Volume 12
Issue 03
September 2024
Inside This Issue
Editorial, 2-3
Technology Corner, 4-5
Tips from the Experts, 6-8
Humanitarian News, 9-14
Best Image Contest, 15
WABIP News, 16
Research, 17-18
Links, 19
To Stent or Not to Stent: That’s Just the First Question
WABIP Newsletter
S E P T E M B E R 2 0 2 4 V O L U M E 1 2 , I S S U E 3
EXECUTIVE BOARD
Stefano Gasparini, MD
Italy, Chair
Pyng Lee, MD, PhD
Singapore, Vice-Chair
Hideo Saka, MD
Japan , Immediate Past-
Chair
Silvia Quadrelli, MD
Membership Commiee
Chair
Jean-Michel Vergnon, MD
Educaon Commiee
Chair
Ali Musani, MD
Finance Commiee Chair
Naofumi Shinagawa, MD
Japan,
Secretary General
Menaldi Rasmin, MD, PhD
Indonesia , President
WCBIP 2024
Rajesh Thomas, MD, PhD
Melbourne , President
WCBIP 2026
STAFF
Michael Mendoza
General Manager
Judy McConnell
Administrator
Kazuhiro Yasufuku
Newsleer Editor-in-chief
P A G E 2
Airway stenng has long been a signicant part of
the intervenonal bronchoscopist’s repertoire when
it comes to endoscopic management of malignant
central airway obstrucon (MCAO). However, there
is marked variability in how and when this interven-
on is oered.
1,2
Even the generally held noon that
stenng should be reserved for cases involving ex-
trinsic airway compression (whether exclusively or
combined with endoluminal disease) is not univer-
sally adhered to.
3
Perhaps unsurprisingly, no guide-
lines had previously been published to help clinicians
on this subject.
That changed with the recent publicaon of the
World Associaon for Bronchology and Intervenon-
al Pulmonology (WABIP) guidelines on airway
stenng for MCAO.
4
A group of 17 experts across 11
countries spanning four connents addressed six
important quesons with mutual consensus (using
the modied Delphi technique) based largely on a
systemac review of published literature.
Only one of the six recommendaons was graded as
“strong” - namely that airway stenng be considered
in paents with MCAO receiving mechanical venla-
on - although the supporng evidence based on
retrospecve observaonal data was judged to be of
low quality (grade 1C recommendaon).
5
Weak rec-
ommendaons - made based on somemes con-
icng evidence that was deemed low-to-moderate
quality - were also made to consider airway stenng
as a means to improve quality of life, performance
status, and survival. Another weak recommendaon
(grade 2C) was to perform surveillance bron-
choscopies to detect stent-related complicaons in
asymptomac paents, with the rst surveillance bronchos-
copy scheduled 4-6 weeks aer stenng. Across all clinical
scenarios, the expert panel determined that no conclusive
evidence supported the selecon of one over the other
commercially available stent type (silicone versus metallic).
Finally, in the absence of any evidence for or against it, the
group’s consensus was in favor of undertaking pulmonary
hygiene measures such as saline nebulizaon to reduce the
risk of stent-related complicaons (nine experts strongly
agreed, seven agreed, and one neither agreed nor disa-
greed).
So, how does the bronchoscopist incorporate these guide-
lines going forward? The very rst queson facing every
bronchoscopist is whether to stent or not. Hamstrung by
limited scienc evidence, these guidelines parally help
answer this queson. However, the role of ascertaining the
degree of airway obstrucon in making treatment decisions
remains unclear. For example, if there is no respiratory dis-
tress or other signs and symptoms aributable to MCAO,
what if any is the minimum degree of extrinsic airway com-
pression that warrants airway stenng? Does that vary with
the locaon of MCAO (e.g., trachea versus a mainstem
bronchus versus bronchus intermedius)? What is an appro-
priate metric for quanfying airway obstrucon? Is it airway
diameter, airway cross-seconal area, length of airway
aected, and/or drop in airway pressure across the aected
segment? What is a valid means of measuring airway size?
Is it the bronchoscopist’s esmate based on endoscopic
images, automated measurements on segmented endo-
scopic images, pre-operave or intraoperave computed
tomographic imaging (stac or dynamic), or bronchoscopic
tools such as a calibrated airway balloon? Of course, lest we
forget, to stent or not to stent is merely the rst queson
out of many. The remaining quesons - including but not
Ali I. Musani MD, FCCP
Professor of Medicine and
Surgery
Director, Complex Airway Pillar
of the Center for Lung and
Breathing
Director, Intervenonal
Pulmonology Program
Director, Global Health
Pathway, Internal Medicine
Residency Program
Division of Pulmonary Sciences
& Crical Care Medicine
Majid Shaq MD, MPH
Medical Director, Bedside
Procedure Service
Medical Director, Intervenonal
Pulmonology
Div. of Pulmonary & Crical Care
Medicine, BWH
Assistant Professor of Medicine,
Harvard Medical School
W A B I P N E W S L E T T E R
P A G E 3
limited to stent selecon, airway hygiene regimen,
duraon of stenng, and surveillance regimen -
warrant further invesgaon as aptly highlighted
by these guidelines.
The WABIP guideline panel is to be congratulated
for producing this valuable addion to the body of
medical literature. In addion to highlighng the
potenal of airway stenng for improving clinical
outcomes, parcularly liberaon from mechanical
venlaon, these guidelines provide needs assess-
ment for future invesgators as they seek more
precise answers to several important quesons.
The journey to discovery is not easy, as exemplied
by a commendable randomized study that was
stopped far short of target enrollment.
3
But, as the
WABIP guideline panel has successfully demon-
strated, there is a rst me for everything.
References:
1. Dutau H et al. Respiraon; internaonal review of
thoracic diseases. 2018;95(1):44-54.
2. Wayne MT et al. Respiraon; internaonal review of
thoracic diseases. 2023;102(8):608-612.
3. Dutau H et al. Respiraon; internaonal review of
thoracic diseases. 2020;99(4):344-352.
4. Chaddha U et al. Respirol. 2024;29(7):563-573.
5. Oki M et al. J Thorac Dis. 2017;9(9):3154-3160.
W A B I P N E W S L E T T E R
P A G E 4
Technology Corner
Review on Arcial Intelligence in Intervenonal Pulmonology
Introducon
Recent advancements in arcial intelligence (AI) technology have signicantly impacted various elds. AI, which emulates human
intelligence through computaonal systems, includes deep learning—a subset of machine learning that employs deep neural net-
works and other models to autonomously learn from data. This powerful technique enhances complex mathemacal models for
predicon, driving progress in areas such as image recognion, disease diagnosis, and prognosis predicon.
