image-150-64c8cd514b55a.webp

image-151-64c8cd518a917.webp

group-1000002623-64c8cd4f8d576.webp

image-146-64c8cd50567b2.webp

image-147-64c8cd501e463.webp

image-148-64c8cd50b2c5c.webp

image-149-64c8cd50d627e.webp

BBC-Logo.webp

The-Independent-Logo.webp

Featured in

AI biomarkers, built for real-world impact and clinical trials. Meet us at ASCO 2026 | Booth #IH12.

Qure.ai Recognized Among TIME’s 100 Most Influential Companies of 2025

Qure.ai partners with Microsoft to increase access to its lung cancer detection and management suite in the U.S.

qTrack

The World Health Organization’s&nbsp;<a href="https://www.who.int/europe/teams/ncd-management/noncommunicable-diseases-(ncds)-management">cancer control framework</a>&nbsp;emphasizes early detection, risk-based screening of defined high-risk cancer populations, and&nbsp;equitable&nbsp;access to diagnostic pathways and&nbsp;timely&nbsp;treatment, particularly in low- and middle-income countries (LMICs) where cancer mortality&nbsp;remains&nbsp;disproportionately high. For lung cancer, this includes populations at elevated risk due to age, environmental exposures, occupational hazards, and comorbidities many of whom fall outside traditional smoking-based screening criteria.

In 2022, countries including India,&nbsp;Turkey, Mexico, Indonesia, and Egypt collectively reported more than 173,000 new lung cancer cases and approximately 153,700 lung cancer-related deaths, reflecting a striking mortality rate of&nbsp;nearly 89%. This imbalance underscores the urgent need for earlier and more targeted approaches to lung cancer detection, particularly those that can be deployed at scale in resource-constrained settings.

While low-dose CT (LDCT) is globally recognized as the gold standard for lung cancer screening, its large-scale adoption&nbsp;remains&nbsp;limited by cost, infrastructure requirements, radiation exposure concerns, and constrained clinical workforce capacity. As a result, chest radiographs (CXRs) continue to serve as the most widely used first-line imaging modality in routine clinical practice across LMICs. National screening and early detection programs therefore require scalable, cost-effective solutions that can&nbsp;optimize&nbsp;existing imaging pathways without placing&nbsp;additional&nbsp;strain on already burdened health systems.

Within this context, the&nbsp;<a href="https://clinicaltrials.gov/study/NCT05817110?term=Create&amp;cond=lung%20cancer&amp;rank=1">CREATE study</a>&nbsp;(NCT05817110) provides important real-world evidence on the role of AI-enabled analysis of chest radiographs for lung cancer risk stratification. Conducted across both resource-limited and resource-adequate settings, CREATE evaluated&nbsp;Qure.ai’s&nbsp;qXR-LNMS (Lung Nodule Malignancy Score) an AI-derived imaging biomarker designed to predict the likelihood of malignancy in pulmonary nodules detected on CXRs against radiologist assessment on LDCT.

The study met its primary endpoint, achieving a positive predictive value of 54.2% and a negative predictive value of 93.4% against predefined thresholds of 20% and 70%, respectively. These results&nbsp;demonstrate&nbsp;the utility of&nbsp;qXR-LNMS in differentiating benign and malignant incidental pulmonary nodules (IPNs) on chest radiographs across diverse healthcare settings.

Notably, among incidentally diagnosed lung cancer cases,&nbsp;nearly 72%&nbsp;were never-smokers&nbsp;and 39% were under the age of 55 populations that are typically excluded from conventional lung cancer screening programs. By applying AI to routine chest X-rays and longitudinally following patients, CREATE&nbsp;demonstrates&nbsp;a pragmatic approach to&nbsp;identifying&nbsp;higher-risk individuals earlier and directing them to LDCT only when clinically&nbsp;warranted. This enables more focused use of specialist resources, supports risk-based screening aligned with WHO guidance, and strengthens national lung cancer control programs by&nbsp;facilitating&nbsp;earlier diagnosis and stage shift. Phase II of the CREATE study is currently ongoing, with interim results presented at&nbsp;<a href="https://www.annalsofoncology.org/article/S0923-7534(25)04511-9/fulltext">ESMO 2025</a>.

large_Qure.ai Goa Lung Cance Web Hero-02.jpg

medium_Qure.ai Goa Lung Cance Web Hero-02.jpg

small_Qure.ai Goa Lung Cance Web Hero-02.jpg

thumbnail_Qure.ai Goa Lung Cance Web Hero-02.jpg

Qure.ai Goa Lung Cance Web Hero-02.jpg

The Invisible Majority: Using AI to Detect Lung Cancer Early in Non-Smokers

Lung cancer remains the leading cause of cancer -related mortality, largely driven by late stage diagnosis when curative treatment options are limited. Although chest X-ray (CXR) is the most widely used imaging modality globally, pulmonary nodules on CXR are frequently missed due to their subtle presentation and high radiology workloads. This case series highlights the role of AI-enabled CXR analysis using qXR-LNMS in enabling opportunistic early detection.

