Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa750
Jemima Hegermann, F. Baty, M. Brutsche
Introduction: Electronic medical records constitutes an important source of information for large scale health care quality studies. Text mining can be applied to automate knowledge extraction from unstructured data included in medical records and generate quality indicators applicable for medical documentation. The objective of the current study was to apply text mining methodology for the analysis of polysomnographic medical reports in order to assess the precision and the inter-physician variability in the diagnostic evaluation of sleep-disordered breathing. Material and Methods: Polysomnography reports of 243 sleep studies scored by 9 trained sleep physicians from the respiratory sleep team of the Sleep Center St. Gallen were analyzed using a text mining approach. After filtering for meaningful discriminating terms, a term-document matrix was generated and analyzed using correspondence analysis. Results: Diagnosis descriptions from the 243 polysomnography reports provided 49 discriminating terms which were used in our analysis (Fig. 1A). Patterns in the usage of these terms allowed for the characterization of the patients’ type of disease (Fig. 1B), disease severity (Fig. 1C) and inter-rater homogeneity (Fig. 1D). Conclusion: Text mining could be used to optimize the quality, as well as the precision and homogeneity of medical reporting of diagnostic procedures - here exemplified with polysomnography.
{"title":"Text mining - a tool for quality assessment of medical records","authors":"Jemima Hegermann, F. Baty, M. Brutsche","doi":"10.1183/13993003.congress-2019.pa750","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa750","url":null,"abstract":"Introduction: Electronic medical records constitutes an important source of information for large scale health care quality studies. Text mining can be applied to automate knowledge extraction from unstructured data included in medical records and generate quality indicators applicable for medical documentation. The objective of the current study was to apply text mining methodology for the analysis of polysomnographic medical reports in order to assess the precision and the inter-physician variability in the diagnostic evaluation of sleep-disordered breathing. Material and Methods: Polysomnography reports of 243 sleep studies scored by 9 trained sleep physicians from the respiratory sleep team of the Sleep Center St. Gallen were analyzed using a text mining approach. After filtering for meaningful discriminating terms, a term-document matrix was generated and analyzed using correspondence analysis. Results: Diagnosis descriptions from the 243 polysomnography reports provided 49 discriminating terms which were used in our analysis (Fig. 1A). Patterns in the usage of these terms allowed for the characterization of the patients’ type of disease (Fig. 1B), disease severity (Fig. 1C) and inter-rater homogeneity (Fig. 1D). Conclusion: Text mining could be used to optimize the quality, as well as the precision and homogeneity of medical reporting of diagnostic procedures - here exemplified with polysomnography.","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"482 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133836482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa749
J. Tolnai, K. Kapus, M. Draskóczy, F. Bari, F. Peták, Z. Novak
{"title":"Teleauscultation: an innovative initiative to categorize and analyse lung sounds","authors":"J. Tolnai, K. Kapus, M. Draskóczy, F. Bari, F. Peták, Z. Novak","doi":"10.1183/13993003.congress-2019.pa749","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa749","url":null,"abstract":"","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130214521","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa743
E. Talboom-Kamp, M. Holstege, M. Kasteleyn
{"title":"Does usage of eHealth improve symptoms of COPD patients?","authors":"E. Talboom-Kamp, M. Holstege, M. Kasteleyn","doi":"10.1183/13993003.congress-2019.pa743","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa743","url":null,"abstract":"","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121238663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa2233
C. Pinheiro, P. Viana, R. Amaral, T. Jacinto
Mobile apps can improve home measurements of pulmonary function via built-in phone sensors, (e.g. microphone). This could promote greater access to health interventions for patients with respiratory diseases, reducing the need of face-to-face visits. To evaluate the efficacy of a smartphone app (m-app) in registering sound records from a forced expiratory maneuver (FEM), according to the 2005 ATS/ERS acceptability criteria and to evaluate its usability. Observational cross-sectional study. Participants randomly performed a FEM in a spirometer and in a smartphone with an m-app, comparing unsatisfactory/unacceptable maneuvers and criteria using chi-square test. The questionnaire “System Usability Scale” assessed the usability, with values between 0 (completely dissatisfied) and 100 (completely satisfied). We’ve included 22 children ( Recording a FEM with the m-app is feasible and the errors are easily rectified. Moreover, it has good usability for the user, encouraging further development and improvement of this technology.
