Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.03.002
Dennis Bontempi PhD , Osbert Zalay PhD , Danielle S Bitterman MD , Nicolai Birkbak PhD , Derek Shyr PhD , Fridolin Haugg MSc , Jack M Qian MD , Hannah Roberts MD , Subha Perni MD , Vasco Prudente MSc , Suraj Pai MSc , Andre Dekker PhD , Benjamin Haibe-Kains PhD , Christian Guthier PhD , Tracy Balboni MD , Laura Warren MD , Monica Krishan MD , Benjamin H Kann MD , Prof Charles Swanton MD , Prof Dirk De Ruysscher MD , Prof Hugo J W L Aerts PhD
<div><h3>Background</h3><div>As humans age at different rates, physical appearance can yield insights into biological age and physiological health more reliably than chronological age. In medicine, however, appearance is incorporated into medical judgements in a subjective and non-standardised way. In this study, we aimed to develop and validate FaceAge, a deep learning system to estimate biological age from easily obtainable and low-cost face photographs.</div></div><div><h3>Methods</h3><div>FaceAge was trained on data from 58 851 presumed healthy individuals aged 60 years or older: 56 304 individuals from the IMDb–Wiki dataset (training) and 2547 from the UTKFace dataset (initial validation). Clinical utility was evaluated on data from 6196 patients with cancer diagnoses from two institutions in the Netherlands and the USA: the MAASTRO, Harvard Thoracic, and Harvard Palliative cohorts FaceAge estimates in these cancer cohorts were compared with a non-cancerous reference cohort of 535 individuals. To assess the prognostic relevance of FaceAge, we performed Kaplan–Meier survival analysis and Cox modelling, adjusting for several clinical covariates. We also assessed the performance of FaceAge in patients with metastatic cancer receiving palliative treatment at the end of life by incorporating FaceAge into clinical prediction models. To evaluate whether FaceAge has the potential to be a biomarker for molecular ageing, we performed a gene-based analysis to assess its association with senescence genes.</div></div><div><h3>Findings</h3><div>FaceAge showed significant independent prognostic performance in various cancer types and stages. Looking older was correlated with worse overall survival (after adjusting for covariates per-decade hazard ratio [HR] 1·151, p=0·013 in a pan-cancer cohort of n=4906; 1·148, p=0·011 in a thoracic cohort of n=573; and 1·117, p=0·021 in a palliative cohort of n=717). We found that, on average, patients with cancer looked older than their chronological age (mean increase of 4·79 years with respect to non-cancerous reference cohort, p<0·0001). We found that FaceAge can improve physicians’ survival predictions in patients with incurable cancer receiving palliative treatments (from area under the curve 0·74 [95% CI 0·70–0·78] to 0·8 [0·76–0·83]; p<0·0001), highlighting the clinical use of the algorithm to support end-of-life decision making. FaceAge was also significantly associated with molecular mechanisms of senescence through gene analysis, whereas age was not.</div></div><div><h3>Interpretation</h3><div>Our results suggest that a deep learning model can estimate biological age from face photographs and thereby enhance survival prediction in patients with cancer. Further research, including validation in larger cohorts, is needed to verify these findings in patients with cancer and to establish whether the findings extend to patients with other diseases. Subject to further testing and validation, approaches such as
{"title":"FaceAge, a deep learning system to estimate biological age from face photographs to improve prognostication: a model development and validation study","authors":"Dennis Bontempi PhD , Osbert Zalay PhD , Danielle S Bitterman MD , Nicolai Birkbak PhD , Derek Shyr PhD , Fridolin Haugg MSc , Jack M Qian MD , Hannah Roberts MD , Subha Perni MD , Vasco Prudente MSc , Suraj Pai MSc , Andre Dekker PhD , Benjamin Haibe-Kains PhD , Christian Guthier PhD , Tracy Balboni MD , Laura Warren MD , Monica Krishan MD , Benjamin H Kann MD , Prof Charles Swanton MD , Prof Dirk De Ruysscher MD , Prof Hugo J W L Aerts PhD","doi":"10.1016/j.landig.2025.03.002","DOIUrl":"10.1016/j.landig.2025.03.002","url":null,"abstract":"<div><h3>Background</h3><div>As humans age at different rates, physical appearance can yield insights into biological age and physiological health more reliably than chronological age. In medicine, however, appearance is incorporated into medical judgements in a subjective and non-standardised way. In this study, we aimed to develop and validate FaceAge, a deep learning system to estimate biological age from easily obtainable and low-cost face photographs.