Background
AI technologies are extensively applied and researched in healthcare, including the eld of pulmonary medicine. Over the past dec-
ade, automated chest X-ray image analysis techniques for early tuberculosis screening have rapidly evolved.
1
Several commercial
products based on deep learning computer-aided detecon (CADe) systems for pulmonary tuberculosis are now available.
2
AI-based
computer-aided diagnosis (CADx) in addion to CADe systems are also employed for lung cancer screening and diagnosis.
3-6
AI in the eld of intervenonal pulmonology
AI in Airway Anatomy
Accurate transbronchial diagnosis for peripheral pulmonary lesions requires selecng the correct airway to reach the target.
7
Tradi-
onally, bronchoscopists view two-dimensional (2-D) axial, coronal, and sagial CT views and reconstruct a 3-D mental image of the
path to the target lesion. However, this is challenging even for experienced bronchoscopists.
8
Automated bronchial tree labeling has
been researched for more than two decades.
9, 10
AI-based bronchial tree labeling has been integrated into current navigaon tech-
nologies; including virtual bronchoscopic navigaon, electromagnec navigaon bronchoscopy, and augmented uoroscopy. Even
during bronchoscopic procedures, it can be dicult to determine which bronchus is being observed which can be especially dicult
for novices or operators unfamiliar with bronchoscopy. AI models have shown high performance in anatomical idencaon using
bronchoscopic video images and trained convoluonal neural networks (CNN).
11-13
These promising results suggest that future AI-
integrated devices could enable real-me bronchial anatomy idencaon. A recent study has trained AI to recognize depth within
the bronchial lumen using bronchoscopic images generated from CT scans, allowing the AI to assist in steering a roboc broncho-
scope to peripheral bronchi while maintaining a clear view of the bronchial lumen.
14
This ‘AI co-pilot’ technology is ancipated to be
used clinically in the future, especially in roboc bronchoscopy, helping improve the learning curve for beginners.
AI in EBUS Images
Convex probe EBUS is widely used for lymph node staging in lung cancer and diagnosing mediasnal diseases due to its diagnosc
ecacy and safety prole.
15
Categorizing lymph node features within EBUS B-mode images has been explored to disnguish be-
tween metastac and benign lymph nodes.
16, 17
However, this method is inherently subjecve and reliant on observer discreon.
CADx has been reported for predicng lung cancer metastasis in lymph nodes using EBUS images. Early studies employed arcial
neural networks with supervised learning on regions of interest from B-mode images.
18
However, these approaches relied on super-
vised selecon of regions of interest, potenally liming real-me applicability. Consequently, some researchers applied CNN to en-
Tsukasa Ishiwata, MD, PhD
Research Fellow,
Division of Thoracic Surgery,
Toronto General Hospital,
University Health Network
Kazuhiro Yasufuku, MD, PhD, FRCSC
Head, Division of Thoracic Surgery,
University Health Network,
Professor and Chair, Division of Thoracic
Surgery, University of Toronto
W A B I P N E W S L E T T E R
P A G E 5
re EBUS images without selecng regions of interest,
19, 20
or to mul-modal images including EBUS, Doppler, and elastographic im-
ages,
21
yielding a high diagnosc accuracy. Similarly, AI-based CADx using radial probe EBUS images for peripheral pulmonary lesion
diagnosis has demonstrated high performance.
22, 23
Currently, no commercially available AI-based systems oer real-me diagnosis
predicon from EBUS images, but it is foreseeable that real-me dierenaon between benign and malignant lesions in EBUS imag-
es will become a valuable clinical tool. This could enhance diagnosc yield by more accurate pre-test probability esmaon before
sampling.
AI in Cytology
Opmal diagnosis in intervenonal pulmonology, parcularly through ssue sampling, can be enhanced by rapid on-site cytologic
evaluaon (ROSE).
24-26
However, the widespread adopon of ROSE faces challenges due to a shortage of experienced on-site cyto-
pathologists and nancial constraints. To address these issues, AI has been integrated with ROSE in research sengs. High diagnosc
performance has been demonstrated by inpung small pixel patches from cytology slide images into CNN models, enabling diagnosis
within seconds.
27
A CNN-based system for categorizing cytology smear images obtained during ROSE in EBUS-TBNA has shown high
accuracy in classifying adequate/inadequate samples, granulomas, and malignant cells.
28
Another AI-based ROSE system has also
shown high diagnosc performance and consistency with experienced cytopathologists using external test datasets.
29
These advance-
ments suggest that AI could enhance diagnosc accuracy and eciency in cytology, providing substanal support in sengs where
experienced cytopathologists are unavailable.
Conclusion
The integraon of AI into intervenonal pulmonology has the potenal to improve diagnosc accuracy and paent safety. However,
the clinical impact of AI-enhanced procedures remains to be fully established. Further research is needed to evaluate the clinical ben-
ets and limitaons of AI in intervenonal pulmonology, as well as to consider the ancipated increase in costs and assess their align-
ment with eecveness.
References
1. Santosh KC et al. J Med Syst. 2022;46(11):82.
2. Hua D et al. Int J Med Inform. 2023;177(105159.
3. Thong LT et al. Lung Cancer. 2023;176(4-13.
4. Mikhael PG et al. J Clin Oncol. 2023;41(12):2191-2200.
5. Kanava F et al. Sci Rep. 2020;10(1):9297.
6. Liu Q et al. Transl Lung Cancer Res. 2020;9(3):549-562.
7. Ishiwata T et al. Expert Rev Respir Med. 2019;13(9):885-897.
8. Dolina MY et al. Chest. 2008;133(4):897-905.
9. Kitaoka H et al. Medical Image Compung and Computer-Assisted Intervenon — MICCAI 2002. Lecture Notes in Computer Science, vol 2489.
Springer.
10. Ota S et al. Medical Imaging and Augmented Reality. MIAR 2008. Lecture Notes in Computer Science, vol 5128. Springer.
