Assessing Diagnostic Accuracy and Viability of AI-Assisted Tuberculosis Detection in Northern Indian Healthcare Facilities: A Multicenter Study

Assessing Diagnostic Accuracy and Viability of AI-Assisted Tuberculosis Detection in Northern Indian Healthcare Facilities: A Multicenter Study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Assessing Diagnostic Accuracy and Viability of AI-Assisted Tuberculosis Detection in Northern Indian Healthcare Facilities: A Multicenter Study Alok Nath, Zia Hashim, Saumya Shukla, Prasanth Areekkara Poduvattil, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4377653/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 Sep, 2024 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Tuberculosis (TB) is the leading cause of mortality among infectious diseases globally. Effectively managing TB requires early identification of high-risk individuals. Resource-constrained settings often lack skilled professionals for interpreting chest X-rays (CXRs) used in TB diagnosis. To address this challenge, we developed “DecXpert” a novel Computer-Aided Detection (CAD) software solution based on deep neural networks for early TB diagnosis from CXRs, aiming to detect subtle abnormalities that may be overlooked by human interpretation alone. This study was conducted on the largest cohort size to date, where the performance of a CAD software (DecXpert) was validated against the gold standard molecular diagnostic technique, GeneXpert MTB/RIF, analyzing data from 4,363 individuals across 12 primary health care centers and one tertiary hospital in North India. DecXpert demonstrated 88% sensitivity (95% CI: 0.85-0.93) and 85% specificity (95% CI: 0.82-0.91) for active TB detection. Incorporating demographics, DecXpert achieved an area under the curve of 0.91 (95% CI: 0.88-0.94), indicating robust diagnostic performance. Our findings establish DecXpert's potential as an accurate, efficient AI solution for early identification of active TB cases. Deployed as a screening tool in resource-limited settings, DecXpert could enable identifying high-risk individuals and facilitate effective TB management where skilled radiological interpretation is limited. Health sciences/Health care/Medical imaging/Radiography Health sciences/Medical research/Translational research Biological sciences/Computational biology and bioinformatics/Image processing Biological sciences/Computational biology and bioinformatics/Software Biological sciences/Computational biology and bioinformatics/Machine learning Tuberculosis (TB) Computer-Aided Detection (CAD) Deep convolutional neural networks (CNN or DCNN) Tuberculosis Screening Radiology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction With an estimated 10.4 million new cases and 1.8 million fatalities [ 1 ] from infectious diseases each year, tuberculosis (TB) is the main cause of infectious disease-related deaths globally, posing a constant challenge to public health [ 2 , 3 ]. Mycobacterium tuberculosis (M.Tb) causes this disease, which can be transmitted via airborne means. Its impact significantly affects nearly 25% of the world's population, particularly in regions marked by socioeconomic deprivation [ 2 ]. The most profound impact of TB is observed in regions with lower to moderate incomes, where approximately two-thirds of all cases are concentrated within eight nations: India, Indonesia, China, Nigeria, Bangladesh, the Philippines, Pakistan, and South Africa [ 2 , 3 ]. Treatment for TB exists, with approximately 80% of infections being effectively managed through a six-month regimen of various antibiotics. To effectively combat TB, early detection and identification of high-risk individuals are essential. Unfortunately, a sizable portion, approximately 30% of the total population suffering from TB, fail to inform the World Health Organization (WHO) regarding their infection [ 4 ], underscoring the problem of underdiagnosis. The main TB screening method is based on chest X-ray (CXR) imaging because of its proven efficacy and affordability [ 5 ]. However, a notable challenge arises from the reliance on skilled human interpreters, such as radiologists, clinicians and/or technicians with training in radiology, to interpret CXR results, particularly due to their scarcity in the most affected regions [ 6 , 7 ]. Artificial intelligence-based solutions designed for resource-constrained settings have seen a noteworthy surge in attention due to the global scarcity of skilled CXR interpreters for TB screening [ 8 – 11 ]. In March 2021, for the first time, the WHO recommended that computer-aided detection (CAD) software can replace human readers in interpreting digital CXRs for screening and triaging pulmonary TB disease [ 12 ]. The WHO recommends that CAD may be used to interpret antero-posterior or postero-anterior views of digital CXRs for pulmonary TB in individuals aged 15 years or older [ 12 ]. The ongoing use of digital radiography is more cost-effective than traditional methods, eliminating ongoing expenses associated with reagent use and radiologist services [ 13 ]. The present TB workflow in resource-limited nations is marked by severe delays in finding and diagnosing TB patients [ 14 ], resulting in a significant proportion of cases being diagnosed at the late stages of the disease [ 15 ]. To address this urgent need and in response to the WHO's endorsement of computer-assisted diagnosis of TB, our primary objective was to conduct TB screening on the most extensive cohort to date, comprising of 4,495 participants and compare the accuracies of our novel AI-based CAD software named “DecXpert” with the gold standard molecular reference technique GeneXpert MTB/RIF [ 16 ] in identification of active TB cases. The secondary objective of this study was to benchmark the diagnostic performance of DecXpert against that of 3 certified radiologists. DecXpert is a specialised deep convolutional neural network with self-attention mechanisms specifically designed for detection of active TB cases. This design holds relevance for real-world TB screening in areas lacking specialised personnel and facing resource limitations. Results Patient Demographics A total of 4,495 participants were prospectively enrolled in the study, with enrollment occurring from January 2018 to November 2023 at the 12 primary health care centers and a single tertiary care center. From the initial cohort of 4,495 participants, 132 individuals were excluded from the analysis for various reasons. Among these exclusions, 81 individuals reported an unproductive cough without an available bronchoalveolar lavage (BAL) sample, while 29 participants declined to participate in the study. Additionally, 4 individuals had a documented history of previous TB treatment, and 17 participants were pregnant. However, 1 individual's inclusion was deemed inconclusive due to uncertain results from either the CXR or GeneXpert MTB/RIF test (refer to Figure 1). A total of 4,363 individuals were ultimately included in the analysis. Among the 4,363 individuals included in the study, 680 had an unproductive cough, but their BAL fluid was accessible and available for analysis. The median age of the participants was 43.1 years, and 2,161 males (49.6%) and 2,202 females (50.4%) were included. Predominantly, fever symptoms were evident in the majority of participants, accounting for 2,565 individuals (58.8%) (refer to Table 1). Hemoptysis and night sweats were reported by 437 (10%) and 306 (7%) participants, respectively. Among the individuals in the study cohort, 2,345 individuals (53.7%) were confirmed to be TB positive, while 2,018 individuals (46.3%) tested negative for TB (refer to Figure 1). Notably, within the subgroup with positive TB results, there were 2,161 (49.6%) males and 2,202 (50.4%) females (Table 1). Gender, fever, cough, hemoptysis, and night sweats exhibited significant associations (P-value <0.05) with the GeneXpert MTB/RIF test results (refer to Table 1). Moreover, gender, age, hemoptysis, night sweats and cough demonstrated significant associations (P-value <0.01) with the radiological and DecXpert results (refer to Table 1). Table 1: Demographic and clinical characteristics of the study population stratified by GeneXpert, radiology, and DecXpert test results. The table presents the demographic data (gender and age groups) and clinical characteristics (symptoms and comorbidities) of the 4,363 individuals included in the study. The data is stratified based on the results of the GeneXpert MTB/RIF test (considered the gold standard), radiology interpretation, and the DecXpert Computer-Aided Detection software. Percentages are provided for each subgroup within the respective categories. The P-values indicate the statistical significance of the differences observed between the subgroups. CAD stands for coronary artery disease and CKD for chronic kidney disease. Total GeneXpert +ve GeneXpert -ve Radiology +ve Radiology -ve DecExpert +ve DecExpert -ve P value Gender * Male 2161(49.6%) 991 (45.9%) 1170 (54.1%) 704 (32.5%) 983 (45.5%) 872 (40.3%) 995 (46%) Female 2202 (50.4%) 1354 (61.5%) 848 (38.5%) 961 (43.6%) 712 (32.3%) 1192 (54.1%) 721 (32.7%) Age * <20 349(8%) 161 (46.1%) 188 (53.9%) 114 (32.6%) 158 (45.2%) 154 (44.1%) 160 (45.8%) 21-40 1484(34%) 686 (46.2%) 798 (53.8%) 487 (32.8%) 601 (40.5%) 604 (40.7%) 677 (45.6%) 41-60 1571(36%) 727 (46.3%) 844 (53.7%) 516 (32.8%) 522 (33.2%) 640 (40.7%) 717 (45.6%) >60 959(22%) 443 (46.2%) 516 (53.8%) 315 (32.8%) 301 (31.3%) 360 (37.5%) 439 (45.7%) Symptoms ** Fever 2565 (58.8%) 1374(53.6%) 1191(46.4%) 1093(42.6%) 1084(42.3%) 1354(52.8%) 1097(42.8) Hemoptysis 437 (10%) 301 (68.9%) 136 (31.1%) 252 (57.6%) 114 (26.1%) 264 (60.4%) 115 (26.3%) Night sweats 306 (7%) 104(34%) 202(66%) 56(18.3%) 82(26.8%) 86(28.1%) 171(55.9%) Cough > 2 wks 3683(84.4%) 2946(80%) 737(20%) 2385(64.8%) 542(14.7%) 2593(70.4%) 353(9.6%) Cough < 2 wks 680 (15.6%) 268(39.4%) 412(60.6%) 174(25.6%) 238(35%) 235(34.6%) 350(51.5%) Comorbidities Hypertension 306 (45%) 200 (65.4%) 106 (34.6%) 85 (27.8%) 72 (23.5%) 106 (34.6%) 73 (23.9%) Diabetes 299 (44%) 104 (34.8%) 195 (65.2%) 74 (24.8%) 164 (54.9%) 92 (30.8%) 166 (55.5%) Chronic lung disease 114 (17%) 22 (19.3%) 92 (80.7%) 16 (14%) 77 (67.5%) 19 (16.7%) 78 (68.4%) Chronic liver disease 18 (2.6%) 2 (11.1%) 16 (88.9%) 1 (5.6%) 13 (72.2%) 1 (5.6%) 14 (77.8%) Malignancy 9 (1.3%) 1(11.1%) 8 (88.9%) 0 7 (77.8%) 1 (11.1%) 7 (77.8%) CAD 94 (13.8%) 42 (44.7%) 52 (55.3%) 30 (31.9%) 44 (46.8%) 37 (39.4%) 44 (46.8%) CKD 96 (14.1%) 54 (56.3%) 42 (43.7%) 38 (39.6%) 35 (36.5%) 47 (49%) 36 (37.5%) Radiologists demonstrated a positive predictive value (PPV) of 71%, suggesting that out of the 2,345 individuals identified as positive by GeneXpert MTB/RIF, 1,665 individuals were confirmed positive by radiologists (refer to Table 2). In contrast, DecXpert exhibited a higher PPV of 88%, where out of the 2,345 individuals identified as positive by GeneXpert MTB/RIF, 2,064 were subsequently confirmed as positive by DecXpert. Regarding the negative predictive value (NPV), radiologists achieved an 83.9% NPV. This indicates that among the 2,018 individuals classified as negative by GeneXpert MTB/RIF, 1,695 were confirmed as negative by radiologists. Conversely, DecXpert exhibited a higher negative predictive value (NPV) of 85%, suggesting that of the 2,018 individuals classified as negative for TB by GeneXpert MTB/RIF, 1,716 were confirmed as negative by DecXpert (refer to Table 2). Table 2: Performance evaluation of DecXpert against GeneXpert MTB/RIF and radiologists. This table compares the performance of the DecXpert Computer-Aided Detection software against the gold standard GeneXpert MTB/RIF test and radiologists’ interpretations. The true positive, true negative, positive predictive value (PPV), and negative predictive value (NPV) are reported for each method. The PPV indicates the probability that a positive test result is truly positive, while the NPV represents the probability that a negative test result is truly negative. MTB stands for Mycobacterium tuberculosis , the causative agent of tuberculosis and RIF for Right iliac fossa. Gene Xpert MTB/RIF Radiologists DecXpert True positives 2,345 1,665 2,064 PPV (%) 100 71 88 True negatives 2,018 1,695 1,716 NPV (%) 100 83.9 85 Quantitative assessment We evaluated the performance of different models for the identification of TB using the DecXpert score as a primary predictor (refer to Figure 2). Initially, Model 1, which relies solely on DecXpert scores for TB detection, demonstrated an area under the receiver operating characteristic (ROC) curve (AUC) of 0.85 (95% CI: 0.82-0.87), indicating a reasonably strong predictive ability (refer to Table 3 and Figure 2). Model 2, which incorporated both the DecXpert score and symptom information, displayed an improved AUC of 0.88 (95% CI: 0.83-0.92). When patient demographic information (specifically age and gender) was integrated with DecXpert scores in Model 3, the AUC further increased to 0.91 (95% CI: 0.88-0.94) (refer to Table 3 and Figure 2), indicating an enhanced predictive performance compared to that of earlier models. Furthermore, the development of a composite Model 4, which integrates the DecXpert score, symptom incidence, age, and gender, resulted in a greater AUC of 0.95 (95% CI: 0.90-0.97) (refer to Table 3 and Figure 2). Table 3 : Summarizes the performance of the DecXpert Computer-Aided Detection system with added patient demographics and clinical information . It shows the Area Under the Receiver Operating Characteristic Curve (AUC) and corresponding 95% confidence intervals (CI) for different combinations of patient demographics and clinical data. The table outlines the AUC values achieved by DecXpert scores alone, DecXpert scores combined with symptom information, DecXpert scores combined with age and gender information, and DecXpert scores combined with symptom information, age, and gender. Higher AUC values indicate better discrimination between individuals with and without active tuberculosis. The integration of additional clinical data progressively enhances DecXpert's diagnostic performance, as evidenced by increasing AUC values across the combinations. Models Components AUC 95% CI Model 1 DecXpert scores 0.85 (0.82–0.87) Model 2 DecXpert scores + Symptom information 0.88 (0.83–0.92) Model 3 DecXpert scores + Age + Gender 0.91 (0.88–0.94) Model 4 DecXpert scores + Symptom information + Age + Gender 0.95 (0.90–0.97) Performance of the DecXpert Algorithm Against the Gold Standard Molecular Reference GeneXpert MTB/RIF Next, we assessed the performance of the proposed DecXpert model that has proven to be highly effective in the identification of TB cases from CXR images. The GeneXpert MTB/RIF test reports served as the established ground truth for evaluation. The DecXpert model, operating solely on CXR imaging data without incorporating patient symptoms, achieved an AUC of 0.85 (95% CI=0.82–0.87), indicated by the blue ROC curve in Figure 