</div></div><div><h3>Methods</h3><div>FaceAge was trained on data from 58 851 presumed healthy individuals aged 60 years or older: 56 304 individuals from the IMDb–Wiki dataset (training) and 2547 from the UTKFace dataset (initial validation). Clinical utility was evaluated on data from 6196 patients with cancer diagnoses from two institutions in the Netherlands and the USA: the MAASTRO, Harvard Thoracic, and Harvard Palliative cohorts FaceAge estimates in these cancer cohorts were compared with a non-cancerous reference cohort of 535 individuals. To assess the prognostic relevance of FaceAge, we performed Kaplan–Meier survival analysis and Cox modelling, adjusting for several clinical covariates. We also assessed the performance of FaceAge in patients with metastatic cancer receiving palliative treatment at the end of life by incorporating FaceAge into clinical prediction models. To evaluate whether FaceAge has the potential to be a biomarker for molecular ageing, we performed a gene-based analysis to assess its association with senescence genes.</div></div><div><h3>Findings</h3><div>FaceAge showed significant independent prognostic performance in various cancer types and stages. Looking older was correlated with worse overall survival (after adjusting for covariates per-decade hazard ratio [HR] 1·151, p=0·013 in a pan-cancer cohort of n=4906; 1·148, p=0·011 in a thoracic cohort of n=573; and 1·117, p=0·021 in a palliative cohort of n=717). We found that, on average, patients with cancer looked older than their chronological age (mean increase of 4·79 years with respect to non-cancerous reference cohort, p<0·0001). We found that FaceAge can improve physicians’ survival predictions in patients with incurable cancer receiving palliative treatments (from area under the curve 0·74 [95% CI 0·70–0·78] to 0·8 [0·76–0·83]; p<0·0001), highlighting the clinical use of the algorithm to support end-of-life decision making. FaceAge was also significantly associated with molecular mechanisms of senescence through gene analysis, whereas age was not.</div></div><div><h3>Interpretation</h3><div>Our results suggest that a deep learning model can estimate biological age from face photographs and thereby enhance survival prediction in patients with cancer. Further research, including validation in larger cohorts, is needed to verify these findings in patients with cancer and to establish whether the findings extend to patients with other diseases. Subject to further testing and validation, approaches such as","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100870"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144035203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.03.005
Shishir Rao DPhil , Yikuan Li DPhil , Mohammad Mamouei PhD , Gholamreza Salimi-Khorshidi DPhil , Malgorzata Wamil PhD , Milad Nazarzadeh DPhil , Christopher Yau DPhil , Gary S Collins PhD , Rod Jackson PhD , Andrew Vickers DPhil , Goodarz Danaei MD ScD , Kazem Rahimi DM FESC
Background
Although statistical models have been commonly used to identify patients at risk of cardiovascular disease for preventive therapy, these models tend to over-recommend therapy. Moreover, in populations with pre-existing diseases, the current approach is to indiscriminately treat all, as modelling in this context is currently inadequate. This study aimed to develop and validate the Transformer-based Risk assessment survival (TRisk) model, a novel deep learning model, for predicting 10-year risk of cardiovascular disease in both the primary prevention population and individuals with diabetes.
Methods
An open cohort of 3 million adults aged 25–84 years was identified using linked electronic health records from 291 general practices, for model development, and 98 general practices, for validation, across England from 1998 to 2015. Comparison against the QRISK3 score and a deep learning derivation of it was done. Additional analyses compared discriminatory performance in other age groups, by sex, and across categories of socioeconomic status.
Findings
TRisk showed superior discrimination (C index in the primary prevention population 0·910; 95% CI 0·906–0·913). TRisk’s performance was found to be less sensitive to population age range than the benchmark models and outperformed other models also in analyses stratified by age, sex, or socioeconomic status. All models were overall well calibrated. In decision curve analyses, TRisk showed a greater net benefit than benchmark models across the range of relevant thresholds. At the widely recommended 10% risk threshold and the higher 15% threshold, TRisk reduced both the total number of patients classified at high risk (by 20·6% and 34·6%, respectively) and the number of false negatives as compared with recommended strategies. TRisk similarly outperformed other models in patients with diabetes. Compared with the widely recommended treat-all policy approach for patients with diabetes, TRisk at a 10% risk threshold would lead to deselection of 24·3% of individuals, with a small fraction of false negatives (0·2% of the cohort).
Interpretation
TRisk enabled a more targeted selection of individuals at risk of cardiovascular disease in both the primary prevention population and cohorts with diabetes, compared with benchmark approaches. Incorporation of TRisk into routine care could potentially reduce the number of treatment-eligible patients by approximately one-third while preventing at least as many events as with currently adopted approaches.
Funding
None.