All consecutive CXRs at participating centers were analyzed using qXR-LNMS, with high-risk cases flagged for clinical review and further investigation. Five asymptomatic patients, who underwent imaging for non-respiratory indications, were identified through this pathway. None had prior suspicion or symptoms of lung cancer. AI flagged high-risk nodules, prompting follow-up with CT imaging and biopsy, which confirmed early-stage lung cancer in all cases. Four patients underwent curative surgery with favorable outcomes, while one received chemotherapy.&nbsp;

Additional cases from Hacettepe University, Turkey, demonstrated similar outcomes, where AI identified unsuspected nodules leading to early diagnosis of Lung Cancer and successful surgical intervention. These findings are supported by data from the CREATE study, which validates the ability of AI to stratify malignancy risk in incidental nodules.

This study demonstrates that AI can act as an opportunistic  screening layer, identifying high-risk nodules in asymptomatic individuals outside traditional screening criteria. Integrating AI into routine workflows that offers a scalable, cost-effective approach to expand early lung cancer detection, particularly in settings without formal screening programs.

large_thumbnail 6 (1).png

medium_thumbnail 6 (1).png

small_thumbnail 6 (1).png

thumbnail_thumbnail 6 (1).png

thumbnail 6 (1).png

A Case Series From a Multicentric Study: Can Artificial Intelligence (AI)-Enabled Chest X-Ray Assist in the Incidental Detection of Early-Stage Lung Cancers?

Incidental pulmonary nodules (IPNs) are frequently detected on chest X-rays (CXRs) performed for routine clinical indications and may represent early-stage lung cancer. While low-dose CT (LDCT) remains limited in resource -constrained settings such as Malaysia. In this context, AI-assisted CXR interpretation offers a scalable approach for opportunistic detection of lung nodules, though local validation is essential for clinical adoption.

This prospective, multicenter cross-sectional study aims to evaluate the diagnostic performance of an AI-based solution, qXR, across hospitals in the Klang Valley region. Approximately 1,000-2,000 consecutive adult CXRs will be analyzed for detection of nodules ≥6 mm. Ground truth will be established through independent review by two thoracic radiologists, with CT correlation where available.&nbsp;

Primary outcomes include sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Secondary analyses will assess performance across nodule size and location, as well as the impact on radiologist reading time. Statistical methods include ROC curve analysis and paired comparison using McNemar testing.

Although results are pending, prior evidence suggests AI sensitivity of ~83-90% for nodule detection, with a 9-10% improvement when used as a second reader. This study is expected to provide critical real-world validation in a multi-ethnic population and demonstrate how AI-enabled CXRs can serve as a scalable alternative to LDCT, supporting earlier lung cancer detection in underserved regions globally.&nbsp;

large_thumbnail 4 (1).png

medium_thumbnail 4 (1).png

small_thumbnail 4 (1).png

thumbnail_thumbnail 4 (1).png

thumbnail 4 (1).png

Performance validation of an artificial intelligence-assisted chest radiograph algorithm for incidental pulmonary nodule detection in Malaysian healthcare facilities: a multicentre cross-sectional study protocol

Early detection of malignant pulmonary nodules remains a significant barrier in lung cancer care, especially in regions where access to low-dose CT (LDCT) is limited. While chest X-rays are widely used and often serve as the first point of imaging, their utility to accurately stratify the malignancy risk of incidental pulmonary nodules (IPNs) is inherently limited.

To address this gap, Qure.ai developed the qXR-LNMS (Lung Nodule Malignancy Risk), an AI-based tool designed to assess the likelihood of malignancy in lung nodules detected on chest X-rays and support clinical decision-making.

The performance of this tool was evaluated in CREATE study, a prospective, real-world, multi-country observational study conducted across Turkey, Mexico, Egypt, Indonesia, and India. The study included 712 participants aged 35 years and above, each presenting incidental pulmonary nodules (8-30 mm) detected on chest X-rays .

