1
|
Kumar S, Kumar H, Kumar G, Singh SP, Bijalwan A, Diwakar M. A methodical exploration of imaging modalities from dataset to detection through machine learning paradigms in prominent lung disease diagnosis: a review. BMC Med Imaging 2024; 24:30. [PMID: 38302883 PMCID: PMC10832080 DOI: 10.1186/s12880-024-01192-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/22/2023] [Accepted: 01/03/2024] [Indexed: 02/03/2024] Open
Abstract
BACKGROUND Lung diseases, both infectious and non-infectious, are the most prevalent cause of mortality overall in the world. Medical research has identified pneumonia, lung cancer, and Corona Virus Disease 2019 (COVID-19) as prominent lung diseases prioritized over others. Imaging modalities, including X-rays, computer tomography (CT) scans, magnetic resonance imaging (MRIs), positron emission tomography (PET) scans, and others, are primarily employed in medical assessments because they provide computed data that can be utilized as input datasets for computer-assisted diagnostic systems. Imaging datasets are used to develop and evaluate machine learning (ML) methods to analyze and predict prominent lung diseases. OBJECTIVE This review analyzes ML paradigms, imaging modalities' utilization, and recent developments for prominent lung diseases. Furthermore, the research also explores various datasets available publically that are being used for prominent lung diseases. METHODS The well-known databases of academic studies that have been subjected to peer review, namely ScienceDirect, arXiv, IEEE Xplore, MDPI, and many more, were used for the search of relevant articles. Applied keywords and combinations used to search procedures with primary considerations for review, such as pneumonia, lung cancer, COVID-19, various imaging modalities, ML, convolutional neural networks (CNNs), transfer learning, and ensemble learning. RESULTS This research finding indicates that X-ray datasets are preferred for detecting pneumonia, while CT scan datasets are predominantly favored for detecting lung cancer. Furthermore, in COVID-19 detection, X-ray datasets are prioritized over CT scan datasets. The analysis reveals that X-rays and CT scans have surpassed all other imaging techniques. It has been observed that using CNNs yields a high degree of accuracy and practicability in identifying prominent lung diseases. Transfer learning and ensemble learning are complementary techniques to CNNs to facilitate analysis. Furthermore, accuracy is the most favored metric for assessment.
Collapse
Affiliation(s)
- Sunil Kumar
- Department of Computer Engineering, J. C. Bose University of Science and Technology, YMCA, Faridabad, India
- Department of Information Technology, School of Engineering and Technology (UIET), CSJM University, Kanpur, India
| | - Harish Kumar
- Department of Computer Engineering, J. C. Bose University of Science and Technology, YMCA, Faridabad, India
| | - Gyanendra Kumar
- Department of Computer and Communication Engineering, Manipal University Jaipur, Jaipur, India
| | | | - Anchit Bijalwan
- Faculty of Electrical and Computer Engineering, Arba Minch University, Arba Minch, Ethiopia.
| | - Manoj Diwakar
- Department of Computer Science and Engineering, Graphic Era Deemed to Be University, Dehradun, India
| |
Collapse
|
2
|
Mostafa FA, Elrefaei LA, Fouda MM, Hossam A. A Survey on AI Techniques for Thoracic Diseases Diagnosis Using Medical Images. Diagnostics (Basel) 2022; 12:diagnostics12123034. [PMID: 36553041 PMCID: PMC9777249 DOI: 10.3390/diagnostics12123034] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/10/2022] [Revised: 11/20/2022] [Accepted: 11/22/2022] [Indexed: 12/12/2022] Open
Abstract
Thoracic diseases refer to disorders that affect the lungs, heart, and other parts of the rib cage, such as pneumonia, novel coronavirus disease (COVID-19), tuberculosis, cardiomegaly, and fracture. Millions of people die every year from thoracic diseases. Therefore, early detection of these diseases is essential and can save many lives. Earlier, only highly experienced radiologists examined thoracic diseases, but recent developments in image processing and deep learning techniques are opening the door for the automated detection of these diseases. In this paper, we present a comprehensive review including: types of thoracic diseases; examination types of thoracic images; image pre-processing; models of deep learning applied to the detection of thoracic diseases (e.g., pneumonia, COVID-19, edema, fibrosis, tuberculosis, chronic obstructive pulmonary disease (COPD), and lung cancer); transfer learning background knowledge; ensemble learning; and future initiatives for improving the efficacy of deep learning models in applications that detect thoracic diseases. Through this survey paper, researchers may be able to gain an overall and systematic knowledge of deep learning applications in medical thoracic images. The review investigates a performance comparison of various models and a comparison of various datasets.