背景：虽然统计模型通常用于识别心血管疾病风险患者进行预防治疗，但这些模型倾向于过度推荐治疗。此外，在已有疾病的人群中，目前的做法是不分青红皂白地治疗所有人，因为在这方面的建模目前是不充分的。本研究旨在开发和验证基于transformer的风险评估生存（TRisk）模型，这是一种新的深度学习模型，用于预测初级预防人群和糖尿病患者10年心血管疾病风险。方法：从1998年至2015年，使用英格兰291个全科诊所的相关电子健康记录确定了300万名25-84岁成年人的开放队列，用于模型开发，98个全科诊所进行验证。与QRISK3分数进行了比较，并对其进行了深度学习推导。其他分析比较了其他年龄组、性别和不同社会经济地位类别的歧视性表现。结果：一级预防人群的风险指数（C指数）为0·910；95% ci 0.906 - 0.913)。研究发现，与基准模型相比，风险模型对人口年龄范围的敏感性较低，在按年龄、性别或社会经济地位分层的分析中，风险模型的表现也优于其他模型。所有模型总体上都得到了很好的校准。在决策曲线分析中，在相关阈值范围内，TRisk显示出比基准模型更大的净收益。在广泛推荐的10%风险阈值和更高的15%阈值下，与推荐的策略相比，TRisk降低了高风险患者的总数（分别减少20.6%和34.6%）和假阴性的数量。在糖尿病患者中，TRisk同样优于其他模型。与广泛推荐的针对糖尿病患者的全面治疗政策方法相比，风险阈值为10%的风险将导致24.3%的个体取消选择，并有一小部分假阴性（0.2%的队列）。解释：与基准方法相比，在初级预防人群和糖尿病人群中，风险使得更有针对性地选择有心血管疾病风险的个体。将风险纳入常规护理可能会使符合治疗条件的患者数量减少约三分之一，同时预防的事件至少与目前采用的方法一样多。资金:没有。
{"title":"Refined selection of individuals for preventive cardiovascular disease treatment with a transformer-based risk model","authors":"Shishir Rao DPhil , Yikuan Li DPhil , Mohammad Mamouei PhD , Gholamreza Salimi-Khorshidi DPhil , Malgorzata Wamil PhD , Milad Nazarzadeh DPhil , Christopher Yau DPhil , Gary S Collins PhD , Rod Jackson PhD , Andrew Vickers DPhil , Goodarz Danaei MD ScD , Kazem Rahimi DM FESC","doi":"10.1016/j.landig.2025.03.005","DOIUrl":"10.1016/j.landig.2025.03.005","url":null,"abstract":"<div><h3>Background</h3><div>Although statistical models have been commonly used to identify patients at risk of cardiovascular disease for preventive therapy, these models tend to over-recommend therapy. Moreover, in populations with pre-existing diseases, the current approach is to indiscriminately treat all, as modelling in this context is currently inadequate. This study aimed to develop and validate the Transformer-based Risk assessment survival (TRisk) model, a novel deep learning model, for predicting 10-year risk of cardiovascular disease in both the primary prevention population and individuals with diabetes.</div></div><div><h3>Methods</h3><div>An open cohort of 3 million adults aged 25–84 years was identified using linked electronic health records from 291 general practices, for model development, and 98 general practices, for validation, across England from 1998 to 2015. Comparison against the QRISK3 score and a deep learning derivation of it was done. Additional analyses compared discriminatory performance in other age groups, by sex, and across categories of socioeconomic status.</div></div><div><h3>Findings</h3><div>TRisk showed superior discrimination (C index in the primary prevention population 0·910; 95% CI 0·906–0·913). TRisk’s performance was found to be less sensitive to population age range than the benchmark models and outperformed other models also in analyses stratified by age, sex, or socioeconomic status. All models were overall well calibrated. In decision curve analyses, TRisk showed a greater net benefit than benchmark models across the range of relevant thresholds. At the widely recommended 10% risk threshold and the higher 15% threshold, TRisk reduced both the total number of patients classified at high risk (by 20·6% and 34·6%, respectively) and the number of false negatives as compared with recommended strategies. TRisk similarly outperformed other models in patients with diabetes. Compared with the widely recommended treat-all policy approach for patients with diabetes, TRisk at a 10% risk threshold would lead to deselection of 24·3% of individuals, with a small fraction of false negatives (0·2% of the cohort).</div></div><div><h3>Interpretation</h3><div>TRisk enabled a more targeted selection of individuals at risk of cardiovascular disease in both the primary prevention population and cohorts with diabetes, compared with benchmark approaches. Incorporation of TRisk into routine care could potentially reduce the number of treatment-eligible patients by approximately one-third while preventing at least as many events as with currently adopted approaches.</div></div><div><h3>Funding</h3><div>None.</div></div>","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100873"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144217272","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.01.014
Prof Israel Júnior Borges do Nascimento MD ClinPath , Hebatullah Mohamed Abdulazeem MD MSc , Ishanka Weerasekara PhD , Prof Jodie Marquez PhD , Lenny T Vasanthan PhD , Genevieve Deeken MSc , Prof Rosemary Morgan PhD , Heang-Lee Tan MPH , Isabel Yordi Aguirre PhD , Lasse Østeengaard MSc , Indunil Kularathne BSc , Natasha Azzopardi-Muscat PhD , Prof Robin van Kessel PhD , Edson Zangiacomi Martinez PhD , Govin Permanand PhD , David Novillo-Ortiz PhD MLIS
We evaluated the effects of digital health technologies (DHTs) on women's health, empowerment, and gender equality, using the scoping review method. Following a search across five databases and grey literature, we analysed 80 studies published up to Aug 18, 2023. The thematic appraisal and quantitative analysis found that DHTs positively affect women's access to health-care services, self-care, and tailored self-monitoring enabling the acquisition of health-related interventions. Use of these technologies is beneficial across various medical fields, including gynaecology, endocrinology, and psychiatry. DHTs also improve women's empowerment and gender equality by facilitating skills acquisition, health education, and social interaction, while allowing cost-effective health services. Overall, DHTs contribute to better health outcomes for women and support the UN Sustainable Development Goals by improving access to health care and financial literacy.