11. Yoo JY et al. Sci Rep. 2021;11(1):23765.
12. Li Y et al. Transl Lung Cancer Res. 2022;11(11):2261-2274.
13. Chen C et al. Ther Adv Chronic Dis. 2023;14(20406223231181495.
14. Zhang J et al. Nat Commun. 2024;15(1):241.
15. Dong X et al. Ann Thorac Surg. 2013;96(4):1502-1507.
16. Fujiwara T et al. Chest. 2010;138(3):641-647.
17. Hylton DA et al. J Thorac Cardiovasc Surg. 2020;159(6):2499-2507.e2493.
18. Tagaya et al. Chest. 2008;133(1):137-142.
19. Churchill IF et al. Ann Thorac Surg. 2022;114(1):248-256.
20. Ito Y et al. Cancers (Basel). 2022;14(14).
21. Li J et al. Endosc Ultrasound. 2021;10(5):361-371.
22. Chen CH et al. Comput Methods Programs Biomed. 2019;177(175-182.
23. Hoa T et al. Sci Rep. 2022;12(1):13710.
24. Oki M et al. Respiraon. 2013;85(6):486-492.
25. Trisolini R et al. Chest. 2015;148(6):1430-1437.
26. Shikano K et al. J Thorac Dis. 2020;12(6):3057-3064.
27. Lin CK et al. Cancer Med. 2021;10(24):9047-9057.
28. Asfahan S et al. Eur Respir J. 2021;58(4).
29. Yan S et al. Front Oncol. 2024;14(1360831.
Tips from the Experts
P A G E 6 V O L U M E 1 2 , I S S U E 3
Introducon:
Subgloc stenosis, which includes benign tracheal stenosis (TS) presents a complex airway condion typically aributed to prior history of
endotracheal intubaon and tracheostomy placement (1). In the absence of a known cause, it is known as idiopathic subgloc stenosis.
Damage to respiratory ssues can be due to pressure necrosis, which triggers inammaon, leading to the development of granulaon s-
sue and eventual formaon of scar ssue that is akin to keloids in the skin (2). Input from a muldisciplinary team involving intervenonal
pulmonology, otolaryngology, and thoracic surgery is oen needed to opmize care for paents with TS. While open surgery oers deni-
ve treatment, paents may not qualify due to their comorbidies and symptom severity necessitang an urgent intervenon. As an alter-
nave, endoscopic procedures focusing on scar resecon, scar lysis, airway dilaon, or a combinaon of these have become more common
pracce. However, a signicant drawback of endoscopic treatment is rate of recurrence and the need for subsequent procedures (3). Serial
intralesional steroid injecon (SILSI), is an aracve opon for decreasing the frequency of recurrence of benign TS in an oce seng with
readily accessible equipment and without the need for general anesthesia.
Raonale for using steroids:
Dermatologists have ulized intralesional steroids since the 1960s to modify skin scars (keloids), achieving response rates ranging from 50%
to 100% with recurrence rates between 9% and 50% (4). Steroids eecvely reduce inammaon that promotes scar formaon by inhibing
collagen and glycosaminoglycan synthesis, encouraging broblast degeneraon, and enhancing collagenase acvity (5). They also inhibit
inammatory cell migraon and cytokine producon, such as transforming growth factor-β and interleukins, crucial in scar development (5).
Steroids induce hypoxia and vasoconstricon, potenally reducing scar volume and soening scar ssue by liming nutrient supply (5). Tri-
amcinolone has been shown to be an eecve intralesional treatment for keloids and hypertrophic scars as well as benign subgloc steno-
sis and TS (3,6,7). Here we discuss our instuon’s pracce of oce-based SILSI for management of TS.
Planning and Paent selecon:
Paents selected for SILSI at our instuon are those with idiopathic subgloc stenosis (iSGS) or iatrogenic SGS who have undergone mul-
ple endoscopic dilaons in the operang room (OR) and wish to prolong intraoperave interval and those with iSGS declining any surgical
intervenon. Franco et al. rst proposed oce-based SILSI as either a sole opon or as an adjunct to surgery for paents with iSGS (3). Their
group ulized spirometry to assess peak inspiratory ow (PIF) and peak expiratory ow (PEF), with the percentage of predicted peak expira-
tory ow (%PEF) serving as an indicator of stenosis severity before, during, and aer treatment. Paents who were symptomac with a %
PEF below 50% and a PIF less than 1.5 L/sec were advised to undergo surgery. Discussing pros and cons of OR treatment vs. awake SILSI and
considering paents’ preference in the context of their current symptom burden is imperave to choosing the opmal treatment opon.
Most paents receive at least one operave dilaon prior to starng SILSI.
Steroid Injecon for Benign Tracheal Stenosis
Megan S. Wu, MD
University of Chicago,
Department of
Otolaryngology
Shreya Podder, MD
University of Chicago,
Department of
Pulmonary and Crical
Care Medicine,
Intervenonal
Pulmonology
Brandon Baird, MD
University of Chicago,
Department of
Otolaryngology
Tips from the Experts
P A G E 7 V O L U M E 1 2 , I S S U E 3
Technique:
We perform our oce-based SILSI through a transcervical approach. The procedure involves two operators: one controls the exible laryngo-
scope while the other performs the injecon. Aer obtaining informed consent, we rst spray a combinaon of oxymetazoline and 4% lido-
caine into the nasal cavity, followed by topical applicaon of 4% lidocaine to anestheze the oropharynx and hypopharynx. Next, we palpate
the anterior neck to locate the cricothyroid space and inltrate approximately 1mL of 2% lidocaine. At this me a tracheal block may also be
performed. A exible laryngoscope is passed through the nose past the level of the larynx to achieve a transgloc view of the subglos. Un-
der direct visualizaon, a 25-gauge needle is passed through the cricothyroid membrane and 2mL of Kenalog (40mg/mL) is injected circum-
ferenally into the subglos. Proper placement of steroid is conrmed by observing ssue blanching and wheal-like expansion within the
subgloc scar. Counseling paents both before and during the procedure is essenal, emphasizing the importance of slow, deep breathing
and maintaining a calming environment.
Quality Control:
Key features of the area of stenosis, including posion and degree of narrowing, are visually compared with video documentaon at each
follow-up visit to determine if there is disease progression, improvement, or stability. Spirometry results and paents’ funconal status are
factored into the decision to connue or stop SILSI. Homan et al. evaluated paents undergoing SILSI for iSGS, 14 of 17 paents (82%) com-
pleng one series of injecons did not require further operave intervenon to date. For paents who completed two sets of three injec-
ons, average stenosis decreased signicantly from 40% to 20% (8). Similarly, Bertelsen’s study found that 17 of 24 paents (71%) did not
need further surgery aer SILSI during mean follow-up me of 32 months. When comparing surgery-free interval (SFI) before and aer SILSI,
SFI improved from 10.1 months before, to 22.6 months aer SILSI (7). Current studies are limited by sample size, length of follow-up and
rater bias of stenosis percentages.
Conclusion:
SILSI is a well-tolerated, safe, and eecve treatment modality with the potenal to alleviate the surgical burden on paents with benign TS.