Collapse
Affiliation(s)
- Fatma A. Mostafa
- Department of Electrical Engineering, Faculty of Engineering at Shoubra, Benha University, Cairo 11672, Egypt
| | - Lamiaa A. Elrefaei
- Department of Electrical Engineering, Faculty of Engineering at Shoubra, Benha University, Cairo 11672, Egypt
| | - Mostafa M. Fouda
- Department of Electrical and Computer Engineering, College of Science and Engineering, Idaho State University, Pocatello, ID 83209, USA
- Correspondence:
| | - Aya Hossam
- Department of Electrical Engineering, Faculty of Engineering at Shoubra, Benha University, Cairo 11672, Egypt
| |
Collapse
|
3
|
Srimahachota S, Krisanachinda A, Roongsangmanoon W, Sansanayudh N, Limpijankit T, Chandavimol M, Athisakul S, Siriyotha S, Rehani MM. Establishment of national diagnostic reference levels for percutaneous coronary interventions (PCIs) in Thailand. Phys Med 2022; 96:46-53. [PMID: 35219961 DOI: 10.1016/j.ejmp.2022.02.013] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 09/15/2021] [Revised: 02/07/2022] [Accepted: 02/12/2022] [Indexed: 12/31/2022] Open
Abstract
PURPOSE To establish national diagnostic reference levels (DRLs) for percutaneous coronary intervention (PCI) in Thailand for lesions of different complexity. METHODS Radiation dose quantity as kerma-area-product (KAP) and cumulative air-kerma at reference point (CAK) from 76 catheterization labs in 38 hospitals in PCI registry of Thailand was transferred online to central data management. Sixteen months data (May 2018 to August 2019) was analyzed. We also investigated role of different factors that influence radiation dose the most. RESULTS Analysis of 22,737 PCIs resulted in national DRLs for PCI of 91.3 Gy.cm2 (KAP) and 1360 mGy (CAK). The NDRLs for KAP for type C, B2, B1 and A lesions were 106.8, 82.6, 67.9, and 45.3 Gy.cm2 respectively and for CAK, 1705, 1247, 962, and 790 mGy respectively. Thus, as compared to lesion A, lesion C had more than double the dose and B2 had nearly 1.6 times and B1 had 1.2 times CAK. Our DRL values are lower than other Asian countries like Japan and Korea and are in the middle range of Western countries. University hospital had significantly higher dose than private or public hospital possibly because of higher load of complex procedures in university hospitals and trainees performing the procedures. Transradial approach showed lower doses than transfemoral approach. CONCLUSIONS This large multi-centric study established DRLs for PCIs which can act as reference for future studies. A hallmark of our study is establishment of reference levels for coronary lesions classified as per ACC/AHA and thus for different complexities.
Collapse
Affiliation(s)
- Suphot Srimahachota
- Cardiac Center and Division of Cardiovascular Medicine, King Chulalongkorn Memorial Hospital and Chulalongkorn University, Bangkok, Thailand.