3. However, upon inclusion of age and gender, the performance of the DecXpert model improved, yielding an AUC of 0.91 (95% CI=0.88–0.94), indicated by the green ROC curve in Figure 3. These results indicate a high level of accuracy ranging between 85% and 91% relative to GeneXpert MTB/RIF test reports, which are considered the gold standard (refer to Figure 3), suggesting that DecXpert reports could potentially serve as a surrogate for GeneXpert MTB/RIF. Notably, when basic patient demographics, specifically age and gender, and patient symptoms such as cough, fever, hemoptysis and night sweats were omitted, the DecXpert model still demonstrated a strong 85% concordance with GeneXpert MTB/RIF testing. Performance of the DecXpert Algorithm against 3 Board-Certified Radiologists Analysing the overall cohort revealed that the DecXpert algorithm successfully identified 2,064 TB patients (88%) out of the total 2,345 GeneXpert MTB/RIF-confirmed positive TB patients, whereas the radiologists identified 1,665 TB patients (71%). Thus, using the DecXpert algorithm increased the overall TB case detection rate by approximately 1.23 times compared to radiologists. Examining the performance of the three board-certified radiologists as illustrated in Figure 4, the first radiologist achieved an AUC of 0.79 (95% CI: 0.74–0.84), the second radiologist achieved an AUC of 0.72 (95% CI: 0.67–0.76), and the third radiologist achieved an AUC of 0.75 (95% CI: 0.71–0.78). Notably, each radiologist's sensitivity/specificity point fell outside the 95% CI space of the ROC curve of the DecXpert model, indicating that their identification performance was inferior to that of the DecXpert model (Figure 4). Furthermore, within the unproductive cough and comorbidity subgroup, there was a considerable improvement in the TB case detection rate, according to DecXpert, which detected 303 patients (71.2%), whereas radiologists were able to detect only 244 patients (57.4%). This observation emphasises the enhanced performance of the DecXpert algorithm compared to that of radiologists, particularly within this subgroup, demonstrating its greater efficiency in identifying TB patients. Qualitative assessment DecXpert went through a validation process that focused on visualizing CXRs and highlighting specific areas in the image that are important for DecXpert to make decisions when classifying TB cases. The assessment highlights the algorithm's reliance on clinically relevant regions within the lung features extracted from CXRs of TB patients to guide its decision-making process. Figure 5 illustrates patient cases presenting highlighted important factors (identified regions) in patients with confirmed TB from the gold standard reference test GeneXpert MTB/RIF. The model relies on accurate visual information and does not consider misleading visual cues such as symbols, motion artifacts, embedded text or symbols, or imaging irregularities when making decisions. This finding demonstrated that DecXpert-related decision-making behaviour is primarily rooted in clinically relevant features. Evaluation of DecXpert Algorithm's Suitability for Deployment in Remote Isolated Regions with Offline and Online Functionality and Minimal Hardware Needs Subsequently, we evaluated the suitability of integrating DecXpert into the pre-existing CXR workflows within primary healthcare facilities and diagnostic centres, particularly for deployment in geographically isolated regions of the nation where computational hardware capabilities are limited. To this end, we examined both online and offline iterations of the DecXpert software and incorporated them into current TB CXR workflows for the purpose of TB screening and diagnosis at primary healthcare facilities and diagnostic centres. The investigation focused on the implementation of DecXpert across seven distinct providers of digital chest X-ray machines, including GE Healthcare™, Siemens™, Philips Healthcare™, FujiFilm Medical Systems™, Shimadzu Corporation™, Toshiba Medical Systems™, and Hitachi Healthcare™. This examination encompassed varying computational hardware setups, ranging from 500 MB to 16 GB of RAM, and spanning different versions of the Windows operating system (2000, 7, 8, and 10). Additionally, the study investigated the compatibility of DecXpert with all five perspectives of CXR images—posteroanterior (PA), anteroposterior (AP), lateral, decubitus, and oblique views. This assessment was conducted at six remote and geographically dispersed locations within the northern region of India. Moreover, we aimed to ascertain the ease of use of DecXpert by the existing X-ray technicians at these sites within their current CXR workflow. DecXpert demonstrated seamless integration and compatibility with all CXR images from the seven vendors, supporting TB screening and diagnostic workflows. Notably, it functioned effectively with basic computational hardware, such as systems with 500 MB RAM and running Windows 2000. Furthermore, the on-site technicians at these healthcare facilities were easily trained on a simple four-step process for processing CXR images (refer to Figure 6-a,b,c,d). DecXpert was made available in both offline and online configurations which had the same functionality, and Figure 6 (a,b,c,d) depicts this straightforward process, wherein both the CXR images and patient demographics are uploaded to the DecXpert software, this process culminates in the creation of a probable diagnostic report in PDF format, which can be disseminated in both electronic and printed forms for assessment by a physician. Discussion Our investigation aimed to evaluate the efficacy of the AI-based DecXpert software in detecting TB in a resource-limited setting with a high disease burden but without HIV co-infection. Notably, this represents the largest cohort study to date, comprising 4,363 participants who underwent GeneXpert MTB/RIF testing for comparison with both DecXpert and radiologist interpretations. DecXpert effectively identified 88% (2,064 out of 2,345) of TB cases diagnosed by GeneXpert MTB/RIF, showcasing its potential to reduce the necessity for expensive molecular tests. The current TB workflow in resource-limited nations suffers from severe diagnostic delays, with a significant proportion of cases detected at late disease stages. This delayed diagnosis poses a significant public health challenge, as it increases the risk of disease transmission and complications. The development of DecXpert, offering both online and offline deployment for automated CXR interpretation, represents a milestone in leveraging technological advancements for large-scale TB screening efforts. By enabling early and accurate identification of TB cases, DecXpert has the potential to revolutionize the existing diagnostic paradigm in resource-constrained settings. Furthermore, the use of DecXpert as a triage tool could enhance case identification and potentially reduce program expenses by optimizing the utilization of costly molecular testing resources such as Cartridge-Based Nucleic Acid Amplification Test (CBNAAT). Our findings demonstrate that DecXpert achieves substantial diagnostic accuracy, with an area under the receiver operating characteristic curve (AUC) of 0.91 (95% confidence interval [CI]: 0.88–0.94) when integrating patient age and gender (Model 3). The incorporation of demographic data facilitated the creation of personalized risk probabilities based on quantitative assessments of TB at the individual CXR level. This approach empowers frontline healthcare personnel with informed decision-making capabilities regarding individual testing prioritization. In settings with limited GeneXpert MTB/RIF testing resources, DecXpert can optimize testing utilization by reducing the number of cartridges used, offering significant benefits in healthcare facilities experiencing high patient burdens. Furthermore, we explored the integration of DecXpert into existing CXR workflows within primary healthcare facilities and diagnostic centers, with a particular focus on deployment in geographically isolated regions with limited computational infrastructure. DecXpert demonstrated seamless integration and compatibility with CXR images from 7 different CXR vendors, supporting TB screening and diagnostic workflows across six remote locations in northern India, even on systems with basic computational hardware, such as those with 500 MB RAM and running Windows 2000. This highlights the versatility and adaptability of the software, making it suitable for deployment in resource-constrained settings with diverse hardware configurations. Notably, on-site technicians at these healthcare facilities were easily trained on a simple four-step process for processing CXR images using DecXpert (Fig. 6 ). The software was made available in both offline and online configurations, offering the same functionality and thereby accommodating diverse operational environments. The streamlined process involved uploading CXR images and patient demographics, culminating in the generation of a probable diagnostic report in a portable document format (PDF) for physician assessment (Fig. 6 ). This user-friendly interface and workflow integration underscore the potential for widespread adoption of DecXpert as a valuable screening and diagnostic tool for TB, particularly in resource-limited and geographically isolated regions. A critical aspect of our study was the validation process, which emphasized DecXpert's reliance on clinically relevant regions within the lung features extracted from CXRs for decision-making. By visualizing the specific areas within CXRs that were crucial for classifying TB cases, as illustrated in Fig. 5 , the study highlights the algorithm's ability to identify and prioritize the relevant radiological features associated with TB disease. Notably, the study found that DecXpert's decision-making process relies on accurate visual information and does not consider misleading visual cues, such as symbols, motion artifacts, embedded text or symbols, or imaging irregularities. The emphasis on clinically relevant regions within the lung features extracted from chest X-rays aligns with the established understanding of TB manifestations and the radiological patterns associated with the disease. By focusing on these specific areas, DecXpert can leverage the diagnostic information contained within the images, potentially enhancing its ability to accurately identify TB cases. These findings contribute to the interpretability and trustworthiness of the system, potentially paving the way for its wider adoption and integration into clinical workflows for TB screening and diagnosis. While the utilization of digital X-rays has expanded the potential for TB diagnosis and garnered community interest, operational considerations, health communication strategies, resource availability, and pathway development for individuals not diagnosed with TB remain crucial elements for the effective implementation of such programs. It is essential to address these factors to ensure that the benefits of DecXpert and similar technologies are fully realized and translated into improved patient outcomes. This study is subject to limitations, including a reliance on GeneXpert MTB/RIF as the reference standard and a setting primarily within a facility-based context rather than a community setting. While GeneXpert MTB/RIF is widely recognized as the gold standard for TB diagnosis, future studies could explore the performance of DecXpert against other diagnostic modalities or in community-based settings to further validate its applicability across diverse scenarios. Nonetheless, the introduction of DecXpert, a deep learning-powered tool for TB screening, demonstrated strong performance in detecting TB cases and explicable decision-making behavior. This technological innovation aims to aid frontline healthcare workers in high-risk regions combating TB, a disease that continues to pose a significant global health challenge. By leveraging the power of artificial intelligence and leveraging its ability to provide interpretable and transparent predictions, DecXpert has the potential to revolutionize TB screening and diagnosis, particularly in resource-limited settings. In summary, our study underscores the potential of DecXpert as a valuable screening tool for TB, particularly in resource-limited settings. Its integration with demographic data shows promise for personalized risk assessment, offering insights that could enhance testing strategies and clinical workflows in the ongoing fight against TB. The robust diagnostic performance, adaptability to diverse healthcare settings, and explicable decision-making process contribute to the overall utility and trustworthiness of DecXpert. As we continue to grapple with the global burden of TB, innovative technologies like DecXpert represent a significant stride towards improving early detection, optimizing resource utilization, and ultimately saving lives. Methods Study Design This study included an extensive passive case-finding investigation of prospectively enrolled patients. This multicenter study was carried out across 12 primary health care (PHC) centres and one tertiary care (TC) centre in the northern Indian region. Patient enrollment occurred from January 2018 to January 2022 at the 12 PHC centres and from April 2022 to November 2023 at the one tertiary care centre. All patients who presented to the concerned centre with a history of fever, cough, expectoration, or constitutional symptoms for more than two weeks were included in the study. Those patients who had all the above symptoms but had nonproductive cough were subjected to bronchoalveolar lavage (BAL) for confirmation of diagnosis if the facility was available at that centre; otherwise, they were not included. The study excluded pregnant females, patients who declined to participate, individuals already undergoing TB treatment, those presenting with a nonproductive cough, and participants for whom BAL fluid samples were unavailable (refer to Fig. 1 ). Patients with TB were identified as individuals aged 15 years and older who had both a CXR and a GeneXpert MTB/RIF test confirming pulmonary TB. Patients without TB were identified as individuals aged 15 years and older who underwent a CXR, with a negative GeneXpert MTB/RIF [ 16 ] test ruling out TB diagnosis. The diagnostic capabilities were evaluated by utilising the sensitivity (Se), specificity (Sp), and the count of false positives, false negatives, true positives, and true negatives (FP, FN, TP, and TN, respectively) in relation to the results of the GeneXpert MTB/RIF test. Scoring Chest X-rays using DecXpert By utilising outputs from its detection systems, the software treats them as descriptive features extracted from images. These features are then used to train a K-nearest neighbors classifier [ 17 ], allowing for the computation of a cumulative abnormality/severity score for each CXR, which ranges from 0 to 100. A