 Participants were initially identified using an FDA-cleared AI detection algorithm (qXR-LNMS) and subsequently AI results were reviewed by radiologists. The cohort consisted of both high-risk and low-risk individuals, enabling robust evaluation across patient segments. The primary objective of this study was to assess the predictive performance of qXR-LNMS against radiologist-confirmed LDCT findings, with additional comparison to established tools such as the Mayo Clinic risk model.&nbsp; &nbsp;

The results demonstrated strong predictive performance:

Positive Predictive Value (PPV): 54.2%&nbsp; 

Importantly, the high NPV indicates a strong ability to rule out malignancy, making it particularly valuable as a triage tool. The AI model also showed 70.6% agreement with the Mayo Clinic risk model, with consistent performance across demographic and clinical subgroups. 

Overall, these findings highlight the potential of qXR-LNMS to enable earlier and more accurate risk stratification of lung nodules directly from chest X-rays. By improving confidence in identifying high-risk cases while safely ruling out low-risk nodules, the tool can support more efficient referral pathways and optimize the use of advanced imaging.

 In essence, this approach transforms a widely available imaging modality into a more powerful screening and decision-support tool particularly impactful in both resource constrained and high-volume healthcare settings.&nbsp;

large_thumbnail 5 (1).png

medium_thumbnail 5 (1).png

small_thumbnail 5 (1).png

thumbnail_thumbnail 5 (1).png

thumbnail 5 (1).png

Evaluation of an Artificial Intelligence Defined Lung Nodule Malignancy Score (qXR-LNMS) in Incidental Pulmonary Nodules: The CREATE Study

Availability of ultraportable chest x-ray (CXR) and advancements in artificial intelligence (AI)-enabled CXR interpretation are promising developments in tuberculosis (TB) active case finding (ACF) but costing and cost-effectiveness analyses are limited. We provide implementation cost and cost-effectiveness estimates of different screening algorithms using symptoms, CXR and AI in Nigeria. People 15 years and older were screened for TB symptoms and offered a CXR with AI-enabled interpretation using qXR v3 (Qure.ai) at lung health camps. Sputum samples were tested on Xpert MTB/RIF for individuals reporting symptoms or with qXR abnormality scores ≥0.30.

We conducted a retrospective costing using a combination of top-down and bottom-up approaches while utilizing itemized expense data from a health system perspective. We estimated costs in five screening scenarios: abnormality score ≥0.30 and ≥0.50; cough ≥ 2 weeks; any symptom; abnormality score ≥0.30 or any symptom. We calculated total implementation costs, cost per bacteriologically-confirmed case detected, and assessed cost-effectiveness using incremental cost-effectiveness ratio (ICER) as additional cost per additional case.

Overall, 3205 people with presumptive TB were identified, 1021 were tested, and 85 people with bacteriologically-confirmed TB were detected. Abnormality ≥ 0.30 or any symptom (US$65704) had the highest costs while cough ≥ 2 weeks was the lowest (US$40740). The cost per case was US$1198 for cough ≥ 2 weeks, and lowest for any symptom (US$635). Compared to baseline strategy of cough ≥ 2 weeks, the ICER for any symptom was US$191 per additional case detected and US$ 2096 for Abnormality ≥0.30 OR any symptom algorithm. 

Using CXR and AI had lower cost per case detected than any symptom screening criteria when asymptomatic TB was higher than 30% of all bacteriologically-confirmed TB detected. Compared to traditional symptom screening, using CXR and AI in combination with symptoms detects more cases at lower cost per case detected and is cost-effective. TB programs should explore adoption of CXR and AI for screening in ACF.

large_WhatsApp Image 2025-10-06 at 11.03.33 AM.webp

medium_WhatsApp Image 2025-10-06 at 11.03.33 AM.webp

small_WhatsApp Image 2025-10-06 at 11.03.33 AM.webp

thumbnail_WhatsApp Image 2025-10-06 at 11.03.33 AM.webp

WhatsApp Image 2025-10-06 at 11.03.33 AM.webp

Implementation costs and cost-effectiveness of ultraportable chest X-ray with artificial intelligence in active case finding for tuberculosis in Nigeria

Pulmonary tuberculosis (PTB) remains a global public health challenge, with 10.8 million new cases reported in 2023. Early diagnosis is crucial for controlling its spread, yet traditional sputum-based tests face limitations in turnaround time and resource availability. Chest X-ray (CXR) is a cost-effective diagnostic tool, but its use in high-tuberculosis (TB) burden regions is restricted by a shortage of radiologists. Artificial intelligence (AI)-based computer-aided detection (CAD) systems, leveraging deep learning, offer a promising solution for automated PTB detection. However, variability in diagnostic performance across AI tools and the need for scenario-specific threshold adjustments remain challenges that need to be addressed. Our meta-analysis evaluated the diagnostic accuracy of five AI-based PTB detection products, aiming to provide insights for advancing AI applications in TB screening and diagnosis.

