1
|
Dikici E, Nguyen XV, Takacs N, Prevedello LM. Prediction of model generalizability for unseen data: Methodology and case study in brain metastases detection in T1-Weighted contrast-enhanced 3D MRI. Comput Biol Med 2023; 159:106901. [PMID: 37068317 DOI: 10.1016/j.compbiomed.2023.106901] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/22/2022] [Revised: 03/08/2023] [Accepted: 04/09/2023] [Indexed: 04/19/2023]
Abstract
BACKGROUND AND PURPOSE A medical AI system's generalizability describes the continuity of its performance acquired from varying geographic, historical, and methodologic settings. Previous literature on this topic has mostly focused on "how" to achieve high generalizability (e.g., via larger datasets, transfer learning, data augmentation, model regularization schemes), with limited success. Instead, we aim to understand "when" the generalizability is achieved: Our study presents a medical AI system that could estimate its generalizability status for unseen data on-the-fly. MATERIALS AND METHODS We introduce a latent space mapping (LSM) approach utilizing Fréchet distance loss to force the underlying training data distribution into a multivariate normal distribution. During the deployment, a given test data's LSM distribution is processed to detect its deviation from the forced distribution; hence, the AI system could predict its generalizability status for any previously unseen data set. If low model generalizability is detected, then the user is informed by a warning message integrated into a sample deployment workflow. While the approach is applicable for most classification deep neural networks (DNNs), we demonstrate its application to a brain metastases (BM) detector for T1-weighted contrast-enhanced (T1c) 3D MRI. The BM detection model was trained using 175 T1c studies acquired internally (from the authors' institution) and tested using (1) 42 internally acquired exams and (2) 72 externally acquired exams from the publicly distributed Brain Mets dataset provided by the Stanford University School of Medicine. Generalizability scores, false positive (FP) rates, and sensitivities of the BM detector were computed for the test datasets. RESULTS AND CONCLUSION The model predicted its generalizability to be low for 31% of the testing data (i.e., two of the internally and 33 of the externally acquired exams), where it produced (1) ∼13.5 false positives (FPs) at 76.1% BM detection sensitivity for the low and (2) ∼10.5 FPs at 89.2% BM detection sensitivity for the high generalizability groups respectively. These results suggest that the proposed formulation enables a model to predict its generalizability for unseen data.
Collapse
Affiliation(s)
- Engin Dikici
- The Ohio State University, College of Medicine, Department of Radiology, Columbus, OH, 43210, USA.
| | - Xuan V Nguyen
- The Ohio State University, College of Medicine, Department of Radiology, Columbus, OH, 43210, USA
| | - Noah Takacs
- The Ohio State University, College of Medicine, Department of Radiology, Columbus, OH, 43210, USA
| | - Luciano M Prevedello
- The Ohio State University, College of Medicine, Department of Radiology, Columbus, OH, 43210, USA
| |
Collapse
|
2
|
Xue P, Si M, Qin D, Wei B, Seery S, Ye Z, Chen M, Wang S, Song C, Zhang B, Ding M, Zhang W, Bai A, Yan H, Dang L, Zhao Y, Rezhake R, Zhang S, Qiao Y, Qu Y, Jiang Y. Unassisted Clinicians Versus Deep Learning-Assisted Clinicians in Image-Based Cancer Diagnostics: Systematic Review With Meta-analysis. J Med Internet Res 2023; 25:e43832. [PMID: 36862499 PMCID: PMC10020907 DOI: 10.2196/43832] [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: 10/26/2022] [Revised: 01/19/2023] [Accepted: 02/13/2023] [Indexed: 02/16/2023] Open
Abstract
BACKGROUND A number of publications have demonstrated that deep learning (DL) algorithms matched or outperformed clinicians in image-based cancer diagnostics, but these algorithms are frequently considered as opponents rather than partners. Despite the clinicians-in-the-loop DL approach having great potential, no study has systematically quantified the diagnostic accuracy of clinicians with and without the assistance of DL in image-based cancer identification. OBJECTIVE We systematically quantified the diagnostic accuracy of clinicians with and without the assistance of DL in image-based cancer identification. METHODS PubMed, Embase, IEEEXplore, and the Cochrane Library were searched for studies published between January 1, 2012, and December 7, 2021. Any type of study design was permitted that focused on comparing unassisted clinicians and DL-assisted clinicians in cancer identification using medical imaging. Studies using medical waveform-data graphics material and those investigating image segmentation rather than classification were excluded. Studies providing binary diagnostic accuracy data and contingency tables were included for further meta-analysis. Two subgroups were defined and analyzed, including cancer type and imaging modality. RESULTS In total, 9796 studies were identified, of which 48 were deemed eligible for systematic review. Twenty-five of these studies made comparisons between unassisted clinicians and DL-assisted clinicians and provided sufficient data for statistical synthesis. We found a pooled sensitivity of 83% (95% CI 80%-86%) for unassisted clinicians and 88% (95% CI 86%-90%) for DL-assisted clinicians. Pooled specificity was 86% (95% CI 83%-88%) for unassisted clinicians and 88% (95% CI 85%-90%) for DL-assisted clinicians. The pooled sensitivity and specificity values for DL-assisted clinicians were higher than for unassisted clinicians, at ratios of 1.07 (95% CI 1.05-1.09) and 1.03 (95% CI 1.02-1.05), respectively. Similar diagnostic performance by DL-assisted clinicians was also observed across the predefined subgroups. CONCLUSIONS The diagnostic performance of DL-assisted clinicians appears better than unassisted clinicians in image-based cancer identification. However, caution should be exercised, because the evidence provided in the reviewed studies does not cover all the minutiae involved in real-world clinical practice. Combining qualitative insights from clinical practice with data-science approaches may improve DL-assisted practice, although further research is required. TRIAL REGISTRATION PROSPERO CRD42021281372; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=281372.
