body as a consequence of venous thromboembolism deposits itself in the tiny blood vessels in the lungs (Kearon, 2003). The blood clot may also arise initially as a deep vein thrombosis mostly in the limbs before being transported to the lungs (Lambrini et al 2018). However, PE may also arise due to other factors such as an emboli arising from the bone marrow or amniotic emboli which may arise during pregnancy or after birth (Girtovitis, 2014). Although PE is prevalent, the main form of PE is that which arises from a venous thrombosis and not the others.

PE as will be discussed exhibits minimal to no consistent symptoms allowing a patient to seek appropriate medical attention. The repercussions of an emboli attaching on the vascular wall of a blood vessel in the lungs are swift and potentially fatal (Lambrini et al 2018). However, traditional symptoms have been observed and studied. These were especially developed in the nineties since there was acute inability to properly diagnose PE. The problem of having constant symptoms is perhaps caused by the fact that the embolism will mostly manifest in short episodes as blood is easily pumped away if the emboli is tiny enough (Menaker et al, 2007). Where the embolus occupies over 80% of the pulmonary artery’s vascular space sudden death would occur whereas a cover of 60% would result into acute low blood pressure occasioned with left-side failure of the lung (Righini et al, 2017). This means therefore that signs of PE will mostly manifest upon having a devastating effect on the patient. Hence prior diagnosis is very much important and life-saving.

The very nature of PE having unspecified signs and symptoms makes it very difficult to clinically assess (Kearon, 2003). Whereas Kearon’s assessment is generally true and probably accurate at the time of writing his research, the developments in diagnosis mechanisms recently has greatly increased the ability of clinicians to assess and suspect PE in a patient. Take for instance the statistics in North America in the late 21st century where up to fifty percent (50%) of all tests returned with diagnosis of PE unlike presently and as late as 2017 where a meagre 5% have returned with a positive diagnosis of PE (Righidi, Robert-Ebadi and Le Gal, 2017).

Thanks to major development in imaging and radiography, particularly in computed tomography (CT) scan which for purposes of this work will be referred to as the computed tomography pulmonary angiography (CTPA), PE has been greatly managed (Remy-Jardin, Pistolesi and Goodman, 2007).

A. axial CTPA image of acute segmental pulmonary embolism on expanded right lower lobe posterior pulmonary artery.

B. coronal CTPA image of the same case in A above along the long axis of a right lower lobe segmental pulmonary artery.

The condition is best suited for this work as it allows widespread use of CTPA yet there exists an alley of diagnostic mechanisms which traditionally were considered more non-intrusive and safer for use. This would include V/Q scan otherwise known as ventilation perfusion lung scintigraphy (The PIOPED investigators, 1990), the D-Dimer testing among others. Hence this offers exceptional opportunity to interrogate the input of CTPA in PE over other diagnostic mechanisms.

=Review of practices=

==Symptomatic features of PE==

As earlier highlighted, studies have shown that PE has very radical symptoms which in some cases result to catastrophic occurrences such as death. However, as PE may occur with relapses, as a single emboli or widespread blockages; it may show certain non-fatal symptoms which would help the clinician to diagnose (Tsilimidos and Marinis, 2013). Akiros, 2009 observed that where there is wide spread PE in a patient certain painful symptoms may be exhibited such as extreme chest pains and extensive hemodynamic restlessness. These, Akiros observed, are mainly the result of cardiac dysfunction caused by the sudden blockage of blood flow.

The pain experienced by the patient usually leads to the suffering of massive decrease in blood pressure, increased tachypnea and due to the lack of oxygenated blood in the right part of the heart a patient would suffer radical electrocardiographic changes (Kapetaniou, 2011).

There is only consistent observation which a doctor would find especially after a patient has had an operation; this is the presence of tachycardia and is mostly the result of small PE (Tsakali, 2007). Moreover where there is sudden shortness of breath, a clinician should suspect the occurrence of PE.