{"title":"Transforming women's health, empowerment, and gender equality with digital health: evidence-based policy and practice","authors":"Prof Israel Júnior Borges do Nascimento MD ClinPath , Hebatullah Mohamed Abdulazeem MD MSc , Ishanka Weerasekara PhD , Prof Jodie Marquez PhD , Lenny T Vasanthan PhD , Genevieve Deeken MSc , Prof Rosemary Morgan PhD , Heang-Lee Tan MPH , Isabel Yordi Aguirre PhD , Lasse Østeengaard MSc , Indunil Kularathne BSc , Natasha Azzopardi-Muscat PhD , Prof Robin van Kessel PhD , Edson Zangiacomi Martinez PhD , Govin Permanand PhD , David Novillo-Ortiz PhD MLIS","doi":"10.1016/j.landig.2025.01.014","DOIUrl":"10.1016/j.landig.2025.01.014","url":null,"abstract":"<div><div>We evaluated the effects of digital health technologies (DHTs) on women's health, empowerment, and gender equality, using the scoping review method. Following a search across five databases and grey literature, we analysed 80 studies published up to Aug 18, 2023. The thematic appraisal and quantitative analysis found that DHTs positively affect women's access to health-care services, self-care, and tailored self-monitoring enabling the acquisition of health-related interventions. Use of these technologies is beneficial across various medical fields, including gynaecology, endocrinology, and psychiatry. DHTs also improve women's empowerment and gender equality by facilitating skills acquisition, health education, and social interaction, while allowing cost-effective health services. Overall, DHTs contribute to better health outcomes for women and support the UN Sustainable Development Goals by improving access to health care and financial literacy.</div></div>","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100858"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144081422","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.03.001
Yuxuan Shi PhD , Zhen Li PhD , Li Wang PhD , Hong Wang PhD , Prof Xiaofeng Liu PhD , Dantong Gu MS , Xiao Chen MS , Xueli Liu PhD , Wentao Gong MS , Xiaowen Jiang MD , Wenquan Li MD , Yongdong Lin BS , Ke Liu MD , Deyan Luo MD , Tao Peng PhD , Xuemei Peng BS , Meimei Tong BS , Huizhen Zheng MD , Xuanchen Zhou MD , Jianrong Wu PhD , Prof Hongmeng Yu PhD
<div><h3>Background</h3><div>Nasopharyngeal carcinoma is highly curable when diagnosed early. However, the nasopharynx’s obscure anatomical position and the similarity of local imaging manifestations with those of other nasopharyngeal diseases often lead to diagnostic challenges, resulting in delayed or missed diagnoses. Our aim was to develop a deep learning algorithm to enhance an otolaryngologist’s diagnostic capabilities by differentiating between nasopharyngeal carcinoma, benign hyperplasia, and normal nasopharynx during endoscopic examination.</div></div><div><h3>Methods</h3><div>In this national, multicentre, model development and validation study, we developed a Swin Transformer-based Nasopharyngeal Diagnostic (STND) system to identify nasopharyngeal carcinoma, benign hyperplasia, and normal nasopharynx. STND was developed with 27 362 nasopharyngeal endoscopic images (10 693 biopsy-proven nasopharyngeal carcinoma, 7073 biopsy-proven benign hyperplasia, and 9596 normal nasopharynx) sourced from eight prominent nasopharyngeal carcinoma centres (stage 1), and externally validated with 1885 prospectively acquired images from ten comprehensive hospitals with a high incidence of nasopharyngeal carcinoma (stage 2). Furthermore, we did a fully crossed, multireader, multicase study involving four expert otolaryngologists from four regional leading nasopharyngeal carcinoma centres, and 24 general otolaryngologists from 24 geographically diverse primary hospitals. This study included 400 images to evaluate the diagnostic capabilities of the experts and general otolaryngologists both with and without the aid of the STND system in a real-world environment.</div></div><div><h3>Findings</h3><div>Endoscopic images used in the internal study (Jan 1, 2017, to Jan 31, 2023) were from 15 521 individuals (9033 [58·2%] men and 6488 [41·8%] women; mean age 47·6 years [IQR 38·4–56·8]). Images from 945 participants (538 [56·9%] men and 407 [43·1%] women; mean age 45·2 years [IQR 35·2– 55·2]) were used in the external validation. STND in the internal dataset discriminated normal nasopharynx images from abnormalities (benign hyperplasia and nasopharyngeal carcinoma) with an area under the curve (AUC) of 0·99 (95% CI 0·99–0·99) and malignant images (ie, nasopharyngeal carcinoma) from non-malignant images (ie, benign hyperplasia and normal nasopharynx) with an AUC of 0·99 (95% CI 0·98–0·99). In the external validation, the system had an AUC for the detection of nasopharyngeal carcinoma of 0·95 (95% CI 0·94–0·96), a sensitivity of 91·6% (95% CI 89·3–93·5), and a specificity of 86·1% (95% CI 84·1–87·9). In the multireader, multicase study, the artificial intelligence (AI)-assisted strategy enhanced otolaryngologists’ diagnostic accuracy by 7·9%, increasing from 83·4% (95% CI 80·1–86·7, without AI assistance) to 91·2% (95% CI 88·6–93·9, with AI assistance; p<0·0001) for primary care otolaryngologists. Reading time per image decreased with the aid of the AI model (mea