higher score indicates a more pronounced abnormality, potentially indicating TB. Individuals with DecXpert scores of 50 or higher were recommended to undergo additional clinical assessment by their physicians (refer to Fig. 7 ). Dataset and Patients Enrolled We prospectively included all patients who met the inclusion criteria across 12 primary health care centres, along with individuals who sought care at the Pulmonary Medicine Department of a tertiary care centre in northern India. All participants underwent CXR and GeneXpert MTB/RIF testing as part of the screening process. Among the initial 4,495 participants screened, 4,364 individuals were recruited into our study. However, the study included 4,363 individuals, and one of the recruited individuals had an inconclusive GeneXpert MTB/RIF test result. Within this cohort, 2,345 CXR images were obtained from patients diagnosed with TB, while 2,018 CXR images were acquired from individuals without TB, as ascertained by their GeneXpert MTB/RIF test results, regarded as the reference gold standard. Deep Learning Algorithm We used a previously constructed deep learning algorithm called DecXpert (version 1.1), which was trained on 9,876 chest X-ray images. Regarding the distribution of data used to train and validate this algorithm, there were 4,932 chest X-ray images from patients with TB and 4,944 chest X-ray images from patients without TB. The patients for the training, validation, and test datasets were selected randomly from the chest X-ray images dataset, constituting 80%, 10%, and 10% of the dataset, respectively. All CXRs that were input into the algorithm were resized to the dimensions of 224x224 pixels. The following guidelines or limitations were established for developing DecXpert algorithm, outlined through an indicator function: (a) ensuring a sensitivity exceeding 80%, (b) achieving a specificity of more than 75%, and (c) constraining the number of parameters to 2 million. To specify the general framework and specific attributes of a deep neural network customised for TB detection, we employed generative synthesis [ 18 ]. Using an optimal generator G, these designs were automatically found. Let A represent the generated architecture of the neural network based on a given set of initial configurations C, aimed at optimising a universal performance metric P [ 19 ]. This optimisation process is subject to specific constraints determined by an indicator function ϕ. $$A=\text{m}\text{a}\text{x}P\left(A\left(c\right)\right)\text{ }\text{s}\text{u}\text{c}\text{h} \text{t}\text{h}\text{a}\text{t}\text{ }{1}_{\varphi }\left(A\left(c\right)\right)=1;\forall c\in C$$ 1 The aim was to find architectures that maximise performance while meeting predetermined criteria. First, the significant diversity of both macro and microarchitectures within the overall network architecture is a clear feature. This variation results from applying a meticulously tailored machine learning approach for TB case detection using CXR images, aiming to strike the best possible balance between efficiency and accuracy. Second, the network architecture mainly comprises of convolutions applied on a depth and pointwise basis. This use of lightweight design patterns illustrates how the machine-driven design exploration strategy can modify the deep neural network microarchitecture in response to the imposed architectural complexity constraints. For DecXpert to potentially be widely adopted, its effective and efficient architecture is especially important. This is because DecXpert frequently needs to be deployed on low-end, low-cost computing devices in resource-constrained regions affected by poverty and economic constraints [ 20 ]. Third, A notable aspect of the DecXpert network architecture includes visual attention condensers [ 21 ], recently introduced, these mechanisms are a type of efficient attention condenser. These visual attention condensers enhance representational ability, requiring less computational and architectural complexity. Finally, this architecture predicts a patient's tuberculosis status as positive or negative. Model training Through the utilisation of stochastic gradient descent optimisation, the proposed neural network structure underwent training [ 22 ]. The training process involved a learning rate of 0.001, a momentum value of 0.6, and a batch size of 12, spanning across 400 epochs. Furthermore, data augmentation techniques were applied, including horizontal flipping, random cropping (15% margin), random contrast shifting (10% margin), and random intensity shifting (10% margin), were applied during the training phase [ 23 ]. After that, each image underwent resizing to a dimension of 224 × 224 pixels. Our experiments have shown that this specific size provides the highest level of performance, allowing for the conservation of important textural characteristics needed to differentiate individuals who are TB-positive from those who are TB-negative. No additional performance improvements were observed when using higher resolutions. The DecXpert deep neural network architecture underwent all construction, training, and evaluation procedures using the TensorFlow deep learning framework [ 24 ]. Statistical analysis: For the statistical analysis, the results were scrutinised using DecXpert, GeneXpert MTB/RIF, and the evaluations of three certified radiologists. Regarding the demographic data, continuous variables were presented as the mean and standard deviation, while categorical variables were described in terms of numbers and percentages. A comparative analysis was conducted between the other methods and GeneXpert, which was considered as the gold standard. Additionally, a comparison was made between DecXpert and the radiologists' assessments. The sensitivity (Sn), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV) were calculated based on the positive and negative results obtained. The performance of these methods, including 95% confidence intervals (CI), was compared in differentiating between positive and negative patients for TB. The Sn, Sp, PPV, NPV, and their corresponding 95% CI were computed and reported. Receiver operating characteristic (ROC) curve analyses and area under curve (AUC) values, along with their respective 95% CI, were presented for the outcomes of positive patients versus negative. All statistical analyses were performed using R programming (version 4.1) [ 25 ], and a statistical significance was determined with a two-tailed p-value below 0.05. Statistical significance was assessed with *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, indicating progressively higher significance levels. Positive Predictive Value (PPV) The positive predictive value is the proportion of true positive cases among all the cases that tested positive. In other words, it measures the probability that a positive result actually indicates the presence of the condition. PPV = (true positives)/(true positives + false positives) Negative Predictive Value (NPV) The negative predictive value is the proportion of true negative cases among all the cases that tested negative. It measures the probability that a negative result actually indicates the absence of the condition. NPV = (true negatives)/(true negatives + false negatives) Molecular validation using GeneXpert MTB/RIF GeneXpert MTB/RIF [ 16 ] was employed to verify the findings derived from DecXpert, encompassing both TB positive and negative cases, thereby enhancing confidence in the results. Radiological validation Three board-certified radiologists analysed and reported the patient’s chest X-rays. The first radiologist had more than 3 years of experience, the second had more than 5 years of experience, and the third had more than 6 years of experience. A majority vote (75%) among the three radiologists was required to classify a CXR as TB or normal. The radiological interpretations of the CXRs were subsequently compared to those of the DecXpert and GeneXpert MTB/RIF results. Ethical compliance The research study obtained ethical clearance from the Institutional Ethics Committee (IEC) at Sanjay Gandhi Post Graduate Institute of Medical Sciences, with the approval code 2022-59-IMP-EXP-46, ensuring adherence to ethical standards. All data related to the study participants was de-identified and anonymized. Additionally, every participant verbally consented to be part of the study after being informed about its details, ensuring compliance with ethical norms set by the ethics committee. The study methodologies adhered to applicable guidelines and regulations. The radiologists and researchers involved in the study only had access to non-identifiable patient data for analysis purposes. In contrast, all identifiable patient screening and diagnostic information remained securely stored on an internally protected server accessible only through credential-based authentication. Declarations Data availability The data produced and analyzed in the present study can be obtained from the corresponding author upon request, subject to reasonable conditions. Acknowledgements We thank Mr. Abid Mohsin Zaidi from the Department of Computer Science, Ambalika Institute of Management & Technology, Lucknow, and Ms. Rachna Shaw from the School of Physical Sciences, Indian Institute of Technology (IIT) Mandi, for their help with data analytics and preparing the figures for this article. Author information Author notes # These authors contributed equally: Alok Nath, Zia Hashim Authors and Affiliations 1 Department of Pulmonary Medicine, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Raebareli Road, Lucknow, Uttar Pradesh, India 226014 2 Indian Institute of Technology Kanpur, Department of Electrical Engineering, Kalyanpur, Kanpur, Uttar Pradesh 208016, India 3 Faculty of Medicine, The University of Queensland, Translational Research Institute, 37 Kent Street, Brisbane, QLD, 4102, Australia 4 Baker Heart and Diabetes Institute, Melbourne, VIC, 3004, Australia. Contributions A.N., Z.H., S.S, P.A.P, N.M. and A.S. designed the study. N.M., M.S., S.S and A.S. performed the analyses. A.S., N.M. and M.S. wrote scripts and provided analysis tools. A.N., Z.H., P.A.P, S.S, N.M, M.S and A.S. provided critical intellectual content for the design of the study. A.N., Z.H., P.A.P, S.S, M.S and A.S. wrote the paper. Corresponding author Correspondence to Ankit Shukla ( [email protected] ) Ethics declarations Competing interests The authors declare no competing interests. References Heslop, R., et al., Changes in Host Cytokine Patterns of TB Patients with Different Bacterial Loads Detected Using 16S rRNA Analysis . PLOS ONE, 2016. 11(12): p. e0168272. WHO, Global Tuberculosis Report 2023. Hillson, R., Tuberculosis and diabetes . Practical Diabetes, 2017. 34(5): p. 149–150. Khan, A.J., et al., Engaging the private sector to increase tuberculosis case detection: an impact evaluation study . Lancet Infect Dis, 2012. 12(8): p. 608–16. Kranzer, K., et al., The benefits to communities and individuals of screening for active tuberculosis disease: a systematic review . Int J Tuberc Lung Dis, 2013. 17(4): p. 432–46. Mollura, D.J., et al., White Paper Report of the RAD-AID Conference on International Radiology for Developing Countries: identifying challenges, opportunities, and strategies for imaging services in the developing world. J Am Coll Radiol, 2010. 7(7): p. 495–500. Candemir, S. and S. Antani, A review on lung boundary detection in chest X-rays . Int J Comput Assist Radiol Surg, 2019. 14(4): p. 563–576. Breuninger, M., et al., Diagnostic accuracy of computer-aided detection of pulmonary tuberculosis in chest radiographs: a validation study from sub-Saharan Africa . PLoS One, 2014. 9(9): p. e106381. Melendez, J., et al., An automated tuberculosis screening strategy combining X-ray-based computer-aided detection and clinical information . Sci Rep, 2016. 6: p. 25265. Muyoyeta, M., et al., The sensitivity and specificity of using a computer aided diagnosis program for automatically scoring chest X-rays of presumptive TB patients compared with Xpert MTB/RIF in Lusaka Zambia . PLoS One, 2014. 9(4): p. e93757. Rahman, M.T., et al., An evaluation of automated chest radiography reading software for tuberculosis screening among public- and private-sector patients . Eur Respir J, 2017. 49(5). WHO consolidated guidelines on tuberculosis Module 2: Screening – Systematic screening for tuberculosis disease. Qin, Z.Z., et al., How is Xpert MTB/RIF being implemented in 22 high tuberculosis burden countries? Eur Respir J, 2015. 45(2): p. 549–54. Sreeramareddy, C.T., et al., Delays in diagnosis and treatment of pulmonary tuberculosis in India: a systematic review . Int J Tuberc Lung Dis, 2014. 18(3): p. 255–266. Vonasek, B., et al., Screening tests for active pulmonary tuberculosis in children . Cochrane Database of Systematic Reviews, 2021. 2021(10). Ioannidis, P., et al., Cepheid GeneXpert MTB/RIF assay for Mycobacterium tuberculosis detection and rifampin resistance identification in patients with substantial clinical indications of tuberculosis and smear-negative microscopy results . J Clin Microbiol, 2011. 49(8): p. 3068–70. Mucherino, A., P.J. Papajorgji, and P.M. Pardalos, k-Nearest Neighbor Classification , in Data Mining in Agriculture , A. Mucherino, P.J. Papajorgji, and P.M. Pardalos, Editors. 2009, Springer New York: New York, NY. p. 83–106. Elsken, T., J. Metzen, and F. Hutter, Neural Architecture Search: A Survey . 2018. Wong, A., et al., FermiNets: Learning generative machines to generate efficient neural networks via generative synthesis . 2018. Howard, A., et al., MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. 2017. Wang, L., Z.Q. Lin, and A. Wong, COVID-Net: a tailored deep convolutional neural network design for detection of COVID-19 cases from chest X-ray images . Scientific Reports, 2020. 10(1): p. 19549. Ruder, S., An overview of gradient descent optimization algorithms. 2016. Shorten, C. and T.M. Khoshgoftaar, A survey on Image Data Augmentation for Deep Learning . Journal of Big Data, 2019. 6(1): p. 60. Abadi, M., et al., TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems. 2016. R Core Team, R., R: A language and environment for statistical computing. 2013. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 05 Sep, 2024 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 11 Jun, 2024 Reviews received at journal 10 Jun, 2024 Reviewers agreed at journal 10 Jun, 2024 Reviews received at journal 26 May, 2024 Reviewers agreed at journal 23 May, 2024 Reviewers invited by journal 23 May, 2024 Editor assigned by journal 23 May, 2024 Editor invited by journal 23 May, 2024 Submission checks completed at journal 23 May, 2024 First submitted to journal 06 May, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4377653","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":301036743,"identity":"6c5f5a92-c56b-4fef-8160-f4a9b5116ac2","order_by":0,"name":"Alok Nath","email":"","orcid":"","institution":"Sanjay Gandhi Post Graduate Institute of Medical Sciences","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Alok","middleName":"","lastName":"Nath","suffix":""},{"id":301036748,"identity":"89ae8583-43b5-427f-9457-10d69e0a8e1e","order_by":1,"name":"Zia Hashim","email":"","orcid":"","institution":"Sanjay Gandhi Post Graduate Institute of Medical Sciences","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Zia","middleName":"","lastName":"Hashim","suffix":""},{"id":301036754,"identity":"2d1d9906-37bd-4fe0-9bf3-b2c3f98d51dd","order_by":2,"name":"Saumya Shukla","email":"","orcid":"","institution":"Sanjay Gandhi Post Graduate Institute of Medical Sciences","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Saumya","middleName":"","lastName":"Shukla","suffix":""},{"id":301036759,"identity":"5bba3dc8-9101-45dd-b974-e46b342cbaf0","order_by":3,"name":"Prasanth Areekkara Poduvattil","email":"","orcid":"","institution":"Sanjay Gandhi Post Graduate Institute of Medical Sciences","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Prasanth","middleName":"Areekkara","lastName":"Poduvattil","suffix":""},{"id":301036764,"identity":"fcea40a3-22f2-4940-b690-dfd8f0e6288a","order_by":4,"name":"Manika Singh","email":"","orcid":"","institution":"Baker IDI Heart and Diabetes Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Manika","middleName":"","lastName":"Singh","suffix":""},{"id":301036771,"identity":"9b92c5c5-7fd0-432d-be51-05facbb31c17","order_by":5,"name":"Nikhil Misra","email":"","orcid":"","institution":"Indian Institute of Technology Kanpur","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Nikhil","middleName":"","lastName":"Misra","suffix":""},{"id":301036775,"identity":"35cea092-3bde-4918-b4f6-35bce79f0608","order_by":6,"name":"Ankit Shukla","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABKklEQVRIie3QwUrDMBgH8ITATmW7dhTWV2gpOIrzXRIG3aV6EWSn2jLIqXpecQ/RR0gp6KWb1xw86N1DYZeOXvzmZlFWp0fB/CEhJPmRLx9CKip/Mh0kYKZIIyEqv5yII0S8ExziebNLjxP0QYj2G9K74ywvUXBhFVm0PuOnZs+4yZ43FRp0JcVr7ZDoT/cCCssvrWU0M875xE4Wq7EdU+T0JSVGC0FyEgIRLH3EIZCcWtI/MaAwlkqK2oi5I8GWzGp3R4Y1kGsgpG4hlvS2hRGWLiNu4P0rBAgsaKftFVt6VBRWzpIi4268gr/Mfacfe7qdFC/cXRySgfSccjoN2G0xzmV1BR3TfbusRiOz+wA7r+2NhvKaFWlaCQOH34HPIT9fUVFRUfl3eQPOn3JI87y9TAAAAABJRU5ErkJggg==","orcid":"","institution":"University of Queensland","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Ankit","middleName":"","lastName":"Shukla","suffix":""}],"badges":[],"createdAt":"2024-05-06 14:47:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4377653/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4377653/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-024-71346-x","type":"published","date":"2024-09-05T15:57:18+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":56487261,"identity":"67199494-0193-4cfb-87f7-ed092500da0c","added_by":"auto","created_at":"2024-05-14 20:57:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":97732,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWorkflow for evaluating the performance of DecXpert CAD software. \u003c/strong\u003eFrom the initially enrolled 4,495 individuals, 4,364 participants were included in the analysis after excluding those with no chest X-ray and/or inconclusive GeneXpert MTB/RIF results. The GeneXpert MTB/RIF test results categorized participants as TB(+) (2,345) or TB(-) (2,018). Radiologists interpreted the chest X-rays, identifying 1,665 as TB(+) and 1,695 as TB(-). The DecXpert CAD system classified 2,064 cases as TB(+) and 1,716 as TB(-). The performance metrics calculated include positive predictive value (PPV) and negative predictive value (NPV) for each method, using the GeneXpert results as the reference standard.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/b13338487203670490138ce0.png"},{"id":56487193,"identity":"9e3b6e05-f735-46b3-bc6c-26d29c44361f","added_by":"auto","created_at":"2024-05-14 20:57:10","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":38085,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIllustrates the diagnostic accuracy of DecXpert assessed through receiver operating characteristic (ROC) curves generated from the evaluated models in this study. \u003c/strong\u003eThe x-axis denotes different combinations of information used, while the y-axis indicates the corresponding AUC values. The bars represent scenarios including DecXpert scores alone, DecXpert scores combined with symptom information, DecXpert scores combined with age and gender information, and DecXpert scores combined with symptom information, age, and gender. Higher AUC values signify improved discrimination between individuals with and without active tuberculosis. The error bars depict the 95% confidence intervals for each AUC value.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/2f78e13d1152b2fa3e93f48f.png"},{"id":56487265,"identity":"428cfd34-b283-49be-a2c8-9466d6ae84f7","added_by":"auto","created_at":"2024-05-14 20:57:20","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":69679,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eShows Receiver Operating Characteristic (ROC) curves for the DecXpert Computer-Aided Detection system\u003c/strong\u003e. The blue curve represents DecXpert scores alone, yielding an Area Under the Curve (AUC) of 0.85 (95% CI: 0.82-0.87), while the green curve represents DecXpert scores combined with age and gender information, resulting in an improved AUC of 0.91 (95% CI: 0.88-0.94). The x-axis illustrates the false positive rate (1 - specificity), while the y-axis depicts the true positive rate (sensitivity). The ROC curve illustrates the trade-off between sensitivity and specificity at different threshold settings of the diagnostic system. Higher AUC values indicate better overall accuracy in discriminating between individuals with and without tuberculosis. Including age and gender information enhances DecXpert's diagnostic performance, as evidenced by the higher AUC value for the green curve.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/4116231ac5931a7d8aa8a866.png"},{"id":56487243,"identity":"90843a25-8ce5-4f03-8fb6-acf29f478d33","added_by":"auto","created_at":"2024-05-14 20:57:14","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":149751,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDiagnostic accuracy of DecXpert and three board certified radiologist experts compared with the reference GeneXpert MTB/RIF. \u003c/strong\u003eThe receiver operating characteristic (ROC) curve illustrates the performance of the DecXpert model (utilising age and gender) alongside the performance metrics of three board-certified radiologist experts, all plotted within the same ROC space. The area under the curve (AUC) quantifies the overall discriminative ability, while CI denotes the confidence interval surrounding the AUC estimation. The asterisk, bubble and square symbols represent the first, second and third radiologists respectively.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/13f34dade8359fbed1e7049b.png"},{"id":56487240,"identity":"06510483-8a79-47df-8cd4-13ed8dfb23e8","added_by":"auto","created_at":"2024-05-14 20:57:12","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":190985,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVisualizing analyzed chest X-rays from sample TB patients. \u003c/strong\u003eSample chest X-ray images of TB patients are shown, with the highlighted areas demonstrating the most important aspects detected by the DecXpert system for identifying tuberculosis-related abnormalities\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/df11a53e54a67a19f1fdc45d.png"},{"id":56487195,"identity":"6c48fd6b-b0ed-47df-8cbb-396fa55dd657","added_by":"auto","created_at":"2024-05-14 20:57:11","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":138584,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUser interface and output displays of the DecXpert software. \u003c/strong\u003e(a) The DecXpert software login screen. (b) The interface for uploading a chest X-ray image and entering patient details. (c) The output screen displaying the analyzed chest X-ray image along with a color-coded risk assessment for tuberculosis and other potential abnormalities. (d) A portable document format (PDF) report providing an overall risk score and quantitative assessment for TB and 18 other abnormalities, as well as a summary of clinical diagnosis.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/699f62d73d1c9ab1778ddc7e.png"},{"id":56487264,"identity":"0b85ef09-4fce-488e-8fbb-383c3c4aad56","added_by":"auto","created_at":"2024-05-14 20:57:20","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":35152,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvaluation of DecXpert CAD performance against the gold standard molecular reference test GeneXpert MTB/RIF. \u003c/strong\u003eThe figure illustrates the workflow followed for comparing DecXpert and GeneXpert MTB/RIF test results, and determining false positives (FP), false negatives (FN), true positives (TP), and true negatives (TN) for DecXpert, using GeneXpert MTB/RIF as the reference standard.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/ad52d26b9f619f34b44f5c15.png"},{"id":64186290,"identity":"eb53208d-2e7b-4a59-9142-625721355004","added_by":"auto","created_at":"2024-09-09 16:26:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1746437,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4377653/v1/bc15557c-96f1-4f45-9cc6-1038edb2b679.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Assessing Diagnostic Accuracy and Viability of AI-Assisted Tuberculosis Detection in Northern Indian Healthcare Facilities: A Multicenter Study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWith an estimated 10.4\u0026nbsp;million new cases and 1.8\u0026nbsp;million fatalities [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e] from infectious diseases each year, tuberculosis (TB) is the main cause of infectious disease-related deaths globally, posing a constant challenge to public health [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003cem\u003eMycobacterium tuberculosis\u003c/em\u003e (M.Tb) causes this disease, which can be transmitted via airborne means. Its impact significantly affects nearly 25% of the world's population, particularly in regions marked by socioeconomic deprivation [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The most profound impact of TB is observed in regions with lower to moderate incomes, where approximately two-thirds of all cases are concentrated within eight nations: India, Indonesia, China, Nigeria, Bangladesh, the Philippines, Pakistan, and South Africa [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Treatment for TB exists, with approximately 80% of infections being effectively managed through a six-month regimen of various antibiotics.\u003c/p\u003e \u003cp\u003eTo effectively combat TB, early detection and identification of high-risk individuals are essential. Unfortunately, a sizable portion, approximately 30% of the total population suffering from TB, fail to inform the World Health Organization (WHO) regarding their infection [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], underscoring the problem of underdiagnosis. The main TB screening method is based on chest X-ray (CXR) imaging because of its proven efficacy and affordability [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. However, a notable challenge arises from the reliance on skilled human interpreters, such as radiologists, clinicians and/or technicians with training in radiology, to interpret CXR results, particularly due to their scarcity in the most affected regions [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Artificial intelligence-based solutions designed for resource-constrained settings have seen a noteworthy surge in attention due to the global scarcity of skilled CXR interpreters for TB screening [\u003cspan additionalcitationids=\"CR9 CR10\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn March 2021, for the first time, the WHO recommended that computer-aided detection (CAD) software can replace human readers in interpreting digital CXRs for screening and triaging pulmonary TB disease [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The WHO recommends that CAD may be used to interpret antero-posterior or postero-anterior views of digital CXRs for pulmonary TB in individuals aged 15 years or older [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The ongoing use of digital radiography is more cost-effective than traditional methods, eliminating ongoing expenses associated with reagent use and radiologist services [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe present TB workflow in resource-limited nations is marked by severe delays in finding and diagnosing TB patients [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], resulting in a significant proportion of cases being diagnosed at the late stages of the disease [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. To address this urgent need and in response to the WHO's endorsement of computer-assisted diagnosis of TB, our primary objective was to conduct TB screening on the most extensive cohort to date, comprising of 4,495 participants and compare the accuracies of our novel AI-based CAD software named \u0026ldquo;DecXpert\u0026rdquo; with the gold standard molecular reference technique GeneXpert MTB/RIF [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] in identification of active TB cases. The secondary objective of this study was to benchmark the diagnostic performance of DecXpert against that of 3 certified radiologists. DecXpert is a specialised deep convolutional neural network with self-attention mechanisms specifically designed for detection of active TB cases. This design holds relevance for real-world TB screening in areas lacking specialised personnel and facing resource limitations.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003ePatient Demographics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 4,495 participants were prospectively enrolled in the study, with enrollment occurring from January 2018 to November 2023 at the 12 primary health care centers and a single tertiary care center. From the initial cohort of 4,495 participants, 132 individuals were excluded from the analysis for various reasons. Among these exclusions, 81 individuals reported an unproductive cough without an available bronchoalveolar lavage (BAL) sample, while 29 participants declined to participate in the study. Additionally, 4 individuals had a documented history of previous TB treatment, and 17 participants were pregnant. However, 1 individual\u0026apos;s inclusion was deemed inconclusive due to uncertain results from either the CXR or GeneXpert MTB/RIF test (refer to Figure 1). A total of 4,363 individuals were ultimately included in the analysis. Among the 4,363 individuals included in the study, 680 had an unproductive cough, but their BAL fluid was accessible and available for analysis. The median age of the participants was 43.1 years, and 2,161 males (49.6%) and 2,202 females (50.4%) were included. Predominantly, fever symptoms were evident in the majority of participants, accounting for 2,565 individuals (58.8%) (refer to Table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHemoptysis and night sweats were reported by 437 (10%) and 306 (7%) participants, respectively. Among the individuals in the study cohort, 2,345 individuals (53.7%) were confirmed to be TB positive, while 2,018 individuals (46.3%) tested negative for TB (refer to Figure 1). Notably, within the subgroup with positive TB results, there were 2,161 (49.6%) males and 2,202 (50.4%) females (Table 1). Gender, fever, cough, hemoptysis, and night sweats exhibited significant associations (P-value \u0026lt;0.05) with the GeneXpert MTB/RIF test results (refer to Table 1). Moreover, gender, age, hemoptysis, night sweats and cough demonstrated significant associations (P-value \u0026lt;0.01) with the radiological and DecXpert results (refer to Table 1).