Thus a mere physical examination of the patient may lead to misdiagnosis of the illness as these symptoms are not consistent from one patient to another (Patakas, 2006). It all depends with the nature of PE that a patient suffers from.

Even though there is minimal exhibition of physical asymptomatic features of PE, most studies have shown the essence of having a clinical assessment done before taking diagnostic measures for the patient as this would inform the choice of the diagnostic test to be carried out (Kearon, 2003); this position remains truthful to date with a slight change in mechanisms. Presently, the clinical assessment will make use of highly sensitive D-dimer test before the application of imaging procedures to conclusively diagnose (Righidi, Robert-Ebadi and Le Gal, 2017).

The ultimate goal of any of the assessments or tests done is to exclude PE rather than include it as the idea is to ensure that intrusive angiography mechanisms have been avoided at all costs (Konstantinides, 2016). This allows the use of clinical probability of PE assessment to recommend the appropriate diagnostic measures to be taken (Weiner, Shwartz and Woloshin, 2011). Moreover patient’s reaction to certain medication or procedures as well as their characteristics may only be assessed clinically before the application of a certain diagnostic test (Wells et al, 2001). Indeed it has been evidenced that for patients suffering from kidney failure or renal failure, the use of CTPA may be inappropriate and damaging (Anderson et al, 2007).

The appropriate clinical assessment of PE probability is guided by two main set of rules; namely the Wells rules and the Geneva rules (Ceriani et al, 2010). Several studies have confirmed the reliability of these rules in their revised versions (Lucassen et al, 2011) in properly diagnosing PE. Moreover as these rules are an intermediary of clinical suspicion of PE, study has shown that they offer similar conclusions in terms of the appropriate diagnostic test to be carried out in similar cases (Douma et al, 2011).

For those patients that exhibit exclusion probability during the clinical assessment a further D-dimer test is usually carried out so as to completely rule out PE (Huisman and Klok, 2013). For patients with PE they are usually associated with high D-dimer levels, i.e. increased fibrin formation which usually results to several symptoms such as malignancy and trauma (Huisman and Klok, 2013). The use of normal, moderate and highly sensitive D-dimer tests would help in completely ruling out PE in patients who exhibited unlikely probability of PE (Jaffrelot et al, 2012).

However the observation made from the rules and tests above is not conclusive and would require imaging to diagnose a patient with PE. These rules re merely to ensure that there is exclusion of mass but aimless testing for PE as they allow a doctor to selectively recommend diagnostic testing after the probability assessment test (Kiok et al, 2008)

Questions of maintaining this status quo and protecting peculiar-characteristic patients are now arising with the widespread prevalence of CTPA as an initial diagnostic measure rather than having clinical assessment (Righidi, Robert-Ebadi and Le Gal, 2017).

=Method=

Upon having an inclusion or likely probability from the clinical assessment, the diagnostic tests can now be applied. Whereas in the recent past imaging technology was not available everywhere (Alikhan et al, 2004) presently the concern is about its misuse in mass testing for PE due to the technological developments in radiography.

However the form of imaging technique deployed has a bearing on the specificity and clarity of the PE concerned. Hence this research will contrast the other diagnostic testing mechanisms against the use of CTPA in PE.

==Computed Tomography Pulmonary Angiography (CTPA)==

It is hailed as being the current modern benchmarking, ‘gold’, standard for PE diagnosis (Moore et al, 2017). It has had quite a development since its invention in the early 1990s where for around a decade it was rarely used due to the inability, then, to distinguish the pulmonary artery with the radiographic contrast during the scanning time which was approximately 3 minutes (de Monye and Pattynama, 2001). With the development of helical CT the ability to perform a continuous scan which took approximately 20 seconds reduced this inefficiency but still it was rarely used or tested properly (Kearon, 2001).

