Diagnostic performance of deep learning for infectious keratitis: a systematic review and meta-analysis

被引:2
作者
Ong, Zun Zheng [1 ]
Sadek, Youssef [2 ]
Qureshi, Riaz [3 ,4 ]
Liu, Su-Hsun [3 ,4 ]
Li, Tianjing [3 ,4 ]
Liu, Xiaoxuan [5 ,6 ,7 ]
Takwoingi, Yemisi [8 ]
Sounderajah, Viknesh [9 ]
Ashrafian, Hutan [9 ]
Ting, Daniel S. W. [10 ,11 ]
Mehta, Jodhbir S. [10 ,11 ]
Rauz, Saaeha [1 ,5 ]
Said, Dalia G. [12 ,13 ]
Dua, Harminder S. [12 ,13 ]
Burton, Matthew J. [14 ,15 ,16 ]
Ting, Darren S. J. [1 ,5 ,11 ,12 ]
机构
[1] Sandwell & West Birmingham NHS Trust, Birmingham & Midland Eye Ctr, Birmingham, England
[2] Univ Birmingham, Coll Med & Hlth, Birmingham Med Sch, Birmingham, England
[3] Univ Coloradom, Dept Epidemiol, Anschutz Med Campus, Aurora, CO USA
[4] Univ Coloradom, Dept Epidemiol, Anschutz Med Campus, Aurora, CO USA
[5] Univ Birmingham, Inst fl ammat & Ageing, Birmingham B15 2TT, England
[6] Univ Hosp Birmingham NHS Fdn Trust, Dept Ophthalmol, Birmingham, England
[7] Hlth Data Res UK, London, England
[8] Univ Birmingham, Dept Appl Hlth Sci, Birmingham, England
[9] Imperial Coll London, Inst Global Hlth Innovat, London, England
[10] Singapore Natl Eye Ctr, Singapore Eye Res Inst, Singapore, Singapore
[11] Duke NUS Med Sch, Ophthalmol & Visual Sci Acad Clin Program, Singapore, Singapore
[12] Univ Nottingham, Sch Med, Acad Ophthalmol, Nottingham, England
[13] Queens Med Ctr, Dept Ophthalmol, Nottingham, England
[14] London Sch Hyg & Trop Med, Int Ctr Eye Hlth, London, England
[15] UCL, Moorfields Eye Hosp NHS Fdn Trust, London, England
[16] UCL Inst Ophthalmol, London, England
基金
英国惠康基金;
关键词
Artificial fi cial intelligence; Corneal infection; Corneal ulcer; Deep learning; Infectious keratitis; Microbial keratitis; FUNGAL KERATITIS; BACTERIAL; ACCURACY; DISEASES; IMAGES;
D O I
10.1016/j.eclinm.2024.102887
中图分类号
R5 [内科学];
学科分类号
1002 ; 100201 ;
摘要
Background Infectious keratitis (IK) is the leading cause of corneal blindness globally. Deep learning (DL) is an emerging tool for medical diagnosis, though its value in IK is unclear. We aimed to assess the diagnostic accuracy of DL for IK and its comparative accuracy with ophthalmologists. Methods In this systematic review and meta-analysis, we searched EMBASE, MEDLINE, and clinical registries for studies related to DL for IK published between 1974 and July 16, 2024. We performed meta-analyses using bivariate models to estimate summary sensitivities and specificities. fi cities. This systematic review was registered with PROSPERO (CRD42022348596). Findings Of 963 studies identified, fi ed, 35 studies (136,401 corneal images from >56,011 patients) were included. Most studies had low risk of bias (68.6%) and low applicability concern (91.4%) in all domains of QUADAS-2, except the index test domain. Against the reference standard of expert consensus and/or microbiological results (seven external validation studies; 10,675 images), the summary estimates (95% CI) for sensitivity and specificity fi city of DL for IK were 86.2% (71.6-93.9) - 93.9) and 96.3% (91.5-98.5). - 98.5). From 28 internal validation studies (16,059 images), summary estimates for sensitivity and specificity fi city were 91.6% (86.8-94.8) - 94.8) and 90.7% (84.8-94.5). - 94.5). Based on seven studies (4007 images), DL and ophthalmologists had comparable summary sensitivity [89.2% (82.2-93.6) - 93.6) versus 82.2% (71.5-89.5); - 89.5); P = 0.20] and specificity fi city [(93.2% (85.5-97.0) - 97.0) versus 89.6% (78.8-95.2); - 95.2); P = 0.45]. Interpretation DL models may have good diagnostic accuracy for IK and comparable performance to ophthalmologists. These fi ndings should be interpreted with caution due to the image-based analysis that did not account for potential correlation within individuals, relatively homogeneous population studies, lack of pre-specification fi cation of DL thresholds, and limited external validation. Future studies should improve their reporting, data diversity, external validation, transparency, and explainability to increase the reliability and generalisability of DL models for clinical deployment.
引用
收藏
页数:17
相关论文
共 77 条
[61]   Clinical Characteristics and Outcomes of Fungal Keratitis in the United Kingdom 2011-2020: A 10-Year Study [J].
Ting, Darren Shu Jeng ;
Galal, Mohamed ;
Kulkarni, Bina ;
Elalfy, Mohamed S. ;
Lake, Damian ;
Hamada, Samer ;
Said, Dalia G. ;
Dua, Harminder S. .
JOURNAL OF FUNGI, 2021, 7 (11)
[62]   Risk Factors, Clinical Outcomes, and Prognostic Factors of Bacterial Keratitis: The Nottingham Infectious Keratitis Study [J].