| | - Anchali Krisanachinda
- Division of Nuclear Medicine, Department of Radiology, Faculty of Medicine, Chulalongkorn University, Bangkok 10330, Thailand
| | - Worawut Roongsangmanoon
- Division of Cardiology, Department of Medicine, Faculty of Medicine, Srinakharinwirot University, Nakornnayok, Thailand
| | - Nakarin Sansanayudh
- Division of Cardiology, Department of Medicine, Faculty of Medicine, Phramongkutklao Hostpital, Bangkok, Thailand
| | - Thosaphol Limpijankit
- Division of Cardiology, Department of Medicine, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
| | - Mann Chandavimol
- Division of Cardiology, Department of Medicine, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
| | - Siriporn Athisakul
- Cardiac Center and Division of Cardiovascular Medicine, King Chulalongkorn Memorial Hospital and Chulalongkorn University, Bangkok, Thailand
| | - Sukanya Siriyotha
- Department of Clinical Epidemiology and Biostatistics, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
| | - Madan M Rehani
- Radiology Department, Massachusetts General Hospital, Boston, MA, USA
| |
Collapse
|
4
|
The Reliability of Two- and Three-Dimensional Cephalometric Measurements: A CBCT Study. Diagnostics (Basel) 2021; 11:diagnostics11122292. [PMID: 34943528 PMCID: PMC8700671 DOI: 10.3390/diagnostics11122292] [Citation(s) in RCA: 12] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/18/2021] [Revised: 11/21/2021] [Accepted: 12/03/2021] [Indexed: 11/16/2022] Open
Abstract
Cephalometry is a standard diagnostic tool in orthodontic and orthognathic surgery fields. However, built-in magnification from the cephalometric machine produces double images from left- and right-side craniofacial structures on the film, which poses difficulty for accurate cephalometric tracing and measurements. The cone-beam computed tomography (CBCT) images not only allow three-dimensional (3D) analysis, but also enable the extraction of two-dimensional (2D) images without magnification. To evaluate the most reliable cephalometric analysis method, we extracted 2D lateral cephalometric images with and without magnification from twenty full-cranium CBCT datasets; images were extracted with magnification to mimic traditional lateral cephalograms. Cephalometric tracings were performed on the two types of extracted 2D lateral cephalograms and on the reconstructed 3D full cranium images by two examiners. The intra- and inter-examiner intraclass correlation coefficients (ICC) were compared between linear and angular parameters, as well as between CBCT datasets of adults and children. Our results showed that overall, tracing on 2D cephalometric images without magnification increased intra- and inter-examiner reliability, while 3D tracing reduced inter-examiner reliability. Angular parameters and children's images had the lowest inter- and intra-examiner ICCs compared with adult samples and linear parameters. In summary, using lateral cephalograms extracted from CBCT without magnification for tracing/analysis increased reliability. Special attention is needed when analyzing young patients' images and measuring angular parameters.
Collapse
|
5
|
Fishman MDC, Rehani MM. Monochromatic X-rays: The future of breast imaging. Eur J Radiol 2021; 144:109961. [PMID: 34562745 DOI: 10.1016/j.ejrad.2021.109961] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/22/2021] [Revised: 09/03/2021] [Accepted: 09/15/2021] [Indexed: 12/16/2022]
Abstract
PURPOSE To present details about the innovative and disruptive technology of monochromatic X-rays and its application to breast imaging. METHODS To analyze results of studies done using a prototype system for breast imaging that generates monochromatic X-rays through fluorescence emission. To assess signal-to-noise ratio (SNR) as a measure of image quality at different doses in breast phantoms of different sizes and review the comparison of parameters with a standard mammography system. RESULTS Monochromatic X-rays reduce the radiation dose per mammogram by a factor of 5 to 10 times. For phantom simulating thick breast (9 cm), the SNR for monochromatic system was 2.6 times higher and the dose 4.2 times lower than the respective values obtained with the conventional system within the same 5 mm × 5 mm square area of the 100% glandular step wedge. For the conventional broadband system to equal the SNR of the monochromatic system, it would require a dose of 19 mGy, 29 times higher than the dose delivered by the monochromatic system. Contrast-enhanced digital mammography with monochromatic X-rays is shown to provide a simpler and more effective technique at substantially lower radiation dose. CONCLUSIONS Lowering radiation dose by a factor of 5 to 10 while maintaining image quality implies a major reduction in total exposure from breast cancer screening and dramatically less risk of radiation-induced cancers in at-risk women. The high SNRs for very thick breast phantoms provide strong evidence that screening with lower breast compression is possible while maintaining image quality.