The ability of CTPA to include and exclude PE diagnosis and its wide availability also has an effect of reducing the cost of PE diagnosis (Mos et al, 2009). CTPA has also reduced the rate at which it gives a false feedback on PE (Dogan et al, 2015). The reason as to why it is mainly used as an excluding test is because of its high negative prevalence rate of approximately 99% (Dogan et al, 2015). Even for patients who may have tested with high D-dimer levels using the high sensitivity test, CTPA can safely be relied to exclude those patients upon a negative PE feedback (Galipienzo et al, 2010). Its high sensitivity and specifity at the rate of 96% even in cases of small blood vessels results into high quality imaging that easily highlights the magnitude of the emboli (Cronin, Weg and Kazerooni, 2008).

The effectiveness of the CTPA also depends on the nature of the CT used in that study has shown that where single-row detector CTPA is used the sensitivity goes as low as 20% for sub segmental emboli. On the other hand, multi-row detector CTPA increases sensitivity by a whopping 96% for segmental and central emboli (Remy-Jardin et al, 2007 and Carrier, Righini and Wells, 2010). However the use of CTPA in sub segmental PE has been greatly questioned as unnecessary due to the proximity of venous thromboembolism to cases of sub segmental PE as that would result to misdiagnosis or partial diagnosis (don Extar et al, 2013).(imaging by single-row detector ctpa contrasted with multi-row detector ctpa in the same case)

The use of CTPA has also helped in analysing the extent of the PE and the location of the embolus. In using the Qanadli scores to measure the PE severity in relation to 30-day to three month mortality rate has shown that the use of CT is very reliable. Where the CTPA gave a feedback of centrally located emboli in the blood vessels thus resulting to acute PE, there was increased mortality rate of up to 95% depending on the Qanadli scores generated that showed the percentage of blockage by the emboli (Vedovati et al, 2013).

==CTPA in relation to other Diagnostic tests==

===V/Q Scan===

Whereas it is acknowledged that CTPA is reliable in many PE cases, the other diagnostic tests are also available options that ought to be considered in relation to CTPA (Menno, Huisman and Klok, 2013). A proper viable option is Ventilation-perfusion lung scintigraphy otherwise known as V/Q scan. The imaging process of V/Q scan has a lower radiation exposure compared to CTPA approximated at 1.2Msv (O’Neill et al, 2005). It involves simultaneous scanning of the pulmonary arteries and airwaves and a feedback to exclude PE with no perfusion defect has a 3 month failure rate of 0.9%. When it is combined with a normal V/Q scan showcasing one segmental perfusion defect, the probability rate of PE diagnosis is at 85-90% (Kruip et al, 2003).

A. V/Q scan images for 41K counts 133-Xenon Energy window showing high perfusion ventilation.

B. V/Q scan images for 13k counts 133-Xenon Energy window showing low perfusion ventilation (Moore et al, 2018).

However V/Q scan has a major drawback due to its low probability PE result when there is a perfusion and ventilation in the same area yet when a further CTPA is conducted 30% of the negative PE turn positive. Further attempts to reduce the rate of negative prevalence in V/Q scan have been made for example through the development of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISAPED) criteria (Sostman et al, 2008). However, the accuracy of this criteria and others are

not intensively tested or studied hence they remain merely theoretical (Menno, Huisman and Klok, 2013).

===Magnetic Resonance Imaging (MRI)===

MRI just like its counterpart CT scan is commonly referred to as Magnetic Resonance Pulmonary Angiography (MRPA) for purposes of PE study. Whereas in the case of CTPA there must be use of contrasting agents injected in the veins, MRI requires no such agents hence it is even more attractable than CTPA at face value as patient characteristic indicate allergic reaction to contrasting agents. The catch in MRPA is that it requires a patient to hold breath for quite some time which is extremely difficult for many PE patients (Menno, Huisman and Klok, 2013). Moreover, studies done have indicated that there is an itching nondiagnostic ratio of up to 30% when compared to CTPA and V/Q scanning (Stein et al, 2010). It becomes completely useless if the patient has some metal implantation in the body as the word magnet would obviously suggest. Hence just like its counterpart V/Q scan, MRPA is not recommended essentially because its high nondiagnostic ratio would require further radiography thus pushing up the cost of PE diagnosis (Revel et al, 2013).