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eTable 1: Demographic and clinical characteristics of the study population stratified by GeneXpert, radiology, and DecXpert test results.\u0026nbsp;\u003c/strong\u003eThe table presents the demographic data (gender and age groups) and clinical characteristics (symptoms and comorbidities) of the 4,363 individuals included in the study. The data is stratified based on the results of the GeneXpert MTB/RIF test (considered the gold standard), radiology interpretation, and the DecXpert Computer-Aided Detection software. Percentages are provided for each subgroup within the respective categories. The P-values indicate the statistical significance of the differences observed between the subgroups. CAD stands for coronary artery disease and CKD for chronic kidney disease.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"653\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eTotal\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGeneXpert +ve\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGeneXpert -ve\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eRadiology +ve\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eRadiology -ve\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDecExpert +ve\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDecExpert \u0026nbsp; \u0026nbsp; -ve\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eP value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGender\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eMale\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e2161(49.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e991 (45.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e1170 (54.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e704 (32.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e983 (45.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e872 (40.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e995 (46%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"15.025041736227045%\" valign=\"top\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.186978297161936%\" valign=\"top\"\u003e\n \u003cp\u003e2202 (50.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.186978297161936%\" valign=\"top\"\u003e\n \u003cp\u003e1354 (61.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.186978297161936%\" valign=\"top\"\u003e\n \u003cp\u003e848 (38.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.517529215358932%\" valign=\"top\"\u003e\n \u003cp\u003e961 (43.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.684474123539232%\" valign=\"top\"\u003e\n \u003cp\u003e712 (32.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.52253756260434%\" valign=\"top\"\u003e\n \u003cp\u003e1192 (54.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.689482470784641%\" valign=\"top\"\u003e\n \u003cp\u003e721 (32.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e349(8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e161 (46.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e188 (53.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e114 (32.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e158 (45.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e154 (44.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e160 (45.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e21-40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e1484(34%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e686 (46.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e798 (53.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e487 (32.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e601 (40.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e604 (40.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e677 (45.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e41-60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e1571(36%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e727 (46.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e844 (53.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e516 (32.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e522 (33.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e640 (40.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e717 (45.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026gt;60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e959(22%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e443 (46.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e516 (53.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e315 (32.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e301 (31.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e360 (37.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e439 (45.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSymptoms\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eFever\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e2565 (58.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e1374(53.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e1191(46.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e1093(42.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e1084(42.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e1354(52.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e1097(42.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eHemoptysis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e437 (10%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e301 (68.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e136 (31.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e252 (57.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e114 (26.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e264 (60.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e115 (26.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eNight sweats\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e306 (7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e104(34%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e202(66%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e56(18.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e82(26.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e86(28.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e171(55.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eCough \u0026gt; 2 wks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e3683(84.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e2946(80%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e737(20%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e2385(64.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e542(14.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e2593(70.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e353(9.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eCough \u0026lt; 2 wks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e680 (15.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e268(39.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e412(60.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e174(25.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e238(35%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e235(34.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e350(51.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eComorbidities\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eHypertension\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e306 (45%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e200 (65.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e106 (34.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e85 (27.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e72 (23.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e106 (34.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e73 (23.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eDiabetes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e299 (44%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e104 (34.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e195 (65.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e74 (24.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e164 (54.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e92 (30.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e166 (55.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eChronic lung disease\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e114 (17%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e22 (19.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e92 (80.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e16 (14%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e77 (67.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e19 (16.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e78 (68.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eChronic liver disease\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e18 (2.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e2 (11.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e16 (88.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e1 (5.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e13 (72.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e1 (5.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e14 (77.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eMalignancy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e9 (1.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e1(11.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e8 (88.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e7 (77.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e1 (11.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e7 (77.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eCAD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e94 (13.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e42 (44.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e52 (55.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e30 (31.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e44 (46.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e37 (39.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e44 (46.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.761467889908257%\" valign=\"top\"\u003e\n \u003cp\u003eCKD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e96 (14.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e54 (56.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.162079510703364%\" valign=\"top\"\u003e\n \u003cp\u003e42 (43.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.63302752293578%\" valign=\"top\"\u003e\n \u003cp\u003e38 (39.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"9.785932721712538%\" valign=\"top\"\u003e\n \u003cp\u003e35 (36.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.385321100917432%\" valign=\"top\"\u003e\n \u003cp\u003e47 (49%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.53822629969419%\" valign=\"top\"\u003e\n \u003cp\u003e36 (37.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.409785932721713%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eRadiologists demonstrated a positive predictive value (PPV) of 71%, suggesting that out of the 2,345 individuals identified as positive by GeneXpert MTB/RIF, 1,665 individuals were confirmed positive by radiologists (refer to Table 2). In contrast, DecXpert exhibited a higher PPV of 88%, where out of the 2,345 individuals identified as positive by GeneXpert MTB/RIF, 2,064 were subsequently confirmed as positive by DecXpert. Regarding the negative predictive value (NPV), radiologists achieved an 83.9% NPV. This indicates that among the 2,018 individuals classified as negative by GeneXpert MTB/RIF, 1,695 were confirmed as negative by radiologists. Conversely, DecXpert exhibited a higher negative predictive value (NPV) of 85%, suggesting that of the 2,018 individuals classified as negative for TB by GeneXpert MTB/RIF, 1,716 were confirmed as negative by DecXpert (refer to Table 2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2: Performance evaluation of DecXpert against GeneXpert MTB/RIF and radiologists.\u0026nbsp;\u003c/strong\u003eThis table compares the performance of the DecXpert Computer-Aided Detection software against the gold standard GeneXpert MTB/RIF test and radiologists\u0026rsquo; interpretations. The true positive, true negative, positive predictive value (PPV), and negative predictive value (NPV) are reported for each method. The PPV indicates the probability that a positive test result is truly positive, while the NPV represents the probability that a negative test result is truly negative. MTB stands for \u003cem\u003eMycobacterium tuberculosis\u003c/em\u003e, the causative agent of tuberculosis and RIF for Right iliac fossa.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"602\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"24.750830564784053%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGene Xpert MTB/RIF\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.41528239202658%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eRadiologists\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDecXpert\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"24.750830564784053%\" valign=\"top\"\u003e\n \u003cp\u003eTrue positives\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e2,345\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.41528239202658%\" valign=\"top\"\u003e\n \u003cp\u003e1,665\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e2,064\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"24.750830564784053%\" valign=\"top\"\u003e\n \u003cp\u003ePPV (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e100\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.41528239202658%\" valign=\"top\"\u003e\n \u003cp\u003e71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e88\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"24.750830564784053%\" valign=\"top\"\u003e\n \u003cp\u003eTrue negatives\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e2,018\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.41528239202658%\" valign=\"top\"\u003e\n \u003cp\u003e1,695\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e1,716\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"24.750830564784053%\" valign=\"top\"\u003e\n \u003cp\u003eNPV (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e100\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.41528239202658%\" valign=\"top\"\u003e\n \u003cp\u003e83.