{"title":"Artificial intelligence-assisted detection of nasopharyngeal carcinoma on endoscopic images: a national, multicentre, model development and validation study","authors":"Yuxuan Shi PhD , Zhen Li PhD , Li Wang PhD , Hong Wang PhD , Prof Xiaofeng Liu PhD , Dantong Gu MS , Xiao Chen MS , Xueli Liu PhD , Wentao Gong MS , Xiaowen Jiang MD , Wenquan Li MD , Yongdong Lin BS , Ke Liu MD , Deyan Luo MD , Tao Peng PhD , Xuemei Peng BS , Meimei Tong BS , Huizhen Zheng MD , Xuanchen Zhou MD , Jianrong Wu PhD , Prof Hongmeng Yu PhD","doi":"10.1016/j.landig.2025.03.001","DOIUrl":"10.1016/j.landig.2025.03.001","url":null,"abstract":"<div><h3>Background</h3><div>Nasopharyngeal carcinoma is highly curable when diagnosed early. However, the nasopharynx’s obscure anatomical position and the similarity of local imaging manifestations with those of other nasopharyngeal diseases often lead to diagnostic challenges, resulting in delayed or missed diagnoses. Our aim was to develop a deep learning algorithm to enhance an otolaryngologist’s diagnostic capabilities by differentiating between nasopharyngeal carcinoma, benign hyperplasia, and normal nasopharynx during endoscopic examination.</div></div><div><h3>Methods</h3><div>In this national, multicentre, model development and validation study, we developed a Swin Transformer-based Nasopharyngeal Diagnostic (STND) system to identify nasopharyngeal carcinoma, benign hyperplasia, and normal nasopharynx. STND was developed with 27 362 nasopharyngeal endoscopic images (10 693 biopsy-proven nasopharyngeal carcinoma, 7073 biopsy-proven benign hyperplasia, and 9596 normal nasopharynx) sourced from eight prominent nasopharyngeal carcinoma centres (stage 1), and externally validated with 1885 prospectively acquired images from ten comprehensive hospitals with a high incidence of nasopharyngeal carcinoma (stage 2). Furthermore, we did a fully crossed, multireader, multicase study involving four expert otolaryngologists from four regional leading nasopharyngeal carcinoma centres, and 24 general otolaryngologists from 24 geographically diverse primary hospitals. This study included 400 images to evaluate the diagnostic capabilities of the experts and general otolaryngologists both with and without the aid of the STND system in a real-world environment.</div></div><div><h3>Findings</h3><div>Endoscopic images used in the internal study (Jan 1, 2017, to Jan 31, 2023) were from 15 521 individuals (9033 [58·2%] men and 6488 [41·8%] women; mean age 47·6 years [IQR 38·4–56·8]). Images from 945 participants (538 [56·9%] men and 407 [43·1%] women; mean age 45·2 years [IQR 35·2– 55·2]) were used in the external validation. STND in the internal dataset discriminated normal nasopharynx images from abnormalities (benign hyperplasia and nasopharyngeal carcinoma) with an area under the curve (AUC) of 0·99 (95% CI 0·99–0·99) and malignant images (ie, nasopharyngeal carcinoma) from non-malignant images (ie, benign hyperplasia and normal nasopharynx) with an AUC of 0·99 (95% CI 0·98–0·99). In the external validation, the system had an AUC for the detection of nasopharyngeal carcinoma of 0·95 (95% CI 0·94–0·96), a sensitivity of 91·6% (95% CI 89·3–93·5), and a specificity of 86·1% (95% CI 84·1–87·9). In the multireader, multicase study, the artificial intelligence (AI)-assisted strategy enhanced otolaryngologists’ diagnostic accuracy by 7·9%, increasing from 83·4% (95% CI 80·1–86·7, without AI assistance) to 91·2% (95% CI 88·6–93·9, with AI assistance; p<0·0001) for primary care otolaryngologists. Reading time per image decreased with the aid of the AI model (mea","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100869"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144340497","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.100889
The Lancet Digital Health
{"title":"Health insights from face photographs","authors":"The Lancet Digital Health","doi":"10.1016/j.landig.2025.100889","DOIUrl":"10.1016/j.landig.2025.100889","url":null,"abstract":"","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100889"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144250373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.02.007
Britte H E A ten Haaft MD , Boris V Janssen BSc , Esther Z Barsom MD PhD , Prof Wouter J K Hehenkamp MD PhD , Prof Mark I van Berge Henegouwen MD PhD , Prof Olivier R Busch MD PhD , Susan van Dieren PhD , Joris I Erdmann MD PhD , Wietse J Eshuis MD PhD , Suzanne S Gisbertz MD PhD , Prof Misha D P Luyer MD PhD , Olga C Damman PhD , Prof Martine C de Bruijne MD PhD , Prof Geert Kazemier MD PhD , Prof Marlies P Schijven MD PhD , Prof Marc G Besselink MD PhD
Background
Online video consultation between patients and health-care providers rapidly gained popularity during the COVID-19 pandemic. However, to our knowledge, there is no high-quality comparative evidence regarding patient satisfaction and quality of information recall with online video consultation and traditional face-to-face consultation. This lack of evidence is especially concerning in the most demanding consultations. We aimed to assess whether online video consultation between patients and surgeons before major abdominal surgery was non-inferior to face-to-face consultation in terms of patient satisfaction, and to assess effects on patient information recall.