==CTPA in relation to PE relapse==

Menno, Huisman and Klok, 2013 opines that the prevalence to conduct trials in first time patients of PE has led to the complete ignorance on the use of CTPA in diagnosing PE recurrence. The available research on this indicates that there are two main pathophysiological factors for this (Menno, Huisman and Klok, 2013). Firstly, the sensitivity of D-dimer test enzymes is greatly decreased in recurring PE (Le Gal et al, 2006). Secondly, PE is in the nature of leaving residues of the embolus therefore it becomes difficult to distinguish between a recurrence and a residual PE occurrence (Stein, Yaekoub and Matta, 2010).

The available studies conducted on this show that where clinical probability assessment and D-dimer are used it returns 0% probability rate in three months hence no need to actually conduct a CTPA although that study also used low number of patients for trials hence may not be entirely accurate (Miniati et al, 2006). A more involving test was conducted on 516 patients out of whom 180 were excluded as low probability of recurrent PE. A further 88 were left unattended and no recurrent PE was recorded. Shockingly in a period of three months venous thromboembolism was diagnosed in up to 3.2% of the 88 patients; and as venous thromboembolism is a precursor of PE for up to 1.2 % of first PE cases (Nijkeuter et al, 2007). These patients had all returned CTPA negative hence causing reason to worry. The lack of accurate CTPA diagnosis for recurrent PE can be remedied by alternative methods of diagnosis and treatment.

Menno, Huisman and Klok, 2013 conducted a case study of a 52 year old patient who had been diagnosed with a CTPA PE four years before and was now experiencing symptoms of another PE and or venous thromboembolism. All procedures for PE diagnosis were conducted including a CTPA but they all returned a low probability PE. The researchers did not recommend alternative diagnostic measures such as ultrasonography but simply monitored him and administering nonsteroidal anti-inflammatory drug (NSAID) for two weeks. The patient experienced no further PE or venous thromboembolism symptoms for six months. The only logical conclusion is that CTPA remains largely untested in recurrent PE and from the existing trials; CTPA is not reliable just like the other diagnostic tests available for PE.

=Evaluation=

The role of CTPA in modern treatment of PE is unquestionably important, however it comes with its own drawbacks that are important to highlight as well as the potential room for improvement.

==Impact of the Contrast Agents used in CTPA==

As earlier highlighted the use of CTPA requires injection of contrast agent for proper visibility and high imaging quality of the concerned blood vessels. Generally there are no known gene mutation effects due to use of iodinated contrast (Dogan et al 2015). Moreover, even for pregnant women there is no adverse effect on the uterus of the iodinated contrast during or after the CTPA or CT generally (Bourjeily et al, 2010). The known health disadvantages are allergic reaction to the contrasting agent and contrast substance-induced nephropathy (CIN) (Dogan et al 2015). Where a PE patient is allergic to iodine, it is highly recommended that V/Q scan is used instead (Torbicki et al, 2008) since it has low radiographic impact on the body of which is a major concern in CTPA.

The necessity to use contrast is so as to create opacification of the pulmonary artery in contention. This allows the proper distinction of the embolus being targeted and the intravenous contrast injected in the blood supply. The density of the contrast is always dependent with the level of the PE diagnosis anticipated so much so that in the case of acute PE, the level of contrast goes as high as 211 HU while mild PE goes as low as 90HU.