The eecveness of SILSI has been shown in TS due to various eologies including idiopathic, rheumatologic, and traumac causes. In our
clinical experience, we recommend SILSI as an adjunct to repeated endoscopic dilaon.
Next steps and future direcons
Further research into the frequency and interval between injecons is needed to help opmize paent’s symptoms and funconal status and
to risk strafy and idenfy possible responders.
References
1. Gelbard A et al. Laryngoscope. 2015; 125(5).
2. Nouraei SAR et al. Laryngoscope. 2006; 116.
3. Franco RA et al. Laryngoscope. 2018;128(3).
4. Robles DT et al. Clin Dermatol. 2007; 25(1).
5. Gauglitz GG et al. Molecular Medicine. 2011;17(1–2).
6. Ledon JA et al. Dermatologic Surgery. 2013; 39.
7. Bertelsen C et al. JAMA Otolaryngol Head Neck Surg. 2018;144(3).
8. Homan MR et al. Laryngoscope. 2017; 127 (2475-2481).
Tips from the Experts
P A G E 8 V O L U M E 1 2 , I S S U E 3
A B
C
D
Figure: A – A paent with idiopathic subgloc stenosis who underwent dilaon in the operang room.
B – The paent retured two months later for SILSI. C - 1 month aer SILSI. D – Signicant improvement
seen 10 months aer two rounds of SILSI.
Humanitarian News
W A B I P N E W S L E T T E R P A G E 9
Democracy Under Siege: The Dual Threat of AI and Disinformaon. Can We
Safeguard Our Future?"
The digital age has ushered in an unprecedented era of technological advancements, revoluonising the ways in which we
communicate, interact, and access informaon. However, this rapid progress has also given rise to a new froner of cyber
threats, characterised by their sophiscaon, adaptability, and far-reaching social implicaons.
The informaon age has inaugurated a period dominated not only by the disseminaon of knowledge but also by the wide-
spread propagaon of misinformaon and disinformaon. The arrival of the informaon age has ushered in an era charac-
terised not only by the worldwide distribuon of knowledge but also by the extensive circulaon of false and misleading
informaon. Facilitated by the fast disseminaon and extensive data processing capabilies of the internet, false informaon
has become a powerful instrument in internaonal compeons and domesc polical disputes. Both governmental and
non-governmental enes deploy false informaon to inuence worldwide public senment, provoke chaos, and undermine
condence. Arcial intelligence (AI), specically machine learning (ML), is intended to enhance these disinformaon cam-
paigns, which are clandesne operaons specically created to deliberately spread inaccurate or decepve material. The
emergence of AI-driven cyber manipulaon is fundamentally changing the current state of aairs, modifying the dynamics of
public discussion, and presenng a substanal obstacle to the integrity of democrac processes on a global scale.
Intenonal disinformaon campaigns are carefully planned and executed through well dened phases. Firstly, operaves
engage in reconnaissance to observe their target area and perform thorough studies of the demographics they aim to ma-
nipulate. A fundamental infrastructure is developed, including of messengers, credible personalies, social media proles,
and groups to eciently disseminate their stories. An uninterrupted stream of material, including wrien arcles as well as
mulmedia elements like photographs, memes, and videos, is crucial to guarantee the spread and acceptance of their mes-
saging. Once disseminated on the internet, these fragments of false informaon are magnied by automated bots, plaorm
algorithms, and social engineering taccs to enhance their eecveness and inuence. However, much beyond simple distri-
buon, the ulmate success oen depends on connuous interacon with unwary users, accomplished through strategies
similar to digital trolling, the combave counterpart in the domain of decepon. In its nal stages, a successful disinfor-
maon campaign alters public opinions, aects unwary individuals, and may even provoke disrupve behaviours, therefore
sustaining disorder.
Irrespecve of their source, disinformaon campaigns that culvate genuine followers can eortlessly assimilate into society
conversaons, obscuring the disncon between reality and falsehood and undermining public condence. This process of
erosion presents a signicant obstacle to a society's capacity to dierenate between truth and decepon, therefore giving
rise to a persistent lack of trust.
Disinformaon campaigns, albeit disnct in their parcularies, exhibit shared features and adhere to idenable paerns
that take advantage of the structural features of social media plaorms. Adopng a comprehensive paradigm to capture all
misinformaon operaons is dicult because of the varied ecosystems of actors, strategies, and plaorms involved. Moreo-
ver, the involvement of convenonal media in this ecosystem, whether as a target, plaorm, or as a deliberate or uninten-
onal facilitator of inuence, introduces an addional level of intricacy. Consequently, the government, corporate sector,
and civil society stakeholders involved in invesgang, disrupng, and countering disinformaon have not yet reached a
consensus on a cohesive methodology to eecvely address the danger posed by disinformaon operaons.
The potenal inuence of generave arcial intelligence (AI), namely large language models (LLMs), on democrac process-
es and cizenship is substanal. It is widely acknowledged among scholars that generave AI will have a signicant impact
on the informaon landscape and social media, both of which are essenal components for democrac cizenship. The au-
thors propose a hybrid methodology for comprehending these eects, acknowledging the complex interdependence of digi-
tal technologies with user interests and the blurred borders between online and oine domains.
Humanitarian News
W A B I P N E W S L E T T E R P A G E 10
Generave arcial intelligence (AI), parcularly large language models (LLMs), is poised to signicantly impact democrac
processes and cizenship. Several researchers agree that generave AI will profoundly inuence the informaon landscape
and social media, which are crucial inputs for democrac cizenship. They advocate for a hybrid approach to understanding
these impacts, recognising the intricate interconnectedness of digital technologies with user interests and the diuse bound-
aries between online and oine spheres.
AI has the capacity to generate certain benecial impacts on democrac cizenship. One potenal benet of AI is its ability
to automate monotonous jobs, therefore allowing cizens to allocate their me and focus towards more signicant polical
involvement. Furthermore, it has the potenal to enable extensive polical deliberaons facilitated by arcial intelligence,
surpassing geographical limitaons. Furthermore, generave arcial intelligence has the capacity to improve the availability
and analysis of informaon, so oering more carefully selected material to guide reasoning. By fostering democrac reci-
procity and elevang the tone of talks, interacve Language Learning Modules (LLMs) have the potenal to develop people'
argumentave and deliberave abilies, hence boosng the quality of online polical discussions. However, the majority of
experts studying the phenomena of new technologies stress the need of examining generave AI in conjuncon with social
media, rather than in isolaon. Their contenon is that the collecve impact of these technologies on the aenon economy
may have substanal social and polical consequences. Within the framework of deliberave democracy, individuals are
expected to culvate certain values, like acknowledging others as equals, creavely embracing other perspecves, and ex-
hibing integrity and respect throughout discussions. Through the provision of balanced arguments, analysis of reasoning
aws, and facilitaon of the comprehension of alternave viewpoints, generave AI has the ability to foster these delibera-
ve virtues. Furthermore, it has the potenal to tackle two notable deciencies in deliberave democracy: argumentaon
inequity and scalability. By supporng individuals with limited communicaon and reasoning abilies, arcial intelligence
(AI) has the potenal to migate disparies in deliberaon. Furthermore, its ability to scale up could surpass the praccal
constraints of extensive, high-quality deliberave parcipaon, so enabling a more inclusive democrac process by providing
informaon and engaging cizens, fostering civility and empathy. This has the potenal to signicantly enhance cizens' abil-
ity to engage in meaningful polical discussions.