The PubMed, Embase, Web of Science, and Cochrane Library databases were searched for literature related to CXR diagnosis of TB based on AI technology published from the establishment day of the database to December 19, 2024. The keywords were “artificial intelligence”, “tuberculosis”, “chest X-ray”, and “diagnosis”. The literature search, screening, data extraction, quality evaluation, and bias risk assessment were conducted independently by two researchers, and Stata 17.0 software (StataCorp) was used to process and analyze the data.

A total of 5,651 references were retrieved, and 21 references were finally selected according to the inclusion and exclusion criteria. The meta-analysis included five software solutions for CXR analysis: JF CXR-1 (JF Healthcare, Nanchang, China), qXR (Qure.ai, Mumbai, India), Lunit INSIGHT CXR (Lunit, Seoul, South Korea), CAD4TB (Delft Imaging, ‘s-Hertogenbosch, Netherlands), and InferRead DR Chest (Infervision, Beijing, China). Their sensitivity and specificity were as follows: JF CXR-1, 86.0% and 80.0%; qXR, 90.0% and 64.0%; Lunit INSIGHT CXR, 90.0% and 63.0%; CAD4TB, 91.0% and 60.0%; InferRead DR Chest, 89.0% and 59.0%.

AI software has demonstrated excellent diagnostic performance in assisting the CXR diagnosis of TB and can help clinicians to make rapid and accurate decisions in screening and treating patients with TB.

large_WhatsApp Image 2025-10-06 at 4.07.16 PM.webp

medium_WhatsApp Image 2025-10-06 at 4.07.16 PM.webp

small_WhatsApp Image 2025-10-06 at 4.07.16 PM.webp

thumbnail_WhatsApp Image 2025-10-06 at 4.07.16 PM.webp

WhatsApp Image 2025-10-06 at 4.07.16 PM.webp

A systematic review and meta-analysis of artificial intelligence software for tuberculosis diagnosis using chest X-ray imaging

The Xpert MTB/RIF Ultra assay (Ultra) is a second-generation molecular diagnostic test for tuberculosis (TB). The “Trace Call” result was added as a semi-quantitative category for extremely low bacillary loads. The prevalence and interpretation of Trace Call results remains insufficiently elucidated in the context of community-based active case finding (ACF).

We organized 56 days of mobile chest X-ray (CXR) screening events in Ho Chi Minh City, Viet Nam, between October 2020 and March 2021. Participants were screened verbally and by CXR and tested by Ultra, if eligible. Persons with a Trace Call were re-tested on Ultra per national guidelines. qXRv3 computer-aided detection software was used for post hoc quality control of CXR interpretation. We calculated descriptive statistics and fitted mixed-effect multivariate regression models to identify factors associated with Trace Call results and confirmatory diagnosis.

A total of 16,698 people were screened by CXR to detect 185 Ultra-positive participants, including 142 persons with a confirmed TB diagnosis. Among Ultra-positive participants, 38.4% (71/185) had Trace Call results. Of these, 85.9% (61/71) were re-tested, and 45.9% (28/61) were bacteriologically-confirmed, comprising 19.7% (28/142) of the final number of confirmed diagnoses. Having a low qXR abnormality score (&lt;0.5) (aOR = 4.97; 95%CI: [1.88, 13.14];&nbsp;p&nbsp;= 0.001) and a history of TB within 5 recent years (aOR = 3.53; 95%CI: [1.69, 7.35];&nbsp;p&nbsp;= 0.001) were associated with an initial Trace Call.

The Trace Call can improve ACF detection, particularly in earlier stages of disease with limited pulmonary deterioration. However, participants with a history of TB had higher rates of Trace Call, reinforcing the need to interpret test results in this group with caution.

large_WhatsApp Image 2025-10-06 at 4.07.15 PM-2.webp

medium_WhatsApp Image 2025-10-06 at 4.07.15 PM-2.webp

small_WhatsApp Image 2025-10-06 at 4.07.15 PM-2.webp

thumbnail_WhatsApp Image 2025-10-06 at 4.07.15 PM-2.webp

WhatsApp Image 2025-10-06 at 4.07.15 PM-2.webp

Prevalence of Xpert MTB/RIF Ultra Trace Call Results and Associated Risk Factors During Active Tuberculosis  Case Finding in Viet Nam: A Programmatic Evaluation

Researchers investigated combining tuberculin skin testing (TST) for tuberculosis infection with an active case-finding program in peri-urban Uganda. The study used mobile chest X-rays analyzed by qXR v3, an AI-powered computer-aided detection system from Qure.ai, to generate risk scores determining who needed further testing. Participants with lower qXR scores indicating minimal active TB disease risk were offered TST to identify latent infection, while those with higher scores received sputum testing. This integrated approach aimed to identify both active TB cases and individuals who could benefit from preventive treatment.