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.916943521594686%\" valign=\"top\"\u003e\n \u003cp\u003e85\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eQuantitative assessment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe evaluated the performance of different models for the identification of TB using the DecXpert score as a primary predictor (refer to Figure 2). Initially, Model 1, which relies solely on DecXpert scores for TB detection, demonstrated an area under the receiver operating characteristic (ROC) curve (AUC) of 0.85 (95% CI: 0.82-0.87), indicating a reasonably strong predictive ability (refer to Table 3 and Figure 2). Model 2, which incorporated both the DecXpert score and symptom information, displayed an improved AUC of 0.88 (95% CI: 0.83-0.92). When patient demographic information (specifically age and gender) was integrated with DecXpert scores in Model 3, the AUC further increased to 0.91 (95% CI: 0.88-0.94) (refer to Table 3 and Figure 2), indicating an enhanced predictive performance compared to that of earlier models. Furthermore, the development of a composite Model 4, which integrates the DecXpert score, symptom incidence, age, and gender, resulted in a greater AUC of 0.95 (95% CI: 0.90-0.97) (refer to Table 3 and Figure 2).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eTable 3\u003c/strong\u003e: \u003cstrong\u003eSummarizes the performance of the DecXpert Computer-Aided Detection system with added patient demographics and clinical information\u003c/strong\u003e. It shows the Area Under the Receiver Operating Characteristic Curve (AUC) and corresponding 95% confidence intervals (CI) for different combinations of patient demographics and clinical data. The table outlines the AUC values achieved by DecXpert scores alone, DecXpert scores combined with symptom information, DecXpert scores combined with age and gender information, and DecXpert scores combined with symptom information, age, and gender. Higher AUC values indicate better discrimination between individuals with and without active tuberculosis. The integration of additional clinical data progressively enhances DecXpert\u0026apos;s diagnostic performance, as evidenced by increasing AUC values across the combinations.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"645\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.643410852713178%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eModels\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"59.689922480620154%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eComponents\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.937984496124031%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.728682170542635%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.643410852713178%\" valign=\"top\"\u003e\n \u003cp\u003eModel 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"59.689922480620154%\" valign=\"top\"\u003e\n \u003cp\u003eDecXpert scores\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.937984496124031%\" valign=\"top\"\u003e\n \u003cp\u003e0.85\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.728682170542635%\" valign=\"top\"\u003e\n \u003cp\u003e(0.82\u0026ndash;0.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.643410852713178%\" valign=\"top\"\u003e\n \u003cp\u003eModel 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"59.689922480620154%\" valign=\"top\"\u003e\n \u003cp\u003eDecXpert scores + Symptom information\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.937984496124031%\" valign=\"top\"\u003e\n \u003cp\u003e0.88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.728682170542635%\" valign=\"top\"\u003e\n \u003cp\u003e(0.83\u0026ndash;0.92)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.643410852713178%\" valign=\"top\"\u003e\n \u003cp\u003eModel 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"59.689922480620154%\" valign=\"top\"\u003e\n \u003cp\u003eDecXpert scores + Age + Gender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.937984496124031%\" valign=\"top\"\u003e\n \u003cp\u003e0.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.728682170542635%\" valign=\"top\"\u003e\n \u003cp\u003e(0.88\u0026ndash;0.94)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"13.643410852713178%\" valign=\"top\"\u003e\n \u003cp\u003eModel 4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"59.689922480620154%\" valign=\"top\"\u003e\n \u003cp\u003eDecXpert scores + Symptom information + Age + Gender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.937984496124031%\" valign=\"top\"\u003e\n \u003cp\u003e0.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.728682170542635%\" valign=\"top\"\u003e\n \u003cp\u003e(0.90\u0026ndash;0.97)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003ePerformance of the DecXpert Algorithm Against the Gold Standard Molecular Reference GeneXpert MTB/RIF\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNext, we assessed the performance of the proposed DecXpert model that has proven to be highly effective in the identification of TB cases from CXR images. The GeneXpert MTB/RIF test reports served as the established ground truth for evaluation. The DecXpert model, operating solely on CXR imaging data without incorporating patient symptoms, achieved an AUC of 0.85 (95% CI=0.82\u0026ndash;0.87), indicated by the blue ROC curve in Figure 3. However, upon inclusion of age and gender, the performance of the DecXpert model improved, yielding an AUC of 0.91 (95% CI=0.88\u0026ndash;0.94), indicated by the green ROC curve in Figure 3. These results indicate a high level of accuracy ranging between 85% and 91% relative to GeneXpert MTB/RIF test reports, which are considered the gold standard (refer to Figure 3), suggesting that DecXpert reports could potentially serve as a surrogate for GeneXpert MTB/RIF. Notably, when basic patient demographics, specifically age and gender, and patient symptoms such as cough, fever, hemoptysis and night sweats were omitted, the DecXpert model still demonstrated a strong 85% concordance with GeneXpert MTB/RIF testing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePerformance of the DecXpert Algorithm against 3 Board-Certified Radiologists\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnalysing the overall cohort revealed that the DecXpert algorithm successfully identified 2,064 TB patients (88%) out of the total 2,345 GeneXpert MTB/RIF-confirmed positive TB patients, whereas the radiologists identified 1,665 TB patients (71%). Thus, using the DecXpert algorithm increased the overall TB case detection rate by approximately 1.23 times compared to radiologists. Examining the performance of the three board-certified radiologists as illustrated in Figure 4, the first radiologist achieved an AUC of 0.79 (95% CI: 0.74\u0026ndash;0.84), the second radiologist achieved an AUC of 0.72 (95% CI: 0.67\u0026ndash;0.76), and the third radiologist achieved an AUC of 0.75 (95% CI: 0.71\u0026ndash;0.78). Notably, each radiologist\u0026apos;s sensitivity/specificity point fell outside the 95% CI space of the ROC curve of the DecXpert model, indicating that their identification performance was inferior to that of the DecXpert model (Figure 4). Furthermore, within the unproductive cough and comorbidity subgroup, there was a considerable improvement in the TB case detection rate, according to DecXpert, which detected 303 patients (71.2%), whereas radiologists were able to detect only 244 patients (57.4%). This observation emphasises the enhanced performance of the DecXpert algorithm compared to that of radiologists, particularly within this subgroup, demonstrating its greater efficiency in identifying TB patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQualitative assessment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDecXpert went through a validation process that focused on visualizing CXRs and highlighting specific areas in the image that are important for DecXpert to make decisions when classifying TB cases. The assessment highlights the algorithm\u0026apos;s reliance on clinically relevant regions within the lung features extracted from CXRs of TB patients to guide its decision-making process. Figure 5 illustrates patient cases presenting highlighted important factors (identified regions) in patients with confirmed TB from the gold standard reference test GeneXpert MTB/RIF. The model relies on accurate visual information and does not consider misleading visual cues such as symbols, motion artifacts, embedded text or symbols, or imaging irregularities when making decisions. This finding demonstrated that DecXpert-related decision-making behaviour is primarily rooted in clinically relevant features.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEvaluation of DecXpert Algorithm\u0026apos;s Suitability for Deployment in Remote Isolated Regions with Offline and Online Functionality and Minimal Hardware Needs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSubsequently, we evaluated the suitability of integrating DecXpert into the pre-existing CXR workflows within primary healthcare facilities and diagnostic centres, particularly for deployment in geographically isolated regions of the nation where computational hardware capabilities are limited. To this end, we examined both online and offline iterations of the DecXpert software and incorporated them into current TB CXR workflows for the purpose of TB screening and diagnosis at primary healthcare facilities and diagnostic centres. The investigation focused on the implementation of DecXpert across seven distinct providers of digital chest X-ray machines, including GE Healthcare\u0026trade;, Siemens\u0026trade;, Philips Healthcare\u0026trade;, FujiFilm Medical Systems\u0026trade;, Shimadzu Corporation\u0026trade;, Toshiba Medical Systems\u0026trade;, and Hitachi Healthcare\u0026trade;. This examination encompassed varying computational hardware setups, ranging from 500 MB to 16 GB of RAM, and spanning different versions of the Windows operating system (2000, 7, 8, and 10).\u003c/p\u003e\n\u003cp\u003eAdditionally, the study investigated the compatibility of DecXpert with all five perspectives of CXR images\u0026mdash;posteroanterior (PA), anteroposterior (AP), lateral, decubitus, and oblique views. This assessment was conducted at six remote and geographically dispersed locations within the northern region of India. Moreover, we aimed to ascertain the ease of use of DecXpert by the existing X-ray technicians at these sites within their current CXR workflow. DecXpert demonstrated seamless integration and compatibility with all CXR images from the seven vendors, supporting TB screening and diagnostic workflows. Notably, it functioned effectively with basic computational hardware, such as systems with 500 MB RAM and running Windows 2000. Furthermore, the on-site technicians at these healthcare facilities were easily trained on a simple four-step process for processing CXR images (refer to Figure 6-a,b,c,d). DecXpert was made available in both offline and online configurations which had the same functionality, and Figure 6 (a,b,c,d) depicts this straightforward process, wherein both the CXR images and patient demographics are uploaded to the DecXpert software, this process culminates in the creation of a probable diagnostic report in PDF format, which can be disseminated in both electronic and printed forms for assessment by a physician.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur investigation aimed to evaluate the efficacy of the AI-based DecXpert software in detecting TB in a resource-limited setting with a high disease burden but without HIV co-infection. Notably, this represents the largest cohort study to date, comprising 4,363 participants who underwent GeneXpert MTB/RIF testing for comparison with both DecXpert and radiologist interpretations. DecXpert effectively identified 88% (2,064 out of 2,345) of TB cases diagnosed by GeneXpert MTB/RIF, showcasing its potential to reduce the necessity for expensive molecular tests.\u003c/p\u003e \u003cp\u003eThe current TB workflow in resource-limited nations suffers from severe diagnostic delays, with a significant proportion of cases detected at late disease stages. This delayed diagnosis poses a significant public health challenge, as it increases the risk of disease transmission and complications. The development of DecXpert, offering both online and offline deployment for automated CXR interpretation, represents a milestone in leveraging technological advancements for large-scale TB screening efforts. By enabling early and accurate identification of TB cases, DecXpert has the potential to revolutionize the existing diagnostic paradigm in resource-constrained settings. Furthermore, the use of DecXpert as a triage tool could enhance case identification and potentially reduce program expenses by optimizing the utilization of costly molecular testing resources such as Cartridge-Based Nucleic Acid Amplification Test (CBNAAT).\u003c/p\u003e \u003cp\u003eOur findings demonstrate that DecXpert achieves substantial diagnostic accuracy, with an area under the receiver operating characteristic curve (AUC) of 0.91 (95% confidence interval [CI]: 0.88\u0026ndash;0.94) when integrating patient age and gender (Model 3). The incorporation of demographic data facilitated the creation of personalized risk probabilities based on quantitative assessments of TB at the individual CXR level. This approach empowers frontline healthcare personnel with informed decision-making capabilities regarding individual testing prioritization. In settings with limited GeneXpert MTB/RIF testing resources, DecXpert can optimize testing utilization by reducing the number of cartridges used, offering significant benefits in healthcare facilities experiencing high patient burdens.\u003c/p\u003e \u003cp\u003eFurthermore, we explored the integration of DecXpert into existing CXR workflows within primary healthcare facilities and diagnostic centers, with a particular focus on deployment in geographically isolated regions with limited computational infrastructure. DecXpert demonstrated seamless integration and compatibility with CXR images from 7 different CXR vendors, supporting TB screening and diagnostic workflows across six remote locations in northern India, even on systems with basic computational hardware, such as those with 500 MB RAM and running Windows 2000. This highlights the versatility and adaptability of the software, making it suitable for deployment in resource-constrained settings with diverse hardware configurations.