However, it must never be forgoen that AI is neither autonomous nor neutral; rather, it is laden with the intenons of
those who shape its content. AI has the potenal to produce low-quality, biased, or incorrect informaon due to hallucina-
ons and biased training data. This could have serious consequences if cizens rely on such outputs for polical delibera-
ons and decision-making. Another signicant threat is the proliferaon of deepfakes and mass disinformaon. The capacity
of Generave AI to produce convincingly false content at scale could undermine trust in digital media and erode the founda-
ons of informed democrac discourse. The potenal for Generave AI to be weaponised in mass disinformaon campaigns
has been highlighted. Well-funded actors could ood the informaon landscape with false narraves, potenally dominang
the aenon economy and intensifying social division and distrust.
An addional risk is that Generave AI may diminish cizens' crical thinking and deliberaon capabilies. By replacing the
cognive work necessary to understand complex issues, AI could se essenal skills for a healthy democracy. There is also
the risk that Generave AI could replace interpersonal interacons in polical discourse, leading to a decline in direct en-
gagement among cizens, which could damage the sense of belonging to a common polical community. Several authors
have warned of the possibility of an "infocalypse," in which cizens become unable to disnguish between real informaon
and that manipulated by AI, potenally leading to a form of "Armed Skepcism," where truth becomes obscured amid con-
icng narraves.
Although disinformaon is not a new phenomenon, the advent of social media technologies has fundamentally altered the
scale, reach, and accuracy of informaon disseminaon in the digital age. The digital world has become a baleground for
control of informaon, with malicious actors exploing AI technology to create and disseminate disinformaon on an un-
precedented scale. By mimicking human behaviours and generang images and messages that resonate with audiences,
these adversaries can subtly alter the truth, blurring the lines between fact and con. The power of AI-driven disinfor-
maon lies in its ability to exploit exisng social divisions, amplifying controversial issues and exacerbang polarizaon. Sa-
ya Umoja Noble's book, "Algorithms of Oppression" (2018), examines how search engines and AI systems can perpetuate
racial and gender biases. She argues that these technologies, far from being neutral, oen reect and amplify social prejudic-
es, which could undermine democrac principles of equality.In what has come to be termed the "informaon war," infor-
maon becomes a weapon, and the minds of cizens become the baleeld. This strategy aims to polarise civil society, sow
Humanitarian News
W A B I P N E W S L E T T E R P A G E 11
chaos, and weaken opposing states. Far from being a mere dystopian hypothesis, this war is ongoing; it has dramacally in-
uenced the posioning of unexpected individuals at the highest levels of government, is dicult to detect, and penetrates
all levels of society at a relavely low cost. The ongoing informaon war, while lacking the overt violence of tradional com-
bat, represents signicant threats to naons, businesses, and cizens alike.
Democrac governments face the challenge of simultaneously ensuring cizens' access to reliable and objecve informaon
while avoiding acts of censorship. This necessity to avoid liming freedom of expression leads to a reliance on large tech
companies to oversee the online informaon environment, thereby granng powerful private enes—oen resource-rich
and with their own agendas and interests—the power of censorship.
One of the most well-known and painful examples for the healthcare community has been the disinformaon campaigns
conducted during the Covid-19 pandemic in Europe, which demonstrated the complexity of idenfying the sources and large
strategies behind these operaons, given their mulfaceted and oen simultaneous nature. However, this also brings to
light the geopolical dimensions of cyberspace, revealing the risks of cybersecurity failures since the mid-2000s. Idenfying
sources of disinformaon is parcularly challenging due to the anonymous nature of cyberspace. Nevertheless, scholars
have idened several categories of actors most likely to produce and disseminate false informaon, including far-right
news sources, disinformaon operaons by foreign states, polical pares promong naonalist content, and prot-driven
content producers exploing the aenon economy.
Cyberspace today represents both a domain for conict and a means for states to advance their interests. The poorly regu-
lated nature of cyberspace poses a signicant challenge, as large-scale disinformaon campaigns do not constute acts of
war that would jusfy convenonal military responses. The unique characteriscs of cyber conicts, including the vulnerabil-
ity of crical infrastructure, the ease of oensive operaons, and the challenges of aribuon and deterrence, render tradi-
onal military strategies insucient. These factors have led some scholars to characterise cyber tools as "weapons of the
weak," empowering states with fewer convenonal military and economic resources.
With the ongoing progress of arcial intelligence technology, the possibility for increasingly intricate and compelling disin-
formaon campaigns expands. The dynamic nature of this environment requires connuous study and the adjustment of
strategies to protect the integrity of democrac procedures and public communicaon from these developing pressures.
Although direct aacks on vong systems may result in me-limited interrupons, their inuence is insignicant when com-
pared to the enduring consequences of disinformaon eorts. Through the exploitaon of contenous maers and the fab-
ricaon of misleading stories, these malevolent acvies gradually diminish condence in the media and government, so
weakening democrac instuons and exerng a more long-lasng impact on elecon results compared to deliberate hack-
ing eorts.
Electoral misinformaon taccs have repercussions that go beyond the mere fabricaon of misleading narraves; they give
rise to wider social discontent with extensive eects. Arcial intelligence (AI) powered operaons exacerbate social ten-
sions, leading to divisions that may be challenging to remediate. It is imperave for policymakers, technology corporaons,
and educators to establish a purposeful posion in order to safeguard against these methods of aack and enhance social
resilience to endure the increasing menace of digital propaganda. Furthermore, as Virginia Eubanks points out in her 2018
book "Automang Inequality," decision-making systems powered by arcial intelligence in public services have the poten-
al to strengthen and worsen pre-exisng social disparies. It is her contenon that these mechanisms frequently exhibit
prejudice against marginalised groups, therefore potenally compromising the democrac ideals of equality and jusce.