The results revealed significant implementation challenges. Of over 33,000 eligible participants, only 12% received TST placement, and just 55% returned for results. While qXR efficiently triaged participants, practical barriers limited program reach. Financial constraints emerged as the primary obstacle, with transportation costs and time away from work cited most frequently. Each additional 10 minutes of travel time reduced TST completion likelihood. However, individuals with previous TB exposure or current symptoms showed greater engagement.

These findings underscore that while qXR enabled efficient screening triage, logistical challenges including site relocations, weather disruptions, and supply shortages compounded difficulties requiring return visits. Without streamlined processes addressing financial and motivational barriers, integrated screening programs may have limited impact. Future TB prevention efforts should focus on understanding local barriers and developing solutions such as point-of-care alternatives to make prevention more accessible in high-burden communities.&nbsp;&nbsp;

large_Evidence Papers on Website_4.webp

medium_Evidence Papers on Website_4.webp

small_Evidence Papers on Website_4.webp

thumbnail_Evidence Papers on Website_4.webp

Evidence Papers on Website_4.webp

Investigating low uptake of tuberculin skin testing when integrated with active case-finding for TB

Tuberculosis remains a public health challenge in Nepal and ranks as the seventh leading cause of death in the country. The END Tuberculosis strategy stresses - the screening for symptoms alone may not suffice; additional screening tools such as a chest radiograph may facilitate referral for diagnosis of tuberculosis. The study aims to evaluate the diagnostic accuracy of artificial intelligence (AI) based Chest X-ray and compare it with the human reading (radiologist), using GeneXpert-MTB RIF Assay for tuberculosis case detection.

Tuberculosis-suspected patients with a history of cough were screened using chest X-rays at two study sites (Dhulikhel Hospital and Nobel Medical College). The reading of AI qXR software was compared with radiologists reading who were blinded of the results generated by the software.Results: The sensitivity of the test by qXR-based AI reading was 100%, (95% CI: 40 – 100%) and specificity 80% (95% CI: 73 – 87%), whereas the sensitivity of the test by the radiologist was 100%, (95% CI: 40 – 100%); and specificity 62% (95% CI: 53 – 70%).

Higher sensitivity and specificity were observed for both qXR-based AI and Radiographer readings for the diagnosis of TB.

large_Evidence Papers on Website-01.webp

medium_Evidence Papers on Website-01.webp

small_Evidence Papers on Website-01.webp

thumbnail_Evidence Papers on Website-01.webp

Evidence Papers on Website-01.webp

Diagnostic Accuracy of Artificial Intelligence-Based Chest X-Ray reading for screening of Tuberculosis

Qure, achieves Health Technology Assessment, in India for AI-Assisted TB Screening. Qure has been evaluated to increase TB detection, while saving costs as compared to the status quo. Qure has also achieved the best overall results, compared to all CDSCO-cleared TB AI solutions in India. 

The Health Technology Assessment (HTA) conducted by ICMR, evaluates the effectiveness of AI tools in improving Tuberculosis (TB) diagnosis care. The study focused on the diagnostic accuracy and cost-effectiveness of AI-assisted CXR interpretation solutions including qXR by&nbsp;Qure.ai. The solutions were also compared to the manual interpretation of the X-rays. 

The results showed that Qure's AI solutions increased TB detection while saving costs. qXR has demonstrated, an ICER of -9,864.77 INR. The study concluded that AI-enabled interpretation could be a potential solution to the issue of human resource constraints and reduce the delays in the diagnosis and treatment of Tuberculosis.&nbsp;

<a href="https://htain.dhr.gov.in/images/pdf/outcome_reports/77.AI_Assisted_CXR.pdf">The full HTA report</a> offers more insights on the study analysis, the challenges of TB screening programs and the role AI can play.

large_IMG_5259.jpg

medium_IMG_5259.jpg

small_IMG_5259.jpg

thumbnail_IMG_5259.jpg

IMG_5259.jpg

Qure.ai increases TB detection, while saving costs, shows evaluation in India

Pulmonary TB (PTB) predominantly affects individuals of working age. We sought to characterise the occupations of people newly diagnosed with PTB in Karachi, Pakistan, by type and physical intensity.

We did a secondary analysis of data from a study evaluating the diagnostic accuracy of artificial intelligence-based chest X-ray (CXR) analysis software, where individuals had been evaluated for active PTB using sputum cultures and had provided information on occupation. We used an accelerometer-validated US National Health and Nutrition Examination Survey-based job categorisation to assign physical activity levels to participant-reported occupations as High, Intermediate, or Low.

