TBscreen: A passive cough classifier for tuberculosis screening with a controlled dataset.
TBscreen: A passive cough classifier for tuberculosis screening with a controlled dataset.
复制标题
TBscreen：一种用于结核病筛查的被动咳嗽分类器，具有受控数据集。
DOI：
10.1126/sciadv.adi0282
复制
发表时间：
2024-01-05
期刊：
影响因子：
13.6
通讯作者：
Horne, David J.
中科院分区：
文献类型：
--
作者：
Sharma, Manuja;Nduba, Videlis;Njagi, Lilian N.;Murithi, Wilfred;Mwongera, Zipporah;Hawn, Thomas R.;Patel, Shwetak N.;Horne, David J.
最近的呼吸系统疾病筛查研究表明咳嗽分类器的性能很有希望，但模型训练和数据集质量的潜在偏差妨碍了可靠的结论。为了检查结核病（TB）咳嗽诊断特征，我们在内罗毕招募了患有肺结核的受试者（N = 149）和患有其他呼吸道疾病的对照受试者（N = 46）。我们收集了一个数据集，其中有33，000例被动咳嗽和1600例强迫咳嗽，这些咳嗽都是在具有相似人口统计学特征的受控环境中进行的。我们使用被动咳嗽尺度图的图像作为输入训练了一个基于ResNet 18的咳嗽分类器，并获得了0.70（±0.11 SD）的五倍交叉验证灵敏度。基于智能手机的模型在具有较高细菌负荷的受试者中具有更好的性能{受试者工作特征-曲线下面积（ROC-AUC）：0.87 [95%置信区间（CI）：0.87至0.88]，P < 0.001}或肺腔[ROC-AUC：0.89（95%CI：0.88至0.89），P < 0.001]。总的来说，我们的数据表明，被动咳嗽特征区分结核病和非结核病受试者，并与细菌负荷和疾病严重程度相关。被动性结核性咳嗽与细菌负荷有关，与其他呼吸道疾病的咳嗽不同。
Recent respiratory disease screening studies suggest promising performance of cough classifiers, but potential biases in model training and dataset quality preclude robust conclusions. To examine tuberculosis (TB) cough diagnostic features, we enrolled subjects with pulmonary TB (N = 149) and controls with other respiratory illnesses (N = 46) in Nairobi. We collected a dataset with 33,000 passive coughs and 1600 forced coughs in a controlled setting with similar demographics. We trained a ResNet18-based cough classifier using images of passive cough scalogram as input and obtained a fivefold cross-validation sensitivity of 0.70 (±0.11 SD). The smartphone-based model had better performance in subjects with higher bacterial load {receiver operating characteristic–area under the curve (ROC-AUC): 0.87 [95% confidence interval (CI): 0.87 to 0.88], P < 0.001} or lung cavities [ROC-AUC: 0.89 (95% CI: 0.88 to 0.89), P < 0.001]. Overall, our data suggest that passive cough features distinguish TB from non-TB subjects and are associated with bacterial burden and disease severity. Passive tuberculosis cough is associated with bacterial burden and differs from coughs in other respiratory conditions.
登录
查看更多内容
影响因子：
3.3
作者：
Arlot, Sylvain;Celisse, Alain
通讯作者：
Celisse, Alain
影响因子：
4.5
作者：
Altshuler, Ellery;Tannir, Bouchra;Jolicoeur, Gisele;Rudd, Matthew;Saleem, Cyrus;Cherabuddi, Kartikeya;Dore, Dominique Helene;Nagarsheth, Parav;Brew, Joe;Small, Peter M.;Morris, J. Glenn;Lapierre, Simon Grandjean
通讯作者：
Lapierre, Simon Grandjean
DOI：
10.1016/j.ijid.2021.02.107
发表时间：
2021-12
期刊：
International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases
影响因子：
--
作者：
Chakaya J;Khan M;Ntoumi F;Aklillu E;Fatima R;Mwaba P;Kapata N;Mfinanga S;Hasnain SE;Katoto PDMC;Bulabula ANH;Sam-Agudu NA;Nachega JB;Tiberi S;McHugh TD;Abubakar I;Zumla A
通讯作者：
Zumla A
DOI：
10.1016/s1473-3099(21)00387-x
发表时间：
2022-04
期刊：
The Lancet. Infectious diseases
影响因子：
--
作者：
Dhana A;Hamada Y;Kengne AP;Kerkhoff AD;Rangaka MX;Kredo T;Baddeley A;Miller C;Singh S;Hanifa Y;Grant AD;Fielding K;Affolabi D;Merle CS;Wachinou AP;Yoon C;Cattamanchi A;Hoffmann CJ;Martinson N;Mbu ET;Sander MS;Balcha TT;Skogmar S;Reeve BWP;Theron G;Ndlangalavu G;Modi S;Cavanaugh J;Swindells S;Chaisson RE;Ahmad Khan F;Howard AA;Wood R;Thit SS;Kyi MM;Hanson J;Drain PK;Shapiro AE;Kufa T;Churchyard G;Nguyen DT;Graviss EA;Bjerrum S;Johansen IS;Gersh JK;Horne DJ;LaCourse SM;Al-Darraji HAA;Kamarulzaman A;Kempker RR;Tukvadze N;Barr DA;Meintjes G;Maartens G
通讯作者：
Maartens G
影响因子：
4.6
作者：
Gabaldon-Figueira, Juan C.;Keen, Eric;Chaccour, Carlos
通讯作者：
Chaccour, Carlos