Collapse
Affiliation(s)
- Michael D C Fishman
- Boston Medical Center, Boston University School of Medicine, Boston, MA, USA
| | | |
Collapse
|
6
|
A Survey of Deep Learning for Lung Disease Detection on Medical Images: State-of-the-Art, Taxonomy, Issues and Future Directions. J Imaging 2020; 6:jimaging6120131. [PMID: 34460528 PMCID: PMC8321202 DOI: 10.3390/jimaging6120131] [Citation(s) in RCA: 20] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/24/2020] [Revised: 11/25/2020] [Accepted: 11/25/2020] [Indexed: 12/24/2022] Open
Abstract
The recent developments of deep learning support the identification and classification of lung diseases in medical images. Hence, numerous work on the detection of lung disease using deep learning can be found in the literature. This paper presents a survey of deep learning for lung disease detection in medical images. There has only been one survey paper published in the last five years regarding deep learning directed at lung diseases detection. However, their survey is lacking in the presentation of taxonomy and analysis of the trend of recent work. The objectives of this paper are to present a taxonomy of the state-of-the-art deep learning based lung disease detection systems, visualise the trends of recent work on the domain and identify the remaining issues and potential future directions in this domain. Ninety-eight articles published from 2016 to 2020 were considered in this survey. The taxonomy consists of seven attributes that are common in the surveyed articles: image types, features, data augmentation, types of deep learning algorithms, transfer learning, the ensemble of classifiers and types of lung diseases. The presented taxonomy could be used by other researchers to plan their research contributions and activities. The potential future direction suggested could further improve the efficiency and increase the number of deep learning aided lung disease detection applications.
Collapse
|
7
|
Caulley L, Ramaswami R, Longo DL, Phimister EG, Ingelfinger JR, Ropper AH, Stern K, Burke AE, Knoper KM, Seals JJJ, Müller DC, Drazen JM. A Look Forward - The Frontiers in Medicine Series. N Engl J Med 2018; 379:85-86. [PMID: 29972760 DOI: 10.1056/nejme1806084] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
|
8
|
ZARDO ERASMODEABREU, ZIEGLER MARCUSSOFIA, SERDEIRA AFRANE, SEVERO CARLOSMARCELODONAZAR, FRAST RODRIGOVALENTE, RECH PAULORENATO, TOFFOLO LAURO, SCALCO RENATASICILIANI, SCHWANKE CARLAHELENAAUGUSTIN. APPLICABILITY OF THE COBB ANGLE MEASUREMENT IN IDIOPATHIC SCOLIOSIS USING SCANNED IMAGING. COLUNA/COLUMNA 2017. [DOI: 10.1590/s1808-185120171601153058] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/22/2022] Open
Abstract
ABSTRACT Objectives: To compare the measurement of the Cobb angle on printed radiographs and on scanned radiographs viewed through the software "PixViewer". Methods: Preoperative radiographs of 23 patients were evaluated on printed films and through the software "PixViewer". The same evaluator, a spine surgeon, chose the proximal and distal limiting vertebrae of the main curve on printed radiographs, without identification of patients, and measured the Cobb angle based on these parameters. The same parameters and measurements were applied to scanned radiographs. The measurements were compared, as well as the choice of limiting vertebrae. Results: The average variation of the Cobb angle between methods was 1.48 ± 1.73°. The intraclass correlation coefficient (ICC) was 0.99, demonstrating excellent reproducibility. Conclusion: The Cobb method can be used to evaluate scoliosis through the "PixViewer" tool with the same reliability as the classic method on printed radiographs.
Collapse
Affiliation(s)
| | | | - AFRANE SERDEIRA
- Pontifícia Universidade Católica do Rio Grande do Sul, Brazil
| | | | | | | | - LAURO TOFFOLO
- Pontifícia Universidade Católica do Rio Grande do Sul, Brazil
| | | | | |
Collapse
|
9
|
Overuse, overdose, overdiagnosis… overreaction? Biomed Imaging Interv J 2011; 6:e8. [PMID: 21611049 PMCID: PMC3097773 DOI: 10.2349/biij.6.3.e8] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.2] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/02/2010] [Accepted: 04/28/2010] [Indexed: 11/24/2022] Open
Abstract
When x-rays were first discovered, the harmful effects of radiation had to be manifest in the early users before they were known. Today, radiation protection and safety have been established and the effects of radiation, as well as its risks, are known. Even so, medical radiation, in particular the growth in the use of computed tomography (CT), has resulted in soaring radiation doses received by the population in general. Inappropriate use has resulted in overuse, overdose and, perhaps, overdiagnosis, especially when used in screening. In the quest to control and curb the use of procedures involving radiation, however, we must be careful not to provoke a pandemic of irrational fear of radiation. Overreaction to the overuse and overdose of radiation might deter patients from life-saving procedures.