Among 272 participants with microbiologically confirmed PTB (women: 130/272, 48%; median age: 29 years, IQR 22–45), 78% (211/272) had smear-positive disease, and 96% (260/272) had data on occupation. Unemployment was common (women: 70/122, 57%; men: 23/138, 17%). Most women reporting an occupation were homemakers (21/52, 40%), and 54% (28/52) had an intermediate- or a high physical activity occupation. Among men reporting an occupation, 35% (40/115) were labourers, and 79% (91/115) had an intermediate- or high-physical activity occupation.

The majority of individuals with PTB were in their working age, had extensive disease, and had intermediate or high physical activity occupations, suggesting economic vulnerability due to physical impairment.

large_Evidence Papers on Website-02.webp

medium_Evidence Papers on Website-02.webp

small_Evidence Papers on Website-02.webp

thumbnail_Evidence Papers on Website-02.webp

Evidence Papers on Website-02.webp

Type and physical intensity of occupations at pulmonary TB diagnosis

The Philippines faces challenges in the screening of tuberculosis (TB), one of them being the shortage in the health workforce who are skilled and allowed to screen TB. Deep learning neural networks (DLNNs) have shown potential in the TB screening process utilizing chest radiographs (CXRs). However, local studies on AI-based TB screening are limited. This study evaluated qXR3.0 technology's diagnostic performance for TB screening in Filipino adults aged 15 and older. Specifically, we evaluated the specificity and sensitivity of qXR3.0 compared to radiologists' impressions and determined whether it meets the World Health Organization (WHO) standards.

A prospective cohort design was used to perform a study on comparing screening and diagnostic accuracies of qXR3.0 and two radiologist gradings in accordance with the Standards for Reporting Diagnostic Accuracy (STARD). Subjects from two clinics in Metro Manila which had qXR 3.0 seeking consultation at the time of study were invited to participate to have CXRs and sputum collected. Radiologists' and qXR3.0 readings and impressions were compared with respect to the reference standard Xpert MTB/RiF assay. Diagnostic accuracy measures were calculated.

With 82 participants, qXR3.0 demonstrated 100% sensitivity and 72.7% specificity with respect to the reference standard. There was a strong agreement between qXR3.0 and radiologists' readings as exhibited by the 0.7895 (between qXR 3.0 and CXRs read by at least one radiologist), 0.9362 (qXR 3.0 and CXRs read by both radiologists), and 0.9403 (qXR 3.0 and CXRs read as not suggestive of TB by at least one radiologist) concordance indices.

qXR3.0 demonstrated high sensitivity to identify presence of TB among patients, and meets the WHO standard of at least 70% specificity for detecting true TB infection. This shows an immense potential for the tool to supplement the shortage of radiologists for TB screening in the country. Future research directions may consider larger sample sizes to confirm these findings and explore the economic value of mainstream adoption of qXR 3.0 for TB screening.

large_WhatsApp Image 2025-10-06 at 4.07.15 PM.webp

medium_WhatsApp Image 2025-10-06 at 4.07.15 PM.webp

small_WhatsApp Image 2025-10-06 at 4.07.15 PM.webp

thumbnail_WhatsApp Image 2025-10-06 at 4.07.15 PM.webp

WhatsApp Image 2025-10-06 at 4.07.15 PM.webp

Diagnostic Performance of a Computer-aided System for Tuberculosis Screening in Two Philippine Cities

This retrospective pilot study compares the performance of AI-powered computer-aided detection (CAD) software (qXR by Qure.ai) with that of two experienced Ethiopian radiologists in diagnosing pulmonary tuberculosis (PTB) using chest X-rays (CXR) from Guinea-Bissau and Ethiopia. The study focused on enhancing TB diagnosis in low-resource, high TB-endemic settings by evaluating the feasibility of using mobile phone-captured analogue CXRs. Two reference standards were used: clinical diagnosis and GeneXpert (GX)-confirmed TB. Out of 498 CXRs, the software identified 81 cases, while the radiologists detected 50 and 99 cases, respectively. The CAD tool achieved an area under the curve (AUC) of 0.84 for GX-confirmed cases, with a sensitivity of 76.5% and specificity of 85.9% at a 0.5 cut-off. Radiologist A showed 64.7% sensitivity and 91.9% specificity, while Radiologist B had 76.5% sensitivity and 82.3% specificity. The agreement between the radiologists was moderate (k=0.45), as was the agreement between the software and individual radiologists (k=0.56). The findings suggest that CAD CXR performs comparably to experienced radiologists, even when applied to mobile phone images of analogue CXRs, making it a promising tool for TB diagnosis in resource-limited settings.<br>

large_Evidence Papers on Website-11.webp

medium_Evidence Papers on Website-11.webp

small_Evidence Papers on Website-11.webp

thumbnail_Evidence Papers on Website-11.webp

Evidence Papers on Website-11.webp

Mobile phone-based computer-aided chest X-ray detects TB as well as trained radiologists