Methods
This open-label, randomised, controlled, non-inferiority trial (VIDEOGO) was conducted at two hospitals (one academic and one regional) in the Netherlands. Adult patients (aged ≥18 years) who required consultation with a surgeon to discuss major abdominal surgery and were able and willing to interact via both online video and face-to-face consultation were eligible for inclusion; patients were excluded if they were unable or unwilling to start or maintain online video consultation. Eligible patients were randomly allocated (1:1) to online video or face-to-face consultation by the study coordinator, using a computer-generated, concealed, permuted-block randomisation method with varying block sizes (two, four, and six patients), stratified by study site. Masking of patients and health-care providers was not possible owing to the nature of the study. The primary outcomes were patient satisfaction (score 0–100; assessed for non-inferiority with a predefined margin of −10%) and information recall (score 0–11), both of which were assessed with online questionnaires and analysed in the intention-to-treat population for whom outcome data were available. Technical adverse events were assessed directly after the consultation as part of the satisfaction questionnaire. This trial is registered with the International Clinical Trial Registry Platform and the Central Committee on Research Involving Human Subjects registry, NL-OMON20092, and is complete.
Findings
Between Feb 13, 2021, and Oct 2, 2023, 120 patients were randomly assigned: 60 to online video consultation and 60 to face-to-face consultation. Outcome data were available for 57 patients in the online video consultation group (20 [35%] female and 37 [65%] male; median age 64·0 [54·5–72·5] years) and 55 patients in the face-to-face group (22 [40%] female and 33 [60%] male; median age 62·0 [56·0–70·0] years). The mean patient satisfaction score was 85·4 out of 100 (SD 12·3) in the online video consultation group and 85·2 (14·2) in the face-to-face group (mean difference 0·2, 95% CI −4·8 to 5·1), which was within the non-inferiority margin of −10% (pnon-inferiority<0·0001). The mean information recall score was 7·30 out of 11 (SD 1·60) in the
{"title":"Online video versus face-to-face preoperative consultation for major abdominal surgery (VIDEOGO): a multicentre, open-label, randomised, controlled, non-inferiority trial","authors":"Britte H E A ten Haaft MD , Boris V Janssen BSc , Esther Z Barsom MD PhD , Prof Wouter J K Hehenkamp MD PhD , Prof Mark I van Berge Henegouwen MD PhD , Prof Olivier R Busch MD PhD , Susan van Dieren PhD , Joris I Erdmann MD PhD , Wietse J Eshuis MD PhD , Suzanne S Gisbertz MD PhD , Prof Misha D P Luyer MD PhD , Olga C Damman PhD , Prof Martine C de Bruijne MD PhD , Prof Geert Kazemier MD PhD , Prof Marlies P Schijven MD PhD , Prof Marc G Besselink MD PhD","doi":"10.1016/j.landig.2025.02.007","DOIUrl":"10.1016/j.landig.2025.02.007","url":null,"abstract":"<div><h3>Background</h3><div>Online video consultation between patients and health-care providers rapidly gained popularity during the COVID-19 pandemic. However, to our knowledge, there is no high-quality comparative evidence regarding patient satisfaction and quality of information recall with online video consultation and traditional face-to-face consultation. This lack of evidence is especially concerning in the most demanding consultations. We aimed to assess whether online video consultation between patients and surgeons before major abdominal surgery was non-inferior to face-to-face consultation in terms of patient satisfaction, and to assess effects on patient information recall.</div></div><div><h3>Methods</h3><div>This open-label, randomised, controlled, non-inferiority trial (VIDEOGO) was conducted at two hospitals (one academic and one regional) in the Netherlands. Adult patients (aged ≥18 years) who required consultation with a surgeon to discuss major abdominal surgery and were able and willing to interact via both online video and face-to-face consultation were eligible for inclusion; patients were excluded if they were unable or unwilling to start or maintain online video consultation. Eligible patients were randomly allocated (1:1) to online video or face-to-face consultation by the study coordinator, using a computer-generated, concealed, permuted-block randomisation method with varying block sizes (two, four, and six patients), stratified by study site. Masking of patients and health-care providers was not possible owing to the nature of the study. The primary outcomes were patient satisfaction (score 0–100; assessed for non-inferiority with a predefined margin of −10%) and information recall (score 0–11), both of which were assessed with online questionnaires and analysed in the intention-to-treat population for whom outcome data were available. Technical adverse events were assessed directly after the consultation as part of the satisfaction questionnaire. This trial is registered with the International Clinical Trial Registry Platform and the Central Committee on Research Involving Human Subjects registry, NL-OMON20092, and is complete.</div></div><div><h3>Findings</h3><div>Between Feb 13, 2021, and Oct 2, 2023, 120 patients were randomly assigned: 60 to online video consultation and 60 to face-to-face consultation. Outcome data were available for 57 patients in the online video consultation group (20 [35%] female and 37 [65%] male; median age 64·0 [54·5–72·5] years) and 55 patients in the face-to-face group (22 [40%] female and 33 [60%] male; median age 62·0 [56·0–70·0] years). The mean patient satisfaction score was 85·4 out of 100 (SD 12·3) in the online video consultation group and 85·2 (14·2) in the face-to-face group (mean difference 0·2, 95% CI −4·8 to 5·1), which was within the non-inferiority margin of −10% (p<sub>non-inferiority</sub><0·0001). The mean information recall score was 7·30 out of 11 (SD 1·60) in the","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100867"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144310552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.100875