{"title":"Measurement of respiratory function with a mobile application: comparison with a conventional spirometer and evaluation of usability","authors":"C. Pinheiro, P. Viana, R. Amaral, T. Jacinto","doi":"10.1183/13993003.congress-2019.pa2233","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa2233","url":null,"abstract":"Mobile apps can improve home measurements of pulmonary function via built-in phone sensors, (e.g. microphone). This could promote greater access to health interventions for patients with respiratory diseases, reducing the need of face-to-face visits. To evaluate the efficacy of a smartphone app (m-app) in registering sound records from a forced expiratory maneuver (FEM), according to the 2005 ATS/ERS acceptability criteria and to evaluate its usability. Observational cross-sectional study. Participants randomly performed a FEM in a spirometer and in a smartphone with an m-app, comparing unsatisfactory/unacceptable maneuvers and criteria using chi-square test. The questionnaire “System Usability Scale” assessed the usability, with values between 0 (completely dissatisfied) and 100 (completely satisfied). We’ve included 22 children ( Recording a FEM with the m-app is feasible and the errors are easily rectified. Moreover, it has good usability for the user, encouraging further development and improvement of this technology.","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132088583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa2240
R. Mcgihon, Jingqin Zhu, M. Lougheed, C. Licskai, Samir Gupta, T. To
Background:Breathe is a mobile health (mHealth) application used for asthma self-management; however, its effects on the use of health care services remain poorly understood. Aim: To determine whether breathe reduces health service use in asthma patients who use the app compared to internal and external controls who do not. Methods: We conducted a pragmatic trial using data from the province of Ontario, Canada. Two groups of subjects who had participated in a randomized clinical trial (RCT) were included: an intervention group of adult asthma patients who used breathe for 12 months, and a control group of patients who did not use the app but received equivalent asthma care. A third, external control group of asthma patients were identified from an existing population cohort and matched to intervention subjects using a propensity-score approach. Generalized linear mixed models were used to determine changes in asthma-related hospitalizations, emergency department (ED) visits and outpatient physician visits over time. Results: A total of 641 individuals with asthma were included in the study. There were no statistically significant differences in the change of asthma-related hospitalization or ED visits between the intervention group and either the internal or external control group. However, compared to the external controls, the intervention group had a significantly greater decrease in the rate of physician office visits for asthma during the study (-10.8 per 100 vs. -4.3 per 100; p=0.018) and 1-year post (-19 per 100 vs. -8 per 100; p=0.032) compared to 1-year prior to the study. Conclusions: Use of breathe is associated with a significant decrease in the rate of physician office visits for asthma.
背景:Breathe是一个用于哮喘自我管理的移动健康(mHealth)应用程序;然而，它对使用保健服务的影响仍然知之甚少。目的:确定与不使用内部和外部控制的哮喘患者相比，使用该应用程序的哮喘患者呼吸是否减少了健康服务的使用。方法:我们使用来自加拿大安大略省的数据进行了一项实用试验。参与随机临床试验(RCT)的两组受试者被纳入:一组是使用呼吸12个月的成年哮喘患者的干预组，另一组是不使用该应用程序但接受同等哮喘护理的患者的对照组。第三，从现有人群队列中确定哮喘患者的外部对照组，并使用倾向评分方法与干预受试者相匹配。使用广义线性混合模型来确定哮喘相关住院、急诊科(ED)就诊和门诊医生就诊随时间的变化。结果:共有641名哮喘患者被纳入研究。干预组与内外对照组哮喘相关住院次数和ED就诊次数的变化均无统计学差异。然而，与外部对照相比，干预组在研究期间因哮喘就诊的比率显著下降(-10.8 / 100 vs -4.3 / 100;P =0.018)和1年职位(每100人中有19人vs.每100人中有-8人;P =0.032)，与研究前一年相比。结论:呼吸的使用与哮喘就诊率的显著降低有关。