This prospective cohort study in South Africa followed 483 household contacts of rifampicin-resistant TB for 4.6 years, evaluating the diagnostic and prognostic accuracy of three CAD software—qXRv3.0.0, CAD4TBv7.0, and Lunit INSIGHT v3.1.4.111. Among participants, 23 had prevalent TB (7 detected routinely), and 38 developed incident TB. qXR outperformed the other CAD solutions, achieving an AUC of 0.91 (95% CI: 0.83–0.99) for detecting prevalent TB, compared to 0.88 for Lunit INSIGHT and 0.87 for CAD4TB. Notably, over 30% of individuals with high qXR scores but negative routine sputum were later diagnosed with TB through enhanced sputum testing or follow-up, underscoring qXR’s superior ability to flag true TB cases missed by routine diagnostics. While the AUC for predicting incident TB ranged between 0.60–0.65 across all software, the study reinforces qXR’s effectiveness as the leading CAD tool for TB screening and triage, highlighting its potential to enhance early detection and intervention in high-burden settings.

large_Evidence Papers on Website-09.webp

medium_Evidence Papers on Website-09.webp

small_Evidence Papers on Website-09.webp

thumbnail_Evidence Papers on Website-09.webp

Evidence Papers on Website-09.webp

Diagnostic Accuracy of Chest X-ray Computer-Aided Detection Software for Detection of Prevalent and Incident Tuberculosis in Household Contacts 

Missed nodules in chest radiographs (CXRs) are common occurrences. We assessed the effect of artificial intelligence (AI) as a second reader on the accuracy of radiologists and non-radiology physicians in lung nodule detection and localization in CXRs.

This retrospective study using the multi-reader multi-case design included 300 CXRs acquired from 40 hospitals across the US. All CXRs had a paired follow-up image (chest CT or CXR) to augment the ground truth establishment for the presence and location of nodules on CXRs by five independent thoracic radiologists. 15 readers (nine radiologists and six non-radiology physicians) read each CXR twice in a second-reader paradigm, once without AI and then immediately with AI assistance. The primary analysis assessed the difference in area-under-the-alternative-free-response-receiver-operating-characteristic-curve (AFROC) of readers with and without AI. Case-level area-under-the-receiver-operating-characteristic-curve (AUROC), sensitivity, and specificity were assessed in secondary analyses.&nbsp;

A total of 300 CXRs (147 with nodules, 153 without nodules) from 300 patients (mean age, 64 years ± 15 [standard deviation]; 174 women) were included. The mean AFROC of readers was 0.73 without AI and 0.81 with AI (95% CI of difference, 0.05–0.10). Case-level AUROC was 0.77 without AI and 0.84 with AI (95% CI of difference, 0.04–0.09). Case-level sensitivity was 72.8% and 83.5% (95% CI of difference, 6.8–14.6) and specificity was 71.1% and 72.0% (95% CI of difference, −0.8–2.6) without and with AI, respectively.&nbsp;

Using AI, readers detected and localized more nodules without any significant difference in false positive interpretations.

large_Web Hero-02.webp

medium_Web Hero-02.webp

small_Web Hero-02.webp

thumbnail_Web Hero-02.webp

Web Hero-02.webp

AI Improves Lung Nodule Detection in Multi-Center Study

Computer-aided detection (CAD) tools for TB detection have the potential to enable screening programmes and reduce the diagnostic gap in settings where access to radiologists is limited. However, there are concerns that other common chest X-ray (CXR) abnormalities not due to TB may be missed.<br>

We assessed the performance of three commercialised CAD tools (qXR, INSIGHT CXR and DrAIDTM TB XR) to detect common non-TB abnormalities against readings with a standardised annotation guide by an expert radiologist. More than 20 well-characterised diagnoses besides TB significant in TB high-burden countries were examined.

The 517 CXRs included were deemed abnormal by the three CAD with a sensitivity of respectively 97% (95% CI 95–98), 94% (95% CI 91–95), and 87% (95% CI 84–90) for INSIGHT CXR, qXR, and DrAID. The CAD generally detected abnormalities in patients with critical diagnoses such as lung cancer or heart failure. Performance for detecting other abnormalities was variable.