Lars Henrik Jensen
{"title":"Video in the clinic: advancing care for patients, professionals, and the planet","authors":"Lars Henrik Jensen","doi":"10.1016/j.landig.2025.100875","DOIUrl":"10.1016/j.landig.2025.100875","url":null,"abstract":"","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100875"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144310553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-01DOI: 10.1016/j.landig.2025.01.013
Prof Richard D Riley PhD , Joie Ensor PhD , Kym I E Snell PhD , Lucinda Archer PhD , Rebecca Whittle PhD , Paula Dhiman PhD , Joseph Alderman MBChB , Xiaoxuan Liu PhD , Laura Kirton MSc , Jay Manson-Whitton , Maarten van Smeden PhD , Prof Karel G Moons PhD , Prof Krishnarajah Nirantharakumar MD , Prof Jean-Baptiste Cazier PhD , Prof Alastair K Denniston PhD , Prof Ben Van Calster PhD , Prof Gary S Collins PhD
Rigorous study design and analytical standards are required to generate reliable findings in healthcare from artificial intelligence (AI) research. One crucial but often overlooked aspect is the determination of appropriate sample sizes for studies developing AI-based prediction models for individual diagnosis or prognosis. Specifically, the number of participants and outcome events required in datasets for model training and evaluation remains inadequately addressed. Most AI studies do not provide a rationale for their chosen sample sizes and frequently rely on datasets that are inadequate for training or evaluating a clinical prediction model. Among the ten principles of Good Machine Learning Practice established by the US Food and Drug Administration, the UK Medicines and Healthcare products Regulatory Agency, and Health Canada, guidance on sample size is directly relevant to at least three principles. To reinforce this recommendation, we outline seven reasons why inadequate sample size negatively affects model training, evaluation, and performance. Using a range of examples, we illustrate these issues and discuss the potentially harmful consequences for patient care and clinical adoption. Additionally, we address challenges associated with increasing sample sizes in AI research and highlight existing approaches and software for calculating the minimum sample sizes required for model training and evaluation.
{"title":"Importance of sample size on the quality and utility of AI-based prediction models for healthcare","authors":"Prof Richard D Riley PhD , Joie Ensor PhD , Kym I E Snell PhD , Lucinda Archer PhD , Rebecca Whittle PhD , Paula Dhiman PhD , Joseph Alderman MBChB , Xiaoxuan Liu PhD , Laura Kirton MSc , Jay Manson-Whitton , Maarten van Smeden PhD , Prof Karel G Moons PhD , Prof Krishnarajah Nirantharakumar MD , Prof Jean-Baptiste Cazier PhD , Prof Alastair K Denniston PhD , Prof Ben Van Calster PhD , Prof Gary S Collins PhD","doi":"10.1016/j.landig.2025.01.013","DOIUrl":"10.1016/j.landig.2025.01.013","url":null,"abstract":"<div><div>Rigorous study design and analytical standards are required to generate reliable findings in healthcare from artificial intelligence (AI) research. One crucial but often overlooked aspect is the determination of appropriate sample sizes for studies developing AI-based prediction models for individual diagnosis or prognosis. Specifically, the number of participants and outcome events required in datasets for model training and evaluation remains inadequately addressed. Most AI studies do not provide a rationale for their chosen sample sizes and frequently rely on datasets that are inadequate for training or evaluating a clinical prediction model. Among the ten principles of Good Machine Learning Practice established by the US Food and Drug Administration, the UK Medicines and Healthcare products Regulatory Agency, and Health Canada, guidance on sample size is directly relevant to at least three principles. To reinforce this recommendation, we outline seven reasons why inadequate sample size negatively affects model training, evaluation, and performance. Using a range of examples, we illustrate these issues and discuss the potentially harmful consequences for patient care and clinical adoption. Additionally, we address challenges associated with increasing sample sizes in AI research and highlight existing approaches and software for calculating the minimum sample sizes required for model training and evaluation.</div></div>","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 6","pages":"Article 100857"},"PeriodicalIF":23.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144217271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-05-01DOI: 10.1016/j.landig.2025.100877
{"title":"Correction to Lancet Digit Health 2024; 6: e386–95","authors":"","doi":"10.1016/j.landig.2025.100877","DOIUrl":"10.1016/j.landig.2025.100877","url":null,"abstract":"","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 5","pages":"Article 100877"},"PeriodicalIF":23.8,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144054723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-05-01DOI: 10.1016/j.landig.2024.12.007
Marco Gasparetto MD , Priya Narula MD , Charlotte Wong MBBS MSc , James Ashton MD PhD , Jochen Kammermeier MD PhD , Prof Marieke Pierik MD PhD , Prof Uri Kopylov MD , Prof Naila Arebi MD PhD