In administration of the contrast for CTPA, there are three main approaches used to realize optimal result in diagnosis. First there is the empirical-based approach which does not take into account physiological factors such as breathing as it requires timing in administration of the IV contrast then a series of imaging is conducted by the CTPA (Ranji et al, 2006). Secondly, the bolus-tracking mechanism tracks the region of interest over the pulmonary artery along its axis. Multiple images are taken in the same position as the administered IV contrast takes effect, this continues until there is maximum opacification of the ROI. Finally the timing bolus mechanism may be relied. With it low content of IV contrast is administered at an approximate scan rate of 5 ml/s. A dynamic scan is conducted with each administered IV contrast material with a ROI over the pulmonary artery. Depending with the predetermined intended diagnosis between 60 to 150 ml of contrast material may be administered in bits. This final mechanism is the most appropriate as it takes care of several physiological factors, most importantly being breathing pattern.

As earlier mentioned, CTPA is preferred over MRPA since in MRPA patients would be required to hold breathe which may be problematic for patients with breathing problems. However, CTPA is also affected by breathing patterns and especially deep breathing patterns. This is because breathing causes artefact motion which may lead to inaccurate diagnosis of PE. Moreover, and most especially in empirical-based approach of contrast administration, it may result to reduced opacification of the pulmonary artery as the increased movement of oxygenated blood increases blood flow. To remedy this, some radiologists may apply the EKG gating to reduce the artefact motion however there has been no evidence that it affects the accuracy of the diagnosis process. (ekg gating reducing artefact motion)

Whereas various CT scans will require different contrast material, for CTPA which scans blood vessels it would require an intravenous contrast material such as an iodine-based or Gadolinium-based contrast material (Radiological Society of North America(RSNA), 2019). Iodine-based contrast, the most relied contrast in CTPA has several side effects which range from short to mild and even long term effects. However, the Radiological Society of North America in 2019 issued guidelines for patients cautioning that patients with preexisting kidney conditions or renal failure should undergo CT scanning under very careful conditions. The rationale was that they are at a high risk of suffering contrast-induced nephropathy (CIN). However, this is mostly associated with older contrast materials no longer in use (RSNA, 2019). Recently developed contrast materials have not shown a preference of causing CIN. Increased intake of fluid after administration of the contrast helps to alleviate most of the side effects associated with iodine-based contrast.

However, the proposal of increased periprocedural period for contrast administration poses another immense threat to the health of the patient; exposure to radioactivity. The radiation exposure of CTPA to the chest is at 10-70 mSv compared to meagre 1.5 mSv in V/Q scanning. The American College of Radiology recommends exposure limit of 3 mSv hence CTPA far much exceeds this limit. The threat is even greater to certain characteristic patients which include age difference. Due to lack of accurate data on the level of radiation exposure during the CTPA procedure, most researchers rely on the radiation level effects in the Hiroshima and Nagasaki atomic bombings in Japan during the Second World War. It was analyzed that exposure to 1 mSv resulted to 5 deaths in a group of 100,000 persons hence in the case of CTPA exposed patients; there would be 10 deaths per 100,000 CTPA patients. This on the face value is an enormous figure. There have been significant attempts to cure this defect by using lower tube voltage in the range of 80-100 kV which would then reduce the dose by 81% and still maintain quality imaging.

However, even with such developments the exposure to radiation for women which results to breast cancer and more so for pregnant women is still immense. Most radiologists try to limit the scope of the CTPA to the chest area with avoidance of the base of the lungs so as to protect the foetus in pregnant women. The current policy on PE imaging is the use of nuclear perfusion where possible instead of CTPA for pregnant women. However, there is no empirical evidence of the difference between the two in terms of radioactivity exposure. It is important therefore to ensure that more is done to improve efficiency of CTPA and reduce the radiation dose. This would require further advancements in CTPA

Ideally the usage of iodinated contrast material is meant to be excreted from the blood after the CTPA through the renal artery into the kidney and passed out as urine. The iodinated contrast material could be excreted entirely through the kidney within a day and more than half within two hours after administration. However, for patients who may have pre-existing glomerular filtration insufficiency, excessive contents of the contrast material may generate excessive osmotic pressure by increasing sodium and water output which increases pressure on the kidney tubular activities. It results to decreased glomerular filtration rate which then kick-starts renal failure (Thomsen and Morcos, 2003). The effects of CIN are dire as there is no known proper medication hence prevention remains to be the key most important approach. It is the third most common cause of hospital-induced renal failure or injury hence most scholars advice radiologists to exercise caution while administering the iodinated contrast material. It is advised that the CTPA procedure period ought to be increased to allow reduced administration of iodinated contrast material hence reducing pressure on the kidney.