{"title":"Effects of an asthma mHealth system on health services use: A pragmatic trial","authors":"R. Mcgihon, Jingqin Zhu, M. Lougheed, C. Licskai, Samir Gupta, T. To","doi":"10.1183/13993003.congress-2019.pa2240","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa2240","url":null,"abstract":"Background:Breathe is a mobile health (mHealth) application used for asthma self-management; however, its effects on the use of health care services remain poorly understood. Aim: To determine whether breathe reduces health service use in asthma patients who use the app compared to internal and external controls who do not. Methods: We conducted a pragmatic trial using data from the province of Ontario, Canada. Two groups of subjects who had participated in a randomized clinical trial (RCT) were included: an intervention group of adult asthma patients who used breathe for 12 months, and a control group of patients who did not use the app but received equivalent asthma care. A third, external control group of asthma patients were identified from an existing population cohort and matched to intervention subjects using a propensity-score approach. Generalized linear mixed models were used to determine changes in asthma-related hospitalizations, emergency department (ED) visits and outpatient physician visits over time. Results: A total of 641 individuals with asthma were included in the study. There were no statistically significant differences in the change of asthma-related hospitalization or ED visits between the intervention group and either the internal or external control group. However, compared to the external controls, the intervention group had a significantly greater decrease in the rate of physician office visits for asthma during the study (-10.8 per 100 vs. -4.3 per 100; p=0.018) and 1-year post (-19 per 100 vs. -8 per 100; p=0.032) compared to 1-year prior to the study. Conclusions: Use of breathe is associated with a significant decrease in the rate of physician office visits for asthma.","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126225803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa739
S. Malinin, E. Furman, E. Rocheva, V. Sokolovsky, G. Furman
{"title":"The home remote diagnostics of bronchial asthma in children with the using of telemedical system","authors":"S. Malinin, E. Furman, E. Rocheva, V. Sokolovsky, G. Furman","doi":"10.1183/13993003.congress-2019.pa739","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa739","url":null,"abstract":"","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114756082","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa742
M. I. Ahmed, H. Hickey, K. Calvert, L. Mogford, B. Elkington, Sarita Makam, K. Jain
Introduction: The paediatric long-term ventilation and complex respiratory care team is responsible for care of 231 children including 75 on long-term ventilation. These children need regular reviews, aimed at preventing infection and hospital admissions and maintaining optimal lung health. Due to their multiple issues, it is difficult for parents to bring them to hospital frequently for follow up appointments. We organised virtual clinics to review these children. A multidisciplinary team (respiratory consultants and trainees, specialist nurses, physiotherapists, psychologist, admin and speech therapists) attends these clinics. Aims: To assess the utility of virtual clinics in management of children with long-term ventilation and complex respiratory needs, and effect on parental and staff satisfaction Methods: We reviewed the data from our virtual clinic between May 2017- December 2018. Data on parental and staff satisfaction were collected. Results: During the 20 month period (May 2017- December 2018), 504 patients were discussed in weekly virtual clinics. Various aspects assessed in these clinics included pertinent microbiology, recent sleep studies and decisions on weaning home oxygen, and review of emergency management plans. Parents were updated on virtual clinic outcomes. Staff and parental satisfaction improved following the introduction of virtual clinic. Conclusion: Virtual clinic ensured timely reviews of our patient cohort and helped us to improve care of children in the community, this improved both parental and staff satisfaction.