This study showed that the three CAD tools identified CXRs as abnormal when diseases other than TB were present. Our findings alleviate ethical concerns of missing abnormalities other than TB when using commercially available CAD for TB screening and show their potential broader applicability.&nbsp;&nbsp;

large_Evidence Papers on Website_8.webp

medium_Evidence Papers on Website_8.webp

small_Evidence Papers on Website_8.webp

thumbnail_Evidence Papers on Website_8.webp

Evidence Papers on Website_8.webp

Detection of other pathologies when utilising computer-assisted digital solutions for TB screening

This was a multicenter retrospective study aiming to evaluate the agreement between qXR and radiologists in classifying CXRs as normal or abnormal (more from a TB angle) for scans that the rads receive from visa screening pathway. Addl, that have assess the perceived benefits and challenges of integrating AI into clinical workflows to streamline the screening process

This prospective study analyzing over 1.3 mn scans from the visa screening population over 18 months (Jan 2021 - June 2022) using qXR to detect various radiological abnormalities in CXRs, thereby enhancing both the speed and accuracy of diagnoses by segregating normals from abnormals.

Integrating AI into radiology workflows addresses issues of burnout, improves diagnostic accuracy, and enhances operational efficiency. In the UAE's 33 radiology centres, qXR streamlined the visa screening process, quickly classifying CXRs and supporting timely clinical decisions while alleviating the workload on radiologists

Survey result: (there was also a survey conducted to understand the experience of rads/PACS &amp; IT managers)The survey indicated that 82% of radiologists believe AI enhances diagnostic accuracy &amp; 88.2% noted a reduction in TAT following AI integration. Our AI demonstrated a high NPV of 99.92%, confirming its effectiveness in identifying normal CXRs with minimal false negatives.

With strict compliance to regulatory standards (since all the data collected during the course of this study was stripped of any patient identifiers in compliance with HIPAA and GDPR guidelines) ; the integration of AI in clinical workflows is rapidly expanding. This study, analyzing over 1.3 million scans from the visa screening population over 18 months (Jan 21 - June 22), highlights the robust capabilities of AI in enhancing efficiency and reducing diagnostic fatigue. By aligning AI interpretations with radiologist assessments, we can minimize the risk of incorrect diagnoses and ultimately improve patient care.

-&gt; 0.99 NPV ---indicating that our AI can very well identify normal CXRs from abnormal with minimum FN

-&gt; 88% of rads report faster turnaround times

-&gt; 82% of rads confirm improved diagnostic accuracy

large_Template 13 (1).webp

medium_Template 13 (1).webp

small_Template 13 (1).webp

thumbnail_Template 13 (1).webp

Template 13 (1).webp

Normal and Abnormal Chest X-Ray Classification at Visa Screening Centers in UAE

The study aimed to understand how informal employment and gender-related issues intersect to affect timely TB diagnosis. Conducting in-depth interviews with 80 female TB patients and 60 of their significant others, the researchers found that normalization of TB symptoms as work-related health issues, fear of job and wage loss, and non-specific symptom narratives led to delays in seeking formal medical care. The study concluded that TB elimination programs should adopt a comprehensive approach that includes workplace interventions, especially in the informal sector, to address these challenges effectively.<br>

large_Evidence-Papers-on-Website_10.webp

medium_Evidence-Papers-on-Website_10.webp

small_Evidence-Papers-on-Website_10.webp

thumbnail_Evidence-Papers-on-Website_10.webp

Evidence-Papers-on-Website_10.webp

Quantifying sputum production success during community-based screening for TB

Artificial intelligence (AI) is transforming diagnostic radiology, particularly in fracture detection, enhancing sensitivity, efficiency, and workflow in clinical settings. AI-assisted fracture detection has demonstrated up to 10.4% improvement in sensitivity and 5% in specificity, reducing radiologist errors and interpretation time while aiding clinicians across specialties. Bounding boxes on images improve explainability and reduce search bias, ensuring critical fractures are not missed. AI has the potential to broaden access to radiographic interpretation, benefiting emergency and family medicine practitioners. However, current models struggle with complex pathologies, such as fracture-dislocations, and may replicate human diagnostic errors if training data lacks CT-based validation. While AI remains a powerful augmentative tool, cautious implementation is essential to mitigate over-reliance and confirmation bias. Future advancements, driven by diverse training datasets and multi-institutional collaboration, will refine AI’s role in radiology workflow and musculoskeletal imaging.<br>

large_Evidence Papers on Website-12.webp

medium_Evidence Papers on Website-12.webp

small_Evidence Papers on Website-12.webp

thumbnail_Evidence Papers on Website-12.webp

Evidence Papers on Website-12.webp

Musculoskeletal fracture detection: artificial intelligence and machine learning-based diagnostic advantages and pitfalls (Texas Tech University Health Sciences Center School of Medicine)

Evidence (Publications & White Papers)