Nevertheless, the intense use of contrasts during CT scans has shown to be of immense value where the CTPA is unable to improve cross-sectional imaging. It has been suggested that intense use of iodinated contrast, will give further details on blood flow intensity including pulmonary perfusion by looking at the concentration of the iodine in various anatomical areas (Wildberger et al, 2005). This may help to distinguish hypertension caused by embolism and non-embolic pulmonary hypertension (Fink et al, 2008).

=Conclusion=

In conclusion, PE is a condition that continues to affect many patients and cause high rate of hospital mortality. The effectiveness of clinical assessment in diagnosis of PE is limited to the extent of only including the likelihood of PE in a patient. As a result there has been increased dependency on imaging to properly diagnose PE.

CTPA is appreciated as the primary diagnostic imaging test to exclude or include PE. The use of CTPA has been effective even in diagnosis of segmental PE. The use of contrast coupled up with development in CT scanning technology; serve to make the use of CTPA even more reliable. Compared with other imaging technics, the accuracy of CTPA is unassailable. V/Q scanning despite having less radiation levels, does not give images with as much resolution as CTPA images.

However, the mass use of CTPA in total ignorance of the clinical assessment procedures continues to put many patients in danger of exposure to extreme radiation. Research has shown tremendous impact of high CTPA-related radiation levels on mostly pregnant women. Limiting the imaging area to slightly above the diaphragm has done little to reduce the risk of radiation-related cancers. Thus it is advised that CTPA should only be used as the first resort for exclusion testing of PE.

Whereas CTPA is a most conclusive diagnostic measure, its effectiveness in recurrent PE is yet to be fully tested. There is need to ensure consistency of places of testing so as to enable radiologists find out the effectiveness of CTPA in recurrent PE. Developing CT technology that has reduced radiation exposure to the levels of or below V/Q scan is paramount in creating a healthy post-CTPA situation.

Continue your exploration of Addressing the Global Health Challenge of Tuberculosis with our related content.

=References=

Bourjeily, G., Paidas, M., Khalil, H., Rosene- Montella, K., Rodger, M. 2010. Pulmonary embolism in pregnancy. Lancet; 375, pp. 500-12

Douma, R., Mos, I., Erkens, P., Nizet, T., Durian, M., Hovens, M., van Houten, A., Hofstee, H., Klok, F., ten Cate, H., Ullmann, E., Buller, H., Kamphuisen, P., Huisman, M., Prometheus Study Group. 2011. Performance of 4 clinical decision rules in the diagnostic management of acute pulmonary embolism: a prospective cohort study. Annals of Internal Medicine; 154(11), pp. 709-18

Huisman, M & Klok, F. 2013. How I diagnose acute pulmonary embolism. Blood; 121(22), pp. 4443-8

Kapetaniou, A. 2011. Pulmonary embolism-Diagnosis and treatment. Medical Analects; 9 (3), pp. 390- 394.