{"title":"Virtual clinic for following up children with long term ventilation and complex respiratory needs","authors":"M. I. Ahmed, H. Hickey, K. Calvert, L. Mogford, B. Elkington, Sarita Makam, K. Jain","doi":"10.1183/13993003.congress-2019.pa742","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa742","url":null,"abstract":"Introduction: The paediatric long-term ventilation and complex respiratory care team is responsible for care of 231 children including 75 on long-term ventilation. These children need regular reviews, aimed at preventing infection and hospital admissions and maintaining optimal lung health. Due to their multiple issues, it is difficult for parents to bring them to hospital frequently for follow up appointments. We organised virtual clinics to review these children. A multidisciplinary team (respiratory consultants and trainees, specialist nurses, physiotherapists, psychologist, admin and speech therapists) attends these clinics. Aims: To assess the utility of virtual clinics in management of children with long-term ventilation and complex respiratory needs, and effect on parental and staff satisfaction Methods: We reviewed the data from our virtual clinic between May 2017- December 2018. Data on parental and staff satisfaction were collected. Results: During the 20 month period (May 2017- December 2018), 504 patients were discussed in weekly virtual clinics. Various aspects assessed in these clinics included pertinent microbiology, recent sleep studies and decisions on weaning home oxygen, and review of emergency management plans. Parents were updated on virtual clinic outcomes. Staff and parental satisfaction improved following the introduction of virtual clinic. Conclusion: Virtual clinic ensured timely reviews of our patient cohort and helped us to improve care of children in the community, this improved both parental and staff satisfaction.","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127230648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa735
Roberta Bodini, M. Grinovero, A. Corsico, M. Marvisi, G. Recchia, Salvatore D'Antonio, A. Vaghi
Introduction: Digital Therapeutics (DTx) is an emerging field of medicine that combines remote monitoring, behaviour modification and personalized intervention to improve health outcomes and health care efficiency. The idea behind DTx is that software can produce drug-like efficacy. DTx are evaluated in clinical trials, regulated by regulatory agencies and prescribed by doctors. The aim of the present review is to analyse the current stat of the art about DTx for asthma and COPD and to provide guidelines for their development. Methods: We searched PubMed, Clinical Trials.gov and Websites to identify DTx for asthma and COPD. Additional information on the development of DTx was requested from various DTx manufacturers. Interviews with research study investigators were performed to get their experience with these digital health technologies. Results: We have identified 5 DTx for asthma and COPD: Hailie (Adherium), Respiro (Amiko), Propeller (Propeller Health), BreatheSmart (Cohero Health) and CareTRx (Teva). All of them combine sensor devices, mApp fors for patients, and cloud-based software for healthcare professionals. Quality and levels of evidence supporting efficacy, safety and value of DTx are highly different and variable. Guidelines for the development of DTx have been developed. Conclusions: DTx have the potential to transform treatment of asthma and COPD. Evidence-based results supporting adoption are highly variable. We proposed guidelines to ensure that DTx manufacturers understand the evidence they need to show to meet the needs of the health and care system, patients, and users.
{"title":"Digital Therapy in the treatment of asthma and COPD - Epidemiology of development and use of an emerging health technology in Respiratory Medicine","authors":"Roberta Bodini, M. Grinovero, A. Corsico, M. Marvisi, G. Recchia, Salvatore D'Antonio, A. Vaghi","doi":"10.1183/13993003.congress-2019.pa735","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa735","url":null,"abstract":"Introduction: Digital Therapeutics (DTx) is an emerging field of medicine that combines remote monitoring, behaviour modification and personalized intervention to improve health outcomes and health care efficiency. The idea behind DTx is that software can produce drug-like efficacy. DTx are evaluated in clinical trials, regulated by regulatory agencies and prescribed by doctors. The aim of the present review is to analyse the current stat of the art about DTx for asthma and COPD and to provide guidelines for their development. Methods: We searched PubMed, Clinical Trials.gov and Websites to identify DTx for asthma and COPD. Additional information on the development of DTx was requested from various DTx manufacturers. Interviews with research study investigators were performed to get their experience with these digital health technologies. Results: We have identified 5 DTx for asthma and COPD: Hailie (Adherium), Respiro (Amiko), Propeller (Propeller Health), BreatheSmart (Cohero Health) and CareTRx (Teva). All of them combine sensor devices, mApp fors for patients, and cloud-based software for healthcare professionals. Quality and levels of evidence supporting efficacy, safety and value of DTx are highly different and variable. Guidelines for the development of DTx have been developed. Conclusions: DTx have the potential to transform treatment of asthma and COPD. Evidence-based results supporting adoption are highly variable. We proposed guidelines to ensure that DTx manufacturers understand the evidence they need to show to meet the needs of the health and care system, patients, and users.","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125623297","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa2227