The implicaons of these taccs are parcularly concerning during electoral periods. A recent survey revealed that 77% of
the French populaon believes that fake news signicantly aects the democrac funconing of society, while 72% express
concern about the inuence of disinformaon on vong. Furthermore, more than half of the respondents (55%) fear that
disinformaon campaigns could challenge the legimacy of the results of the European elecons. In 2021, 51% of internet
users in France reported encountering news they considered false or unreliable on social media or news sites in the previous
three months. The proliferaon of false news has been alarming; between 2016 and 2017, the volume of tweets referencing
false news doubled in France and quintupled globally. By 2019, there were approximately 45.5 million tweets worldwide
discussing false news, with 1.7 million originang solely in France. In 2023, the monitoring organisaon NewsGuard iden-
ed nearly 800 websites dedicated exclusively to publishing false stories.
Humanitarian News
W A B I P N E W S L E T T E R P A G E 12
Generave AI technologies, capable of producing text, images, and videos, have become widely accessible and are increas-
ingly used to carry out disinformaon campaigns. Large language models (LLMs), such as GPT-3, and generave adversarial
networks (GANs) enable the creaon of highly realisc fake news and manipulated media. These tools facilitate the dissemi-
naon of disinformaon through social media plaorms, including Facebook, X (formerly Twier), and Instagram, where
they can easily inuence public opinion and electoral outcomes. The French Ministry of the Interior has idened foreign
interference operaons as a growing threat to democrac stability, underscoring the urgency of addressing this form of
"informaonal insecurity." The challenge lies not only in the creaon of disinformaon but also in developing eecve re-
sponses that leverage both AI and human analysis to detect, idenfy, and counter these threats.
A mulfaceted strategy is essenal to safeguard the authencity of democrac processes in the face of the growing threat of
AI-driven disinformaon. This requires strategies that encompass collaboraon across various sectors to eecvely migate
these threats.
Some of the proposed strategies to combat AI-driven disinformaon include:
• Cyber Awareness Educaon from an Early Age
Incorporang cyber awareness into educaonal curricula is imperave, given the alarming increase in AI-driven intru-
sions. This proacve educaonal strategy must extend beyond basic digital literacy to encompass crical thinking skills
that empower students to queson the validity and inherent biases of digital content. By fostering an environment
where quesoning and verifying online informaon becomes roune, we can equip future generaons with the neces-
sary tools to navigate the complexies of digital content. This training will enable them to discern reliable informaon
from potenal falsehoods, thereby enhancing their resilience against disinformaon.
• Public Campaigns to Promote Crical Thinking and Media Literacy
Public awareness campaigns aimed at improving media literacy across all age groups are vital. These campaigns should
foster a comprehensive understanding of how informaon is created and disseminated online, enabling individuals to
assess the credibility of the sources and content with which they engage. By culvang an informed electorate, these
iniaves can empower cizens to resist the inuence of misleading data, which oen has roots in decepon. Such
eorts not only reinforce individual crical thinking skills but also contribute to collecve social resilience against the
omnipresent threat of disinformaon.
• Collaboraon Between Governments, Technology Companies, and Civil Society
Addressing the rampant spread of AI-driven disinformaon requires collaborave approaches that bring together gov-
ernments, technology companies, and civil society. These partnerships are essenal for developing robust technological
soluons and eecve regulatory frameworks. By promong the exchange of best pracces and advancements in AI
management, stakeholders can establish resilient systems that not only idenfy and counter disinformaon but also
uphold the principles of free expression and the disseminaon of authenc informaon. Such collaboraon is crucial for
creang a united front against the mulfaceted challenges posed by disinformaon.
The complex issue of AI-driven disinformaon represents a signicant threat to the fundamental principles of democrac
sociees. The dual challenges of cyberaacks and the widespread disseminaon of false informaon underscore the necessi-
ty of educaon as a dynamic defence strategy. By empowering individuals with the scepcism and knowledge required to
navigate digital content intelligently, we can strengthen our democrac values. Policymakers, educators, and technology
experts must priorise investments in praccal social soluons that address current threats while ancipang future risks
associated with AI. Implemenng regulaons to hold plaorms accountable for the content they generate is essenal in this
endeavour. By promong awareness and fostering collaboraon, we can forfy our democracies against the profound im-
pacts of AI-facilitated disinformaon, ensuring that the integrity of democrac processes is preserved for future generaons.
A disnguished invesgave journalist such as Carole Cadwalladr, in her work on the Cambridge Analyca scandal, highlight-
ed the potenal of AI-driven microtargeng to manipulate democrac processes, prompng global debate and regulatory
acons. Meanwhile, Jamie Susskind, in "Future Polics" (2018), explores how digital technologies are reshaping power rela-
onships in society and warns of their potenal to fundamentally alter our understanding and pracce of democracy, which
could lead to new forms of digital authoritarianism if not managed properly.
Humanitarian News
W A B I P N E W S L E T T E R P A G E 13
Advancements in arcial intelligence have revoluonised the ability to map and assess channels and messages of decep-
on. Natural language processing (NLP) allows for the examinaon of communicaon exchanged on social media plaorms,
while machine learning methods, such as decision trees and long short-term memory (LSTM) networks, improve the iden-
caon of altered photos and texts. Arcial intelligence (AI) systems can undergo training to detect recurring paerns of
false informaon on social media plaorms and provide alerts on potenally detrimental material. Through the examinaon
of data obtained from plaorms like Twier and Telegram, these systems have the capability to idenfy certain stylisc fea-
tures commonly associated with false informaon. Consequently, they aid human operators in comprehending the charac-
teriscs of the content that is being agged. Nevertheless, the ecacy of arcial intelligence (AI) in this eld relies on the
calibre of the training data and the cooperaon between AI tools and human prociency.
Undoubtedly, in order to avoid these possible negave consequences, this emerging phenomena requires detailed examina-
on of laws and ethical principles. The updated EU AI Act seeks to govern deepfake operaons by mandang that consumers
of AI-generated material reveal any alteraons made. Although regulaon plays a crucial role in addressing disinformaon, it
is unable to completely reduce the dangers presented by enemies using comparable AI capabilies. Therefore, the bale
against disinformaon must employ the same AI-powered technologies that bad actors use, based on strong ethical princi-
ples that strengthen naonal sovereignty. The OECD report released on 4 March 2024 emphasises the need of a synchro-
nised naonal approach to monitor disinformaon on social media and ensure that cizens are informed about connuous
aempts to intervene. Furthermore, this approach should priorise the improvement of crical thinking abilies within the
populace, namely through educaonal programs designed to increase knowledge of manipulaon techniques.