Ting, Darren Shu Jeng ;
Cairns, Jessica ;
Gopal, Bhavesh P. ;
Ho, Charlotte Shan ;
Krstic, Lazar ;
Elsahn, Ahmad ;
Lister, Michelle ;
Said, Dalia G. ;
Dua, Harminder S. .
FRONTIERS IN MEDICINE, 2021, 8
[63]   12-year analysis of incidence, microbiological profiles and in vitro antimicrobial susceptibility of infectious keratitis: the Nottingham Infectious Keratitis Study [J].
Ting, Darren Shu Jeng ;
Ho, Charlotte Shan ;
Cairns, Jessica ;
Elsahn, Ahmad ;
Al-Aqaba, Mouhamed ;
Boswell, Tim ;
Said, Dalia G. ;
Dua, Harminder Singh .
BRITISH JOURNAL OF OPHTHALMOLOGY, 2021, 105 (03) :328-333
[64]   Infectious keratitis: an update on epidemiology, causative microorganisms, risk factors, and antimicrobial resistance [J].
Ting, Darren Shu Jeng ;
Ho, Charlotte Shan ;
Deshmukh, Rashmi ;
Said, Dalia G. ;
Dua, Harminder S. .
EYE, 2021, 35 (04) :1084-1101
[65]   Differentiation of Active Corneal Infections from Healed Scars Using Deep Learning [J].
Tiwari, Mo ;
Piech, Chris ;
Baitemirova, Medina ;
Prajna, Namperumalsamy, V ;
Srinivasan, Muthiah ;
Lalitha, Prajna ;
Villegas, Natacha ;
Balachandar, Niranjan ;
Chua, Janice T. ;
Redd, Travis ;
Lietman, Thomas M. ;
Thrun, Sebastian ;
Lin, Charles C. .
OPHTHALMOLOGY, 2022, 129 (02) :139-146
[66]   Deep learning model for extensive smartphone-based diagnosis and triage of cataracts and multiple corneal diseases [J].
Ueno, Yuta ;
Oda, Masahiro ;
Yamaguchi, Takefumi ;
Fukuoka, Hideki ;
Nejima, Ryohei ;
Kitaguchi, Yoshiyuki ;
Miyake, Masahiro ;
Akiyama, Masato ;
Miyata, Kazunori ;
Kashiwagi, Kenji ;
Maeda, Naoyuki ;
Shimazaki, Jun ;
Noma, Hisashi ;
Mori, Kensaku ;
Oshika, Tetsuro .
BRITISH JOURNAL OF OPHTHALMOLOGY, 2024, 108 (10) :1406-1413
[67]   Infectious corneal ulceration: a proposal for neglected tropical disease status [J].
Ung, Lawson ;
Acharya, Nisha R. ;
Agarwal, Tushar ;
Alfonso, Eduardo C. ;
Bagga, Bhupesh ;
Bispo, Paulo J. M. ;
Burton, Matthew J. ;
Dart, John K. G. ;
Thuy Doan ;
Fleiszig, Suzanne M. J. ;
Garg, Prashant ;
Gilmore, Michael S. ;
Gritz, David C. ;
Hazlett, Linda D. ;
Iovieno, Alfonso ;
Jhanji, Vishal ;
Kempen, John H. ;
Lee, Cecilia S. ;
Lietman, Thomas M. ;
Margolis, Todd P. ;
McLeod, Stephen D. ;
Mehta, Jod S. ;
Miller, Darlene ;
Pearlman, Eric ;
Prajna, Lalitha ;
Prajna, N. Venkatesh ;
Seitzman, Gerami D. ;
Shanbhag, Swapna S. ;
Sharma, Namrata ;
Sharma, Savitri ;
Srinivasan, Muthiah ;
Stapleton, Fiona ;
Tan, Donald T. H. ;
Tandon, Radhika ;
Taylor, Hugh R. ;
Tu, Elmer Y. ;
Tuli, Sonal S. ;
Vajpayee, Rasik B. ;
Van Gelder, Russell N. ;
Watson, Stephanie L. ;
Zegans, Michael E. ;
Chodosh, James .
BULLETIN OF THE WORLD HEALTH ORGANIZATION, 2019, 97 (12) :854-856
[68]   The persistent dilemma of microbial keratitis: Global burden, diagnosis, and antimicrobial resistance [J].
Ung, Lawson ;
Bispo, Paulo J. M. ;
Shanbhag, Swapna S. ;
Gilmore, Michael S. ;
Chodosh, James .
SURVEY OF OPHTHALMOLOGY, 2019, 64 (03) :255-271
[69]   Feasibility assessment of infectious keratitis depicted on slit-lamp and smartphone photographs using deep learning [J].
Wang, Lei ;
Chen, Kuan ;
Wen, Han ;
Zheng, Qinxiang ;
Chen, Yang ;
Pu, Jiantao ;
Chen, Wei .
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS, 2021, 155
[70]   Development and multi-center validation of machine learning model for early detection of fungal keratitis [J].
Wei, Zhenyu ;
Wang, Shigeng ;
Wang, Zhiqun ;
Zhang, Yang ;
Chen, Kexin ;
Gong, Lan ;
Li, Guigang ;
Zheng, Qinxiang ;
Zhang, Qin ;
He, Yan ;
Zhang, Qi ;
Chen, Di ;
Cao, Kai ;
Pang, Jinding ;
Zhang, Zijun ;
Wang, Leying ;
Ou, Zhonghong ;
Liang, Qingfeng .
EBIOMEDICINE, 2023, 88