\u003c/p\u003e \u003cp\u003eNotably, on-site technicians at these healthcare facilities were easily trained on a simple four-step process for processing CXR images using DecXpert (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The software was made available in both offline and online configurations, offering the same functionality and thereby accommodating diverse operational environments. The streamlined process involved uploading CXR images and patient demographics, culminating in the generation of a probable diagnostic report in a portable document format (PDF) for physician assessment (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). This user-friendly interface and workflow integration underscore the potential for widespread adoption of DecXpert as a valuable screening and diagnostic tool for TB, particularly in resource-limited and geographically isolated regions.\u003c/p\u003e \u003cp\u003eA critical aspect of our study was the validation process, which emphasized DecXpert's reliance on clinically relevant regions within the lung features extracted from CXRs for decision-making. By visualizing the specific areas within CXRs that were crucial for classifying TB cases, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the study highlights the algorithm's ability to identify and prioritize the relevant radiological features associated with TB disease. Notably, the study found that DecXpert's decision-making process relies on accurate visual information and does not consider misleading visual cues, such as symbols, motion artifacts, embedded text or symbols, or imaging irregularities. The emphasis on clinically relevant regions within the lung features extracted from chest X-rays aligns with the established understanding of TB manifestations and the radiological patterns associated with the disease. By focusing on these specific areas, DecXpert can leverage the diagnostic information contained within the images, potentially enhancing its ability to accurately identify TB cases. These findings contribute to the interpretability and trustworthiness of the system, potentially paving the way for its wider adoption and integration into clinical workflows for TB screening and diagnosis.\u003c/p\u003e \u003cp\u003eWhile the utilization of digital X-rays has expanded the potential for TB diagnosis and garnered community interest, operational considerations, health communication strategies, resource availability, and pathway development for individuals not diagnosed with TB remain crucial elements for the effective implementation of such programs. It is essential to address these factors to ensure that the benefits of DecXpert and similar technologies are fully realized and translated into improved patient outcomes. This study is subject to limitations, including a reliance on GeneXpert MTB/RIF as the reference standard and a setting primarily within a facility-based context rather than a community setting. While GeneXpert MTB/RIF is widely recognized as the gold standard for TB diagnosis, future studies could explore the performance of DecXpert against other diagnostic modalities or in community-based settings to further validate its applicability across diverse scenarios.\u003c/p\u003e \u003cp\u003eNonetheless, the introduction of DecXpert, a deep learning-powered tool for TB screening, demonstrated strong performance in detecting TB cases and explicable decision-making behavior. This technological innovation aims to aid frontline healthcare workers in high-risk regions combating TB, a disease that continues to pose a significant global health challenge. By leveraging the power of artificial intelligence and leveraging its ability to provide interpretable and transparent predictions, DecXpert has the potential to revolutionize TB screening and diagnosis, particularly in resource-limited settings.\u003c/p\u003e \u003cp\u003eIn summary, our study underscores the potential of DecXpert as a valuable screening tool for TB, particularly in resource-limited settings. Its integration with demographic data shows promise for personalized risk assessment, offering insights that could enhance testing strategies and clinical workflows in the ongoing fight against TB. The robust diagnostic performance, adaptability to diverse healthcare settings, and explicable decision-making process contribute to the overall utility and trustworthiness of DecXpert. As we continue to grapple with the global burden of TB, innovative technologies like DecXpert represent a significant stride towards improving early detection, optimizing resource utilization, and ultimately saving lives.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003eStudy Design\u003c/h2\u003e\n \u003cp\u003eThis study included an extensive passive case-finding investigation of prospectively enrolled patients. This multicenter study was carried out across 12 primary health care (PHC) centres and one tertiary care (TC) centre in the northern Indian region. Patient enrollment occurred from January 2018 to January 2022 at the 12 PHC centres and from April 2022 to November 2023 at the one tertiary care centre. All patients who presented to the concerned centre with a history of fever, cough, expectoration, or constitutional symptoms for more than two weeks were included in the study. Those patients who had all the above symptoms but had nonproductive cough were subjected to bronchoalveolar lavage (BAL) for confirmation of diagnosis if the facility was available at that centre; otherwise, they were not included. The study excluded pregnant females, patients who declined to participate, individuals already undergoing TB treatment, those presenting with a nonproductive cough, and participants for whom BAL fluid samples were unavailable (refer to Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003ePatients with TB were identified as individuals aged 15 years and older who had both a CXR and a GeneXpert MTB/RIF test confirming pulmonary TB. Patients without TB were identified as individuals aged 15 years and older who underwent a CXR, with a negative GeneXpert MTB/RIF [\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e] test ruling out TB diagnosis.\u003c/p\u003e\n \u003cp\u003eThe diagnostic capabilities were evaluated by utilising the sensitivity (Se), specificity (Sp), and the count of false positives, false negatives, true positives, and true negatives (FP, FN, TP, and TN, respectively) in relation to the results of the GeneXpert MTB/RIF test.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eScoring Chest X-rays using DecXpert\u003c/h2\u003e\n \u003cp\u003eBy utilising outputs from its detection systems, the software treats them as descriptive features extracted from images. These features are then used to train a K-nearest neighbors classifier [\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e], allowing for the computation of a cumulative abnormality/severity score for each CXR, which ranges from 0 to 100. A higher score indicates a more pronounced abnormality, potentially indicating TB. Individuals with DecXpert scores of 50 or higher were recommended to undergo additional clinical assessment by their physicians (refer to Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eDataset and Patients Enrolled\u003c/h2\u003e\n \u003cp\u003eWe prospectively included all patients who met the inclusion criteria across 12 primary health care centres, along with individuals who sought care at the Pulmonary Medicine Department of a tertiary care centre in northern India. All participants underwent CXR and GeneXpert MTB/RIF testing as part of the screening process.\u003c/p\u003e\n \u003cp\u003eAmong the initial 4,495 participants screened, 4,364 individuals were recruited into our study. However, the study included 4,363 individuals, and one of the recruited individuals had an inconclusive GeneXpert MTB/RIF test result. Within this cohort, 2,345 CXR images were obtained from patients diagnosed with TB, while 2,018 CXR images were acquired from individuals without TB, as ascertained by their GeneXpert MTB/RIF test results, regarded as the reference gold standard.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eDeep Learning Algorithm\u003c/h2\u003e\n \u003cp\u003eWe used a previously constructed deep learning algorithm called DecXpert (version 1.1), which was trained on 9,876 chest X-ray images. Regarding the distribution of data used to train and validate this algorithm, there were 4,932 chest X-ray images from patients with TB and 4,944 chest X-ray images from patients without TB. The patients for the training, validation, and test datasets were selected randomly from the chest X-ray images dataset, constituting 80%, 10%, and 10% of the dataset, respectively. All CXRs that were input into the algorithm were resized to the dimensions of 224x224 pixels. The following guidelines or limitations were established for developing DecXpert algorithm, outlined through an indicator function: (a) ensuring a sensitivity exceeding 80%, (b) achieving a specificity of more than 75%, and (c) constraining the number of parameters to 2\u0026nbsp;million.\u003c/p\u003e\n \u003cp\u003eTo specify the general framework and specific attributes of a deep neural network customised for TB detection, we employed generative synthesis [\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e]. Using an optimal generator G, these designs were automatically found. Let A represent the generated architecture of the neural network based on a given set of initial configurations C, aimed at optimising a universal performance metric P [\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e]. This optimisation process is subject to specific constraints determined by an indicator function ϕ.\u003c/p\u003e\n \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e$$A=\\text{m}\\text{a}\\text{x}P\\left(A\\left(c\\right)\\right)\\text{ }\\text{s}\\text{u}\\text{c}\\text{h} \\text{t}\\text{h}\\text{a}\\text{t}\\text{ }{1}_{\\varphi }\\left(A\\left(c\\right)\\right)=1;\\forall c\\in C$$\u003c/div\u003e\n \u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003eThe aim was to find architectures that maximise performance while meeting predetermined criteria. First, the significant diversity of both macro and microarchitectures within the overall network architecture is a clear feature. This variation results from applying a meticulously tailored machine learning approach for TB case detection using CXR images, aiming to strike the best possible balance between efficiency and accuracy.\u003c/p\u003e\n \u003cp\u003eSecond, the network architecture mainly comprises of convolutions applied on a depth and pointwise basis. This use of lightweight design patterns illustrates how the machine-driven design exploration strategy can modify the deep neural network microarchitecture in response to the imposed architectural complexity constraints. For DecXpert to potentially be widely adopted, its effective and efficient architecture is especially important. This is because DecXpert frequently needs to be deployed on low-end, low-cost computing devices in resource-constrained regions affected by poverty and economic constraints [\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eThird, A notable aspect of the DecXpert network architecture includes visual attention condensers [\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e], recently introduced, these mechanisms are a type of efficient attention condenser. These visual attention condensers enhance representational ability, requiring less computational and architectural complexity. Finally, this architecture predicts a patient\u0026apos;s tuberculosis status as positive or negative.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003eModel training\u003c/h2\u003e\n \u003cp\u003eThrough the utilisation of stochastic gradient descent optimisation, the proposed neural network structure underwent training [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e]. The training process involved a learning rate of 0.001, a momentum value of 0.6, and a batch size of 12, spanning across 400 epochs. Furthermore, data augmentation techniques were applied, including horizontal flipping, random cropping (15% margin), random contrast shifting (10% margin), and random intensity shifting (10% margin), were applied during the training phase [\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e]. After that, each image underwent resizing to a dimension of 224 \u0026times; 224 pixels. Our experiments have shown that this specific size provides the highest level of performance, allowing for the conservation of important textural characteristics needed to differentiate individuals who are TB-positive from those who are TB-negative. No additional performance improvements were observed when using higher resolutions. The DecXpert deep neural network architecture underwent all construction, training, and evaluation procedures using the TensorFlow deep learning framework [\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eStatistical analysis:\u003c/h2\u003e\n \u003cp\u003eFor the statistical analysis, the results were scrutinised using DecXpert, GeneXpert MTB/RIF, and the evaluations of three certified radiologists. Regarding the demographic data, continuous variables were presented as the mean and standard deviation, while categorical variables were described in terms of numbers and percentages. A comparative analysis was conducted between the other methods and GeneXpert, which was considered as the gold standard. Additionally, a comparison was made between DecXpert and the radiologists\u0026apos; assessments. The sensitivity (Sn), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV) were calculated based on the positive and negative results obtained. The performance of these methods, including 95% confidence intervals (CI), was compared in differentiating between positive and negative patients for TB. The Sn, Sp, PPV, NPV, and their corresponding 95% CI were computed and reported. Receiver operating characteristic (ROC) curve analyses and area under curve (AUC) values, along with their respective 95% CI, were presented for the outcomes of positive patients versus negative. All statistical analyses were performed using R programming (version 4.1) [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e], and a statistical significance was determined with a two-tailed p-value below 0.05. Statistical significance was assessed with *P\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **P\u0026thinsp;\u0026lt;\u0026thinsp;0.01, ***P\u0026thinsp;\u0026lt;\u0026thinsp;0.001, and ****P\u0026thinsp;\u0026lt;\u0026thinsp;0.0001, indicating progressively higher significance levels.