Designing cung-edge technologies to counter disinformaon must adhere to legal and ethical norms, disnguishing demo-
crac sociees from non-democrac civilisaons. Ensuring a climate of trust among stakeholders, such as researchers, indus-
try leaders, and legislators, is essenal for promong collaboraon across dierent sectors. A collaborave ecosystem
should be established in order to develop ecient strategies to combat disinformaon.
The inuence of generave AI on the future of democracies has been the subject of extensive research and reecon within
the academic sphere. Various social sciensts have oered diverse perspecves on the present and future of democracies
under the inuence of arcial intelligence (AI).
Just to menon some few examples Shoshana Zubo, in her seminal work "The Age of Surveillance Capitalism" (2019), ar-
gues that AI-driven data collecon and analysis by tech giants pose a signicant threat to democracy. She contends that
these companies' ability to predict and inuence human behaviour undermines individual autonomy and democrac deci-
sion-making, while warning that this "surveillance capitalism" could lead to a new form of social order that is incompable
with democrac governance. Meanwhile, Cass Sunstein, in "Republic.com 2.0" (2007), explores the concepts of "echo cham-
bers" and "lter bubbles," predicng that these technologies could lead to greater polarizaon and fragmentaon of the
public sphere, potenally weakening democrac discourse. Digital plaorms, facilitated by AI algorithms, consolidate echo
chambers that reinforce cognive biases and limit exposure to diverse perspecves. This not only intensies polarizaon but
also complicates cizens' ability to discern between truthful and false informaon.AI-driven disinformaon contributes to a
growing ideological divide, both polically and socially. This generates hosle environments where construcve dialogue is
hindered, impeding the necessary collaboraon to address global challenges such as climate change and economic inequali-
ty.
In "Network Propaganda" (2018), Yochai Benkler examines how digital technologies, including AI, aect the disseminaon of
disinformaon and emphasises that the problem is not solely technological but also rooted in polical and media structures.
He proposes strengthening tradional journalisc instuons and media literacy to help migate the negave impacts of AI
on democrac discourse.Finally, Evgeny Morozov, in works such as "To Save Everything, Click Here" (2013), criques the no-
on that technology can solve complex social and polical problems. He warns against "technological soluonism" and ar-
gues that excessive reliance on AI and other technologies could weaken democrac instuons and civic engagement.
This concise reference to academic stances indicates that the majority of scholars concur that, although AI poses substanal
obstacles to democrac processes, its inuence is not predesned. The importance of strong regulaon, improved digital
literacy, and a rethinking of democrac instuons is underscored to guarantee that AI funcons to augment rather than
weaken democrac principles. Furthermore, many emphasise the need of preserving human agency and supervision in AI-
powered systems, especially those that impact public decision-making procedures. The consensus among these intellectuals
Humanitarian News
W A B I P N E W S L E T T E R P A G E 14
is that the fate of democracy in a world inuenced by arcial intelligence will mostly hinge on the decisions made by socie-
es regarding the development, implementaon, and regulaon of these technologies. They advocate for connuous public
discussion and involvement in these maers to mould a future in which AI reinforces rather than replaces democrac ideals.
The responsibility lies with us to promote legislaon that converts this powerful instrument into a mechanism for enhancing
the quality of life for all individuals, or into another unfortunate failure that exacerbates disparies and gives rise to an au-
thoritarian totalitarianism camouaged as technological advancement.
References:
1. Benkler Y et al. (2018). Network propaganda: Manipulaon, disinformaon, and radicalizaon in American polics. Oxford University
Press.
2. Domme KE. Froners in Polical Science. 2021: 3, 634074.
3. Eubanks V. (2018). Automang inequality: How high-tech tools prole, police, and punish the poor. St. Marn's Press.
4. Feldstein S. (2021). Arcial intelligence and democrac values: What's at stake? Carnegie Endowment for Internaonal Peace.
5. Kaye D. (2020). Arcial intelligence and its impact on freedom of expression. United Naons Human Rights Council.
6. Kreiss D et al. Polical Communicaon. 2020: 37(5), 651-660.
7. Morozov E. (2013). To save everything, click here: The folly of technological soluonism. PublicAairs.
8. Smith J. Journal of Polical Science. 2023: 45(2), 123-145.
9. Sunstein CR. (2007). Republic.com 2.0. Princeton University Press.
10. Susskind J. (2018). Future polics: Living together in a world transformed by tech. Oxford University Press.
11. West DM. (2020). Arcial intelligence and the future of democracy. Brookings Instuon.
*The views expressed in this arcle are those of the author (Silvia Quadrelli) and do not necessarily reect the ocial posi-
ons of the Execuve Board or Internaonal Board of Regents of the WABIP.
Best Image Contest 2024 (3 of 3)
Category: Central Airway Diseases
Descripon: Endobronchial tumor lesion obstrucng 80% of the lumen of the le
main bronchus (metastac myxoid epitheloid leiomyosarcoma).
Submier(s): Claudia Liliana Moreno Diaz
Best Image Contest
P A G E 15
This image is 1 of 3 selected among 100+ submissions to our Best Image Contest held in late 2023. Our next
Image Contest will open later this year. We look forward to receiving your image submissions.
P A G E 16
WABIP News
WCBIP 2024 CONGRESS
Preparaons for the upcoming World Congress
on Bronchology and Intervenonal Pulmonology
(WCBIP) are well underway. The WABIP Execu-
ve Board and Board of Regents meengs are
scheduled for October 23, 2024, in Bali. Pulmo-
nology experts from around the world have been
invited to serve as speakers and chairs for vari-
ous sessions. Conrmaons have been received,
and a comprehensive list is being compiled to
ensure a diverse and knowledgeable panel. The
abstract submission process has concluded. Ac-
cepted abstracts have been compiled, and issues
such as duplicate submissions have been re-
solved. The nal list will be published for regis-
tered parcipants. The WABIP Media Commiee
is promong the congress on social plaorms.
Visit the Bali Indonedia WCBIP website at
hps://www.wcbip2024.com
VIDEO FESTIVAL
The WABIP has nalized the results for its Video Fesval, recognizing top contribuons in various
categories. The "Best Overall" video was awarded to "Bronchoscopic Recanalizaon of Complex
Complete Tracheal Stenosis with Montgomery T-Tube Inseron and Follow-Up – A Case Study."
This video will be shown at the Bali Congress. Other categories include "Best Scienc Content,"
"Best Innovaon," and "Best Imaging," with winners being noed and cercates being prepared
for presentaon at the WCBIP 2024 in Bali.