Kourkouta, L., Koukourikos, K., Iliadis, .C, Ouzounakis, P., Tsaloglidou, A. 2018. Pulmonary Embolism: A Literature Review. American Journal of Nursing Science. Special Issue: Nursing Education and Research. Vol. 7, No. 3-1, pp. 57-61. doi: 10.11648/j.ajns.s.2018070301.19

Kruip, M., Leclercg, M., van der Heul, C., Prins, M., Buller, H. 2003. Diagnostic strategies for excluding pulmonary embolism in clinical outcome studies. A systematic review. Annals of Internal Medicine; 138(12), pp. 941-51

Lucassen, W., Geersing, G., Erkens, P., Reitsma, J., Moons, K., Buller, H., van Weert, H. 2011. Clinical decision rules for excluding pulmonary embolism: a meta- analysis. Annal of Internal Medicine; 155(7), pp. 448-60

Menno, V., Huisman., Frederikus, A. 2013. How I diagnose acute pulmonary embolism Blood. First Edition paper. DOI 10.1182/blood-2013-03-453050.

Miniati, M., Monti, S., Bottai, M., Scoscia, E., Bauleo, C., Tonelli, L., Dainelli, A., Giuntini, C. 2006. Survival and Restoration of Pulmonary Perfusion in a Long- Term Follow- up of Patients After Acute Pulmonary Embolism. Medicine; 85(5), pp. 253-262 (viewed 05 April 2020) Available from:

Nijkeuter, M., Sohne, M., Tick, L., Kamphuisen, P., Kramer, M., Laterveer, L., van Houten, A., Kruip, M., Leebeek, F., Buller, H., Huisman, M., Christopher Study Investigators. 2007. The natural course of hemodynamically stable pulmonary embolism: Clinical outcome and risk factors in a large prospective cohort study. Chest; 131(2), pp. 517-23

Patakas, D. 2006. Otorhinolaryngology. Publishing University Studio Press, Thessaloniki

PIOPED Investigators. 1990. Tissue plasminogen activator for the treatment of acute pulmonary embolism. Chest; 97, pp. 528-33

Remy-Jardin, M., Pistolesi, M., Goodman, L.R. 2007. Management of suspected acute pulmonary embolism in the era of CT angiography: a statement from the Fleischner Society. Radiology; 245:315– 329.

Righini, M., Robert-Ebadi, H., Le Gal, G. 2017. Diagnosis of acute pulmonary embolism. J Thromb Haemost; 15, pp. 1251–61

Sostman, H., Miniati, M., Gottschalk, A., Matta, F., Stein, P., Pistolesi, M. 2008. Sensitivity and specificity of perfusion scintigraphy combined with chest readiography for acute pulmonary embolism in PIOPED II. J Nucl Med; 49(11), pp. 1741-8

Stagaki, E., Apostolakis, S., Karakontaki, F., Tsaramirsi, M., Poluchronopoulos, B., Konstantinidis, S. 2012.

Pulmonary Thromboembolic Disease. Hellenic Cardiological Review; 53, pp. 159-173

Stein, P.D., Yaekoub, A.Y., Matta, F. 2010. Resolution of pulmonary embolism on CT pulmonary angiography. AJR Am J Roentgenol; 194(5): 1263-1268.

Tsilimidos, G., Marinis, A. 2013. Review of the Diagnostic and Therapeutic Approach of Pulmonary Embolism. Epistimonika Xronika; 18(2), pp. 88-91.

Tsakali, A., Kefalopoyloy, Z., Papadoulas, S., Argitis, B., Liamis, A., Tsolakis, I. 2007. Deep venous thrombosis: etiology, prophylaxis, treatment. Greek Vascular Surgery; 8 (9), pp. 251-258.

Tsuchiya, N., van Beek, J.R., Ohno, Y., Hatabu, H., Kauczor, H., Swift, A., Vogel-Claussen, J., Biederer, J., Wild, J., Wielputz, M.O. and Schiebler, M.L. 2018. Magnetic resonance angiography for the primary diagnosis of pulmonary embolism: a review from the international workshop for pulmonary functional imaging. World J Radiol; 10(6), pp.52-64

Vedovati, M., Germini, F., Agnelli, G., Becattini, C. 2013. Prognostic role of embolic burden assessed at computed tomography angiography in patients with acute pulmonary embolism: a systematic review and meta- analysis. J Thromb Haemost; 11(12), pp. 2092-102