Ngeow Alvin Jia Hao, Moosa Aminath Shiwaza, Tan Mary Grace, Zou Lin, Goh Millie Ming Rong, Lim Gek Hsiang, Tagamolila Vina, Ereno Imelda, Durnford Jared Ryan, Cheung Samson Kei Him, Hong Nicholas Wei Jie, Soh Ser Yee, Tay Yih Yann, Chang Zi Ying, Ong Ruiheng, Tsang Li Ping Marianne, Yip Benny K L, Chia Kuok Wei, Yap Kelvin, Lim Ming Hwee, Ta Andy Wee An, Goh Han Leong, Yeo Cheo Lian, Chan Daisy Kwai Lin, Tan Ngiap Chuan
Department of Neonatal and Developmental Medicine, Singapore General Hospital, Singapore.
Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
JAMA Netw Open. 2024 Dec 2;7(12):e2450260. doi: 10.1001/jamanetworkopen.2024.50260.
This diagnostic study describes the merger of domain knowledge (Kramer principle of dermal advancement of icterus) with current machine learning (ML) techniques to create a novel tool for screening of neonatal jaundice (NNJ), which affects 60% of term and 80% of preterm infants.
This study aimed to develop and validate a smartphone-based ML app to predict bilirubin (SpB) levels in multiethnic neonates using skin color analysis.
DESIGN, SETTING, AND PARTICIPANTS: This diagnostic study was conducted between June 2022 and June 2024 at a tertiary hospital and 4 primary-care clinics in Singapore with a consecutive sample of neonates born at 35 or more weeks' gestation and within 21 days of birth.
The smartphone-based ML app captured skin images via the central aperture of a standardized color calibration sticker card from multiple regions of interest arranged in a cephalocaudal fashion, following the Kramer principle of dermal advancement of icterus. The ML model underwent iterative development and k-folds cross-validation, with performance assessed based on root mean squared error, Pearson correlation, and agreement with total serum bilirubin (TSB). The final ML model underwent temporal validation.
Linear correlation and statistical agreement between paired SpB and TSB; sensitivity and specificity for detection of TSB equal to or greater than 17mg/dL with SpB equal to or greater than 13 mg/dL were assessed.
The smartphone-based ML app was validated on 546 neonates (median [IQR] gestational age, 38.0 [35.0-41.0] weeks; 286 [52.4%] male; 315 [57.7%] Chinese, 35 [6.4%] Indian, 169 [31.0%] Malay, and 27 [4.9%] other ethnicities). Iterative development and cross-validation was performed on 352 neonates. The final ML model (ensembled gradient boosted trees) incorporated yellowness indicators from the forehead, sternum, and abdomen. Temporal validation on 194 neonates yielded a Pearson r of 0.84 (95% CI, 0.79-0.88; P < .001), 82% of data pairs within clinically acceptable limits of 3 mg/dL, sensitivity of 100%, specificity of 70%, positive predictive value of 10%, negative predictive value of 100%, positive likelihood ratio of 3.3, negative likelihood ratio of 0, and area under the receiver operating characteristic curve of 0.89 (95% CI, 0.82-0.96).
In this diagnostic study of a new smartphone-based ML app, there was good correlation and statistical agreement with TSB with sensitivity of 100%. The screening tool has the potential to be an NNJ screening tool, with treatment decisions based on TSB (reference standard). Further prospective studies are needed to establish the generalizability and cost-effectiveness of the screening tool in the clinical setting.
这项诊断性研究描述了领域知识（黄疸皮肤进展的克莱默原理）与当前机器学习（ML）技术的融合，以创建一种用于筛查新生儿黄疸（NNJ）的新型工具，该病影响60%的足月儿和80%的早产儿。
本研究旨在开发并验证一款基于智能手机的ML应用程序，通过皮肤颜色分析预测多种族新生儿的胆红素（SpB）水平。
设计、设置和参与者：这项诊断性研究于2022年6月至2024年6月在新加坡的一家三级医院和4家基层医疗诊所进行，连续纳入孕周为35周或以上且出生后21天内的新生儿样本。
基于智能手机的ML应用程序按照黄疸皮肤进展的克莱默原理，通过标准化颜色校准贴纸卡的中心孔径从以头尾方向排列的多个感兴趣区域捕获皮肤图像。ML模型进行了迭代开发和k折交叉验证，基于均方根误差、皮尔逊相关性以及与总血清胆红素（TSB）的一致性来评估性能。最终的ML模型进行了时间验证。
评估配对的SpB和TSB之间的线性相关性和统计一致性；以及当SpB等于或大于13mg/dL时检测TSB等于或大于17mg/dL的敏感性和特异性。
基于智能手机的ML应用程序在546名新生儿中得到验证（中位[四分位间距]孕周为38.0[35.0 - 41.0]周；286名[52.4%]为男性；315名[57.7%]为华裔，35名[6.4%]为印度裔，169名[31.0%]为马来裔，27名[4.9%]为其他种族）。对352名新生儿进行了迭代开发和交叉验证。最终的ML模型（集成梯度提升树）纳入了前额、胸骨和腹部的黄度指标。对194名新生儿的时间验证得出皮尔逊r为0.84（95%CI，0.79 - 0.88；P < .001），82%的数据对在3mg/dL的临床可接受范围内，敏感性为100%，特异性为70%，阳性预测值为10%，阴性预测值为100%，阳性似然比为3.3，阴性似然比为0，以及受试者工作特征曲线下面积为0.89（95%CI，0.82 - 0.96）。
在这项针对新型基于智能手机的ML应用程序的诊断性研究中，与TSB有良好的相关性和统计一致性，敏感性为100%。该筛查工具有可能成为一种NNJ筛查工具，治疗决策基于TSB（参考标准）。需要进一步的前瞻性研究来确定该筛查工具在临床环境中的可推广性和成本效益。