WABIP INTERVENTIONAL PULMONOLOGY INSTITUTE (IPI)
WABIP is delighted to commence the acceptance process for its IPI fellowship program, with ten
(10) candidates conrming their parcipaon across various quarters of 2025. Addionally, inquir-
ies regarding parcipaon in the fellowship program for 2026 have been received from prospec-
ve candidates.
WABIP PAPER SUBMISSIONS
WABIP has published guidelines on airway stenng for malignant central airway obstrucon
(MCAO) for publicaon in the journal "Respirology," aimed at improving understanding and man-
agement pracces in this area. Addionally, a white paper on radiaon safety in bronchoscopy is
available in its nal version as an open-access arcle in "Respiraon." It aims to educate pul-
monologists on minimizing radiaon exposure during procedures and has received endorsements
from the European Associaon for Bronchology and Intervenonal Pulmonology (EABIP), with
pending endorsements from the American Associaon for Bronchology and Intervenonal Pulmo-
nology (AABIP). The topic will be addressed at a symposium during the WCBIP meeng in Bali.
Arcial Intelligence is Not Yet More Intelligent than Organic Intelligence!
Arcial Intelligence (AI) was coined in the 1950s and has been a slowly evolving eld since then. There is increasing interest in ulizing AI in medi-
cal applicaons. As in other elds, a me of great change may be approaching. Within the eld of arcial intelligence, there are mulple subsets,
such as machine learning, which is composed of the study of neural networks and further subdivided into deep learning, which ulmately branches
into generave arcial intelligence. Each branch of the eld of AI subdivides further as dierent techniques are applied.
Increasingly, there has been interest in applying dierent types of arcial intelligence approaches to specic problems in medicine. This free up
clinicians me from repeve and mundane aspects that can be automated. The detecon of lung nodules on chest imaging studies can be an ex-
ample of computer vision, where a model is trained on ways to idenfy lung nodules. This was rst performed with image idencaon via a con-
voluonal neural network in the ImageNet Trial. Applying these techniques to chest radiology is sll nascent, but some approaches may have clini-
cal applicaons sooner rather than later. There are already commercially available technologies, but it is important to understand the current tech-
nology and limitaons to determine how best to implement the technology.
Currently, depending on the technology/technique, ulizaon of AI in chest radiology does not have a perfect sensivity or specicity for idenfy-
ing benign vs malignant lung nodules. However, there has been steady progress. There are two signicant challenges: the rst is idenfying a lung
nodule, and the second is risk strafying the lung nodule to malignant vs non-malignant. Some studies, such as ANODE09 and LUNA16, focused on
idenfying lung nodules. Mulple compeons have ulized datasets of varying sizes to train AI models and compare them to experienced Radiol-
ogists. Although the inial trials did not demonstrate an advantage with AI, subsequent trials with the renement of algorithms and AI techniques
have resulted in situaons where an AI model can perform similarly to a well-trained Radiologist. We have gone from receiver operang character-
isc curves of 0.5 up to 0.9 in a short period of me.
Editor-in-Chief: Dr. Kazuhiro Yasufuku
Research
Primary Business Address:
Kazuhiro Yasufuku, Editor-in-Chief WABIP
Newsleer
c/o Judy McConnell
200 Elizabeth St, 9N-957
Toronto, ON M5G 2C4 Canada
E-mail: newsleer@wabip.com
P A G E 17
Associate editor:
Dr. Ali Musani
Associate editor:
Dr. Sepmiu Murgu
Ali I. Musani MD, FCCP
Professor of Medicine and Surgery,
University of Colorado School of
Medicine, Denver
Niral M. Patel
Assistant Professor of Medicine,
University of California San Diego
These are promising results from these small studies, and there are some products commercially available that ulize some AI techniques. It
may sll take me for an AI model to work independently of a Radiologist, but with the current technology available, there is a role for assis-
ve technology and workow improvement. Certain AI models are able to easily idenfy lung nodules/abnormalies and then pass the task
of interpretaon to the radiologist. Some commercially available products can remove bone and/or blood vessels as part of imaging post-
processing, which allows for idencaon of the nodules present. Other models have the opposite approach with nodules idened and the
AI model idenfying concerning characteriscs. With either approach or a concurrent approach, workow improvements in Radiology reads
with improved standardizaon can be made. If the me needed to review, process, and interpret Radiology studies is reduced, this would
ulmately allow for higher throughput and reducon in cost, all while further standardizing and opmizing the output. There are also studies
that have demonstrated improved results with Radiologists working with AI tools. In this case, the sum is truly greater than the individual
parts.
The use of AI in Radiology oers a glimpse into what a future of AI tools ulized in medicine could look like. There is, of course, skepcism
and concern over how AI will change medical pracce, but as with the industrial revoluon, these new tools will allow us to be more produc-
ve. They will undoubtedly have unexpected consequences, but the technology is likely here to stay. The more we understand and ulize
these new technologies as they connue the improve, the beer we will be able to help the paents in front of us.
References:
1. Kaul V et al. Gastrointest Endosc. 2020;92(4):807-812.
2. Liu JA et al. American Journal of Roentgenology. 2022;219(5):703-712.
3. Schreuder A et al. Transl Lung Cancer Res. 2021;10(5):2378-2388.
Research
P A G E 18
P A G E
19
WABIP ACADEMY- WEBCASTS
The WABIP has started a new educaon project recently: THE WABIP ACADEMY. The WABIP Academy will pro-
vide free online webcasts with new and hot topics that will interest pulmonologists and intervenonalists.
Current webcast topic: Tissue acquision for biomarker directed therapy of NSCLC
You can reach these webcasts by using this link: hp://www.wabipacademy.com/webcast/
www.bronchology.com Home of the Journal of Bronchology
www.bronchoscopy.org Internaonal educaonal website for
bronchoscopy training with u-tube and
facebook interfaces, numerous teachiing
videos, and step by step tesng and assess
ment tools
www.aabronchology.org American Associaon for Bronchology and I
ntervenonal Pulmonology (AABIP)
www.eabip.org European Associaon for Bronchology and
Intervenonal Pulmonology
W A B I P N E W S L E T T E R
Links
www.chestnet.org Intervenonal Chest/Diagnosc Procedures (IC/DP)
NetWork
www.thoracic.org American Thoracic Society
www.ctsnet.org The leading online resource of educaonal and
scienc research informaon for cardiothoracic
surgeons.
www.jrs.or.jp The Japanese Respirology Society
sites.google.com/site/asendoscopiarespiratoria/
Asociación Sudamericana de Endoscopía Respiratoria
P A G E 19