Collapse
|
10
|
Leung RS, Fairhurst J, Johnson K, Landes C, Moon L, Sprigg A, Offiah AC. Teleradiology: a modern approach to diagnosis, training, and research in child abuse? Clin Radiol 2011; 66:546-50. [PMID: 21310398 DOI: 10.1016/j.crad.2010.11.014] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.2] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/01/2010] [Revised: 11/22/2010] [Accepted: 11/25/2010] [Indexed: 10/18/2022]
Affiliation(s)
- R S Leung
- Department of Radiology, Great Ormond Street Hospital for Children, London, UK
| | | | | | | | | | | | | |
Collapse
|
11
|
Soh HS, Ung NM, Ng KH. The characteristics of Fuji IP Cassette Type PII and application for radiation oncology quality assurance tests and portal imaging. AUSTRALASIAN PHYSICAL & ENGINEERING SCIENCES IN MEDICINE 2008; 31:146-50. [PMID: 18697706 DOI: 10.1007/bf03178589] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.2] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/28/2022]
Abstract
The advancement of digital imaging has prompted more medical institutions to go filmless. The computed radiography (CR) system is becoming an important tool not only in diagnostic imaging, but also in radiation oncology. A new CR system that was specially designed for the use in radiation oncology, Fuji IP cassette type PII has been introduced to the market in the middle of year 2006. This project aimed to study some basic physical characteristics of this new type of cassette and explore its application for performing quality assurance (QA) tests and portal imaging in radiotherapy. All the images were read by FCR 5000 Plus reader. The image was found to reach its saturation value of 1023 (due to the image was stored in 10 bits data) by depending on the sensitivity value being adjusted. The uniformity test gave the result of 0.12%. The cassette was used to perform the QA tests which were previously performed using film. All the results met the specification as stated in AAPM Task Group 40. The comparison for the portal images of PortalVision contrast-detail phantom showed that the spatial resolution of the images obtained by CR system (Fujifilm Co., Ltd., Tokyo, Japan) were better than the EPID (Varian Medical Systems, Inc., Palo Alto, USA) and film system (Eastman Kodak Co., New York, USA). The IP cassette type PII was found to be suitable as an alternative QA test tool and portal imaging in radiotherapy.
Collapse
Affiliation(s)
- H S Soh
- Department of Biomedical Imaging, University of Malaya, Kuala Lumpur, Malaysia
| | | | | |
Collapse
|
12
|
QURAISHI NA, SAHU M, ROBINSON AHN. Do patients with fractures see their digital radiographs and does it help? Br J Radiol 2008; 81:436. [DOI: 10.1259/bjr/71709494] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/05/2022] Open
|
13
|
Rehani MM. The IAEA's activities in radiological protection in digital imaging. RADIATION PROTECTION DOSIMETRY 2008; 129:22-28. [PMID: 18440964 DOI: 10.1093/rpd/ncn155] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.1] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/26/2023]
Abstract
The success achieved to minimise the radiation exposure to staff to levels much below the prescribed dose limits encourages a similar approach in patient protection where dose limits do not exist but where reference levels are provided. A number of radiation injuries among patients have been reported, per capita dose is increasing and medical exposure has become the largest contributor to population dose. The International Atomic Energy Agency (IAEA) has supported a number of research projects, produced a wide range of guidance publications, conducted over 60 training courses in the past 5 y, produced for free dissemination training CDs on radiation protection and established a dedicated Web site on the radiological protection of patients (http://rpop.iaea.org). The number of countries that is currently receiving assistance in this field exceeds 80, as compared to about half a dozen 5 y ago. Significant results on patient dose management, demonstrating dose reduction of more than 30% while maintaining image quality, have been made available through IAEA projects.
Collapse
Affiliation(s)
- Madan M Rehani
- Radiological Protection of Patients Unit, International Atomic Energy Agency, PO Box 100, A-1400 Vienna, Austria.
| |
Collapse
|