The use of digital health technology (DHT) is increasing worldwide. Clinical trials assessing available health tools for the management of patients with inflammatory bowel disease (IBD) are sparse, with limited evidence-based outcome data. In this umbrella review, we investigated the effectiveness of DHT in the care of patients with IBD and identified areas for future research following the Joanna Briggs Institute methodology. Systematic reviews published between January, 2012, and September, 2024, were identified through searches across nine databases (Ovid Embase, Ovid MEDLINE, ProQuest PsycINFO, Epistemonikos, Cochrane, Health Evidence, DoPHER, PROSPERO, and CINAHL via EBSCO), and the results were imported into Covidence software. Inclusion criteria included systematic reviews of randomised controlled trials (RCTs) involving patients of all ages with Crohn’s disease or ulcerative colitis, using DHT for diagnostics, treatment support, monitoring, self-management, or increasing participation in research studies, compared with standard care or alternative interventions. Outcomes included the efficacy and effectiveness of digital interventions, as reported in the studies. The primary outcome was clinical efficacy reported as one or more of the following: clinical response or remission, disease activity, flare-ups or relapses, and quality of life. Secondary outcomes included medication adherence, number of health-care visits, patient engagement (satisfaction and adherence or compliance with interventions), attendance for all terms of engagement, rate of interactions, knowledge improvement, psychological outcomes, and cost or cost–time effectiveness. The review protocol was registered in PROSPERO (registration number: CRD42023417525). AMSTAR-2 was used for methodological quality assessment. Nine relevant reviews were included, including five with meta-analyses comprising 13–19 RCTs in each review; four reviews were rated as high quality and five as critically low quality. DHT was not directly beneficial in achieving or maintaining clinical remission in IBD. In four trials, DHT use was associated with a reduced number of hospital attendances and increased treatment adherence, supporting its role as an adjuvant to standard clinical practice in IBD. Although current evidence from several RCTs and systematic reviews does not indicate better clinical outcomes with DHT in maintaining IBD remission and reducing relapse rates, DHT could be used as an adjuvant resource contributing towards treatment adherence and reducing hospital visits.
{"title":"Efficacy of digital health technologies in the management of inflammatory bowel disease: an umbrella review","authors":"Marco Gasparetto MD , Priya Narula MD , Charlotte Wong MBBS MSc , James Ashton MD PhD , Jochen Kammermeier MD PhD , Prof Marieke Pierik MD PhD , Prof Uri Kopylov MD , Prof Naila Arebi MD PhD","doi":"10.1016/j.landig.2024.12.007","DOIUrl":"10.1016/j.landig.2024.12.007","url":null,"abstract":"<div><div>The use of digital health technology (DHT) is increasing worldwide. Clinical trials assessing available health tools for the management of patients with inflammatory bowel disease (IBD) are sparse, with limited evidence-based outcome data. In this umbrella review, we investigated the effectiveness of DHT in the care of patients with IBD and identified areas for future research following the Joanna Briggs Institute methodology. Systematic reviews published between January, 2012, and September, 2024, were identified through searches across nine databases (Ovid Embase, Ovid MEDLINE, ProQuest PsycINFO, Epistemonikos, Cochrane, Health Evidence, DoPHER, PROSPERO, and CINAHL via EBSCO), and the results were imported into Covidence software. Inclusion criteria included systematic reviews of randomised controlled trials (RCTs) involving patients of all ages with Crohn’s disease or ulcerative colitis, using DHT for diagnostics, treatment support, monitoring, self-management, or increasing participation in research studies, compared with standard care or alternative interventions. Outcomes included the efficacy and effectiveness of digital interventions, as reported in the studies. The primary outcome was clinical efficacy reported as one or more of the following: clinical response or remission, disease activity, flare-ups or relapses, and quality of life. Secondary outcomes included medication adherence, number of health-care visits, patient engagement (satisfaction and adherence or compliance with interventions), attendance for all terms of engagement, rate of interactions, knowledge improvement, psychological outcomes, and cost or cost–time effectiveness. The review protocol was registered in PROSPERO (registration number: CRD42023417525). AMSTAR-2 was used for methodological quality assessment. Nine relevant reviews were included, including five with meta-analyses comprising 13–19 RCTs in each review; four reviews were rated as high quality and five as critically low quality. DHT was not directly beneficial in achieving or maintaining clinical remission in IBD. In four trials, DHT use was associated with a reduced number of hospital attendances and increased treatment adherence, supporting its role as an adjuvant to standard clinical practice in IBD. Although current evidence from several RCTs and systematic reviews does not indicate better clinical outcomes with DHT in maintaining IBD remission and reducing relapse rates, DHT could be used as an adjuvant resource contributing towards treatment adherence and reducing hospital visits.</div></div>","PeriodicalId":48534,"journal":{"name":"Lancet Digital Health","volume":"7 5","pages":"Article 100843"},"PeriodicalIF":23.8,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144081271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","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：