N. Das, M. Topalovic, J. Raskin, J. Aerts, T. Troosters, W. Janssens
Background: Previous work demonstrated the possibility to automate pulmonary function test interpretation to diagnose respiratory disease using machine learning (ML). Since ML is a black box approach, understanding the reasoning behind a prediction is critical for generating trust, and is fundamental if one plans to take action based on a prediction. Objectives: We investigated a technique called locally interpretable model-agnostic explanation (LIME) to explain the predictions of a ML classifier that takes PFT data (spirometry, resistance, lung volumes, diffusion capacity) as input to suggest a diagnosis. Methods: We developed a ML classifier using 1400 historical cases. We tested our classifier in 50 randomly selected subjects with respiratory problems who completed PFT. An expert panel produced gold standard diagnoses from clinical, PFT and other test data. We applied LIME technique to generate interpretative explanations for each classifier prediction. Results: The classifier accuracy was 76%. LIME showed a high FEV1 Z-score (0.41±0.71) and TLCO%pred (92±15%) as the top explanatory feature for normal and asthma prediction while a low FEV1/FVC (50±11%) and RV%pred (70±15%) for COPD and ILD prediction respectively. Three predictions were incorrect when the top feature was negative (Fig 1b). Conclusion: By providing intuitive explanations, LIME builds trust for clinical application of a ML-based PFT interpretation algorithm.
{"title":"Explaining predictions of an automated pulmonary function test interpretation algorithm","authors":"N. Das, M. Topalovic, J. Raskin, J. Aerts, T. Troosters, W. Janssens","doi":"10.1183/13993003.congress-2019.pa2227","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa2227","url":null,"abstract":"Background: Previous work demonstrated the possibility to automate pulmonary function test interpretation to diagnose respiratory disease using machine learning (ML). Since ML is a black box approach, understanding the reasoning behind a prediction is critical for generating trust, and is fundamental if one plans to take action based on a prediction. Objectives: We investigated a technique called locally interpretable model-agnostic explanation (LIME) to explain the predictions of a ML classifier that takes PFT data (spirometry, resistance, lung volumes, diffusion capacity) as input to suggest a diagnosis. Methods: We developed a ML classifier using 1400 historical cases. We tested our classifier in 50 randomly selected subjects with respiratory problems who completed PFT. An expert panel produced gold standard diagnoses from clinical, PFT and other test data. We applied LIME technique to generate interpretative explanations for each classifier prediction. Results: The classifier accuracy was 76%. LIME showed a high FEV1 Z-score (0.41±0.71) and TLCO%pred (92±15%) as the top explanatory feature for normal and asthma prediction while a low FEV1/FVC (50±11%) and RV%pred (70±15%) for COPD and ILD prediction respectively. Three predictions were incorrect when the top feature was negative (Fig 1b). Conclusion: By providing intuitive explanations, LIME builds trust for clinical application of a ML-based PFT interpretation algorithm.","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"138 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125752099","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-28DOI: 10.1183/13993003.congress-2019.pa745
Rishi J Khusial, P. Honkoop, O. Usmani, Marcia Soares, M. Biddiscombe, S. Meah, M. Bonini, Antonios Lalas, J. Koopmans, J. Snoeck-Stroband, Steffen Ortmann, K. Moustakas, K. Votis, D. Tzovaras, K. Chung, S. Fowler, J. Sont
{"title":"myAirCoach: mHealth assisted self-management in patients with uncontrolled asthma, a randomized control trial","authors":"Rishi J Khusial, P. Honkoop, O. Usmani, Marcia Soares, M. Biddiscombe, S. Meah, M. Bonini, Antonios Lalas, J. Koopmans, J. Snoeck-Stroband, Steffen Ortmann, K. Moustakas, K. Votis, D. Tzovaras, K. Chung, S. Fowler, J. Sont","doi":"10.1183/13993003.congress-2019.pa745","DOIUrl":"https://doi.org/10.1183/13993003.congress-2019.pa745","url":null,"abstract":"","PeriodicalId":129661,"journal":{"name":"M-health/e-health","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121711510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：