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003ePositive Predictive Value (PPV)\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eThe positive predictive value is the proportion of true positive cases among all the cases that tested positive. In other words, it measures the probability that a positive result actually indicates the presence of the condition.\u003c/p\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003ePPV = (true positives)/(true positives\u0026thinsp;+\u0026thinsp;false positives)\u003c/p\u003e\n \u003c/div\u003e\n \u003cp\u003e\u003cstrong\u003eNegative Predictive Value (NPV)\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eThe negative predictive value is the proportion of true negative cases among all the cases that tested negative. It measures the probability that a negative result actually indicates the absence of the condition.\u003c/p\u003e\n \u003cp\u003eNPV = (true negatives)/(true negatives\u0026thinsp;+\u0026thinsp;false negatives)\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eMolecular validation using GeneXpert MTB/RIF\u003c/h2\u003e\n \u003cp\u003eGeneXpert MTB/RIF [\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e] was employed to verify the findings derived from DecXpert, encompassing both TB positive and negative cases, thereby enhancing confidence in the results.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003eRadiological validation\u003c/h2\u003e\n \u003cp\u003eThree board-certified radiologists analysed and reported the patient\u0026rsquo;s chest X-rays. The first radiologist had more than 3 years of experience, the second had more than 5 years of experience, and the third had more than 6 years of experience. A majority vote (75%) among the three radiologists was required to classify a CXR as TB or normal. The radiological interpretations of the CXRs were subsequently compared to those of the DecXpert and GeneXpert MTB/RIF results.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003eEthical compliance\u003c/h2\u003e\n \u003cp\u003eThe research study obtained ethical clearance from the Institutional Ethics Committee (IEC) at Sanjay Gandhi Post Graduate Institute of Medical Sciences, with the approval code 2022-59-IMP-EXP-46, ensuring adherence to ethical standards. All data related to the study participants was de-identified and anonymized. Additionally, every participant verbally consented to be part of the study after being informed about its details, ensuring compliance with ethical norms set by the ethics committee. The study methodologies adhered to applicable guidelines and regulations. The radiologists and researchers involved in the study only had access to non-identifiable patient data for analysis purposes. In contrast, all identifiable patient screening and diagnostic information remained securely stored on an internally protected server accessible only through credential-based authentication.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data produced and analyzed in the present study can be obtained from the corresponding author upon request, subject to reasonable conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank Mr. Abid Mohsin Zaidi from the Department of Computer Science, Ambalika Institute of Management \u0026amp; Technology, Lucknow, and Ms. Rachna Shaw from the School of Physical Sciences, Indian Institute of Technology (IIT) Mandi, for their help with data analytics and preparing the figures for this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAuthor notes\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e#\u003c/sup\u003eThese authors contributed equally: Alok Nath, Zia Hashim\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors and Affiliations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1\u003c/sup\u003eDepartment of Pulmonary Medicine, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Raebareli Road, Lucknow, Uttar Pradesh, India 226014\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e2\u003c/sup\u003eIndian Institute of Technology Kanpur, Department of Electrical Engineering, Kalyanpur, Kanpur, Uttar Pradesh 208016, India\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e3\u003c/sup\u003eFaculty of Medicine, The University of Queensland, Translational Research Institute, 37 Kent Street, Brisbane, QLD, 4102, Australia\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e4\u003c/sup\u003eBaker Heart and Diabetes Institute, Melbourne, VIC, 3004, Australia.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.N., Z.H., S.S, P.A.P, N.M. and A.S. designed the study. N.M., M.S., S.S and A.S. \u0026nbsp;performed the analyses. A.S., N.M. and \u0026nbsp;M.S. wrote scripts and provided analysis tools. A.N., Z.H., P.A.P, S.S, N.M, M.S and A.S. provided critical intellectual content for the design of the study. A.N., Z.H., P.A.P, S.S, M.S and A.S. wrote the paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding author\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to Ankit Shukla ([email protected])\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHeslop, R., et al., \u003cem\u003eChanges in Host Cytokine Patterns of TB Patients with Different Bacterial Loads Detected Using 16S rRNA Analysis\u003c/em\u003e. PLOS ONE, 2016. 11(12): p. e0168272.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHO, \u003cem\u003eGlobal Tuberculosis Report 2023.\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHillson, R., \u003cem\u003eTuberculosis and diabetes\u003c/em\u003e. Practical Diabetes, 2017. 34(5): p. 149\u0026ndash;150.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan, A.J., et al., \u003cem\u003eEngaging the private sector to increase tuberculosis case detection: an impact evaluation study\u003c/em\u003e. Lancet Infect Dis, 2012. 12(8): p. 608\u0026ndash;16.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKranzer, K., et al., \u003cem\u003eThe benefits to communities and individuals of screening for active tuberculosis disease: a systematic review\u003c/em\u003e. Int J Tuberc Lung Dis, 2013. 17(4): p. 432\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMollura, D.J., et al., \u003cem\u003eWhite Paper Report of the RAD-AID Conference on International Radiology for Developing Countries: identifying challenges, opportunities, and strategies for imaging services in the developing world.\u003c/em\u003e J Am Coll Radiol, 2010. 7(7): p. 495\u0026ndash;500.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCandemir, S. and S. Antani, \u003cem\u003eA review on lung boundary detection in chest X-rays\u003c/em\u003e. Int J Comput Assist Radiol Surg, 2019. 14(4): p. 563\u0026ndash;576.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreuninger, M., et al., \u003cem\u003eDiagnostic accuracy of computer-aided detection of pulmonary tuberculosis in chest radiographs: a validation study from sub-Saharan Africa\u003c/em\u003e. PLoS One, 2014. 9(9): p. e106381.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMelendez, J., et al., \u003cem\u003eAn automated tuberculosis screening strategy combining X-ray-based computer-aided detection and clinical information\u003c/em\u003e. Sci Rep, 2016. 6: p. 25265.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMuyoyeta, M., et al., \u003cem\u003eThe sensitivity and specificity of using a computer aided diagnosis program for automatically scoring chest X-rays of presumptive TB patients compared with Xpert MTB/RIF in Lusaka Zambia\u003c/em\u003e. PLoS One, 2014. 9(4): p. e93757.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahman, M.T., et al., \u003cem\u003eAn evaluation of automated chest radiography reading software for tuberculosis screening among public- and private-sector patients\u003c/em\u003e. Eur Respir J, 2017. 49(5).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u003cem\u003eWHO consolidated guidelines on tuberculosis Module 2: Screening \u0026ndash; Systematic screening for tuberculosis disease.\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin, Z.Z., et al., \u003cem\u003eHow is Xpert MTB/RIF being implemented in 22 high tuberculosis burden countries?\u003c/em\u003e Eur Respir J, 2015. 45(2): p. 549\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSreeramareddy, C.T., et al., \u003cem\u003eDelays in diagnosis and treatment of pulmonary tuberculosis in India: a systematic review\u003c/em\u003e. Int J Tuberc Lung Dis, 2014. 18(3): p. 255\u0026ndash;266.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVonasek, B., et al., \u003cem\u003eScreening tests for active pulmonary tuberculosis in children\u003c/em\u003e. Cochrane Database of Systematic Reviews, 2021. 2021(10).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIoannidis, P., et al., \u003cem\u003eCepheid GeneXpert MTB/RIF assay for Mycobacterium tuberculosis detection and rifampin resistance identification in patients with substantial clinical indications of tuberculosis and smear-negative microscopy results\u003c/em\u003e. J Clin Microbiol, 2011. 49(8): p. 3068\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMucherino, A., P.J. Papajorgji, and P.M. Pardalos, \u003cem\u003ek-Nearest Neighbor Classification\u003c/em\u003e, in \u003cem\u003eData Mining in Agriculture\u003c/em\u003e, A. Mucherino, P.J. Papajorgji, and P.M. Pardalos, Editors. 2009, Springer New York: New York, NY. p. 83\u0026ndash;106.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElsken, T., J. Metzen, and F. Hutter, \u003cem\u003eNeural Architecture Search: A Survey\u003c/em\u003e. 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWong, A., et al., \u003cem\u003eFermiNets: Learning generative machines to generate efficient neural networks via generative synthesis\u003c/em\u003e. 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward, A., et al., \u003cem\u003eMobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications.\u003c/em\u003e 2017.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, L., Z.Q. Lin, and A. Wong, \u003cem\u003eCOVID-Net: a tailored deep convolutional neural network design for detection of COVID-19 cases from chest X-ray images\u003c/em\u003e. Scientific Reports, 2020. 10(1): p. 19549.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRuder, S., \u003cem\u003eAn overview of gradient descent optimization algorithms.\u003c/em\u003e 2016.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShorten, C. and T.M. Khoshgoftaar, \u003cem\u003eA survey on Image Data Augmentation for Deep Learning\u003c/em\u003e. Journal of Big Data, 2019. 6(1): p. 60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbadi, M., et al., \u003cem\u003eTensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems.\u003c/em\u003e 2016.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR Core Team, R., \u003cem\u003eR: A language and environment for statistical computing.\u003c/em\u003e 2013.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Tuberculosis (TB), Computer-Aided Detection (CAD), Deep convolutional neural networks (CNN or DCNN), Tuberculosis Screening, Radiology","lastPublishedDoi":"10.21203/rs.3.rs-4377653/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4377653/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTuberculosis (TB) is the leading cause of mortality among infectious diseases globally. Effectively managing TB requires early identification of high-risk individuals. Resource-constrained settings often lack skilled professionals for interpreting chest X-rays (CXRs) used in TB diagnosis. To address this challenge, we developed “DecXpert” a novel Computer-Aided Detection (CAD) software solution based on deep neural networks for early TB diagnosis from CXRs, aiming to detect subtle abnormalities that may be overlooked by human interpretation alone.\u003c/p\u003e\n\u003cp\u003eThis study was conducted on the largest cohort size to date, where the performance of a CAD software (DecXpert) was validated against the gold standard molecular diagnostic technique, GeneXpert MTB/RIF, analyzing data from 4,363 individuals across 12 primary health care centers and one tertiary hospital in North India. DecXpert demonstrated 88% sensitivity (95% CI: 0.85-0.93) and 85% specificity (95% CI: 0.82-0.91) for active TB detection. Incorporating demographics, DecXpert achieved an area under the curve of 0.91 (95% CI: 0.88-0.94), indicating robust diagnostic performance.\u003c/p\u003e\n\u003cp\u003eOur findings establish DecXpert's potential as an accurate, efficient AI solution for early identification of active TB cases. Deployed as a screening tool in resource-limited settings, DecXpert could enable identifying high-risk individuals and facilitate effective TB management where skilled radiological interpretation is limited.\u003c/p\u003e","manuscriptTitle":"Assessing Diagnostic Accuracy and Viability of AI-Assisted Tuberculosis Detection in Northern Indian Healthcare Facilities: A Multicenter Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-14 20:56:45","doi":"10.21203/rs.3.rs-4377653/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-06-11T06:46:19+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-06-10T10:11:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"158376139649340213009323720926640631569","date":"2024-06-10T10:09:27+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-26T05:03:53+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"96579352550985680492465971610335726106","date":"2024-05-23T12:10:52+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-05-23T08:24:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-05-23T08:10:29+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-05-23T07:20:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-05-23T07:13:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-05-06T14:45:54+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a1e9275b-477a-4f89-97b6-893d9f445f5c","owner":[],"postedDate":"May 14th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":31766690,"name":"Health sciences/Health care/Medical imaging/Radiography"},{"id":31766691,"name":"Health sciences/Medical research/Translational research"},{"id":31766692,"name":"Biological sciences/Computational biology and bioinformatics/Image processing"},{"id":31766694,"name":"Biological sciences/Computational biology and bioinformatics/Software"},{"id":31766696,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"}],"tags":[],"updatedAt":"2024-09-09T16:19:03+00:00","versionOfRecord":{"articleIdentity":"rs-4377653","link":"https://doi.org/10.1038/s41598-024-71346-x","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2024-09-05 15:57:18","publishedOnDateReadable":"September 5th, 2024"},"versionCreatedAt":"2024-05-14 20:56:45","video":"","vorDoi":"10.1038/s41598-024-71346-x","vorDoiUrl":"https://doi.org/10.1038/s41598-024-71346-x","workflowStages":[]},"version":"v1","identity":"rs-4377653","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4377653","identity":"rs-4377653","version":["v1"]},"buildId":"cBFmMYwuxLRRLfASyISRj","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}