Collapse
Affiliation(s)
- Peng Xue
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Mingyu Si
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Dongxu Qin
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Bingrui Wei
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Samuel Seery
- Faculty of Health and Medicine, Division of Health Research, Lancaster University, Lancaster, United Kingdom
| | - Zichen Ye
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Mingyang Chen
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Sumeng Wang
- Department of Cancer Epidemiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Cheng Song
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Bo Zhang
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Ming Ding
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Wenling Zhang
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Anying Bai
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Huijiao Yan
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Le Dang
- Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Yuqian Zhao
- Sichuan Cancer Hospital & Institute, Sichuan Cancer Center, School of Medicine, University of Electronic Science & Technology of China, Sichuan, China
| | - Remila Rezhake
- Affiliated Cancer Hospital, The 3rd Affiliated Teaching Hospital of Xinjiang Medical University, Xinjiang, China
| | - Shaokai Zhang
- Henan Cancer Hospital, Affiliated Cancer Hospital of Zhengzhou University, Henan, China
| | - Youlin Qiao
- Center for Global Health, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Yimin Qu
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| | - Yu Jiang
- Department of Epidemiology and Biostatistics, School of Population Medicine and Public Health, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
| |
Collapse
|
3
|
Lung Cancer Nodules Detection via an Adaptive Boosting Algorithm Based on Self-Normalized Multiview Convolutional Neural Network. JOURNAL OF ONCOLOGY 2022; 2022:5682451. [PMID: 36199795 PMCID: PMC9529389 DOI: 10.1155/2022/5682451] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 05/20/2022] [Revised: 06/28/2022] [Accepted: 07/19/2022] [Indexed: 11/18/2022]
Abstract
Lung cancer is the deadliest cancer killing almost 1.8 million people in 2020. The new cases are expanding alarmingly. Early lung cancer manifests itself in the form of nodules in the lungs. One of the most widely used techniques for both lung cancer early and noninvasive diagnosis is computed tomography (CT). However, the intensive workload of radiologists to read a large number of scans for nodules detection gives rise to issues like false detection and missed detection. To overcome these issues, we proposed an innovative strategy titled adaptive boosting self-normalized multiview convolution neural network (AdaBoost-SNMV-CNN) for lung cancer nodules detection across CT scans. In AdaBoost-SNMV-CNN, MV-CNN function as a baseline learner while the scaled exponential linear unit (SELU) activation function normalizes the layers by considering their neighbors' information and a special drop-out technique (α-dropout). The proposed method was trained and tested using the widely Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) and Early Lung Cancer Action Program (ELCAP) datasets. AdaBoost-SNMV-CNN achieved an accuracy of 92%, sensitivity of 93%, and specificity of 92% for lung nodules detection on the LIDC-IDRI dataset. Meanwhile, on the ELCAP dataset, the accuracy for detecting lung nodules was 99%, sensitivity 100%, and specificity 98%. AdaBoost-SNMV-CNN outperformed the majority of the model in accuracy, sensitivity, and specificity. The multiviews confer the model's good generalization and learning ability for diverse features of lung nodules, the model architecture is simple, and has a minimal computational time of around 102 minutes. We believe that AdaBoost-SNMV-CNN has good accuracy for the detection of lung nodules and anticipate its potential application in the noninvasive clinical diagnosis of lung cancer. This model can be of good assistance to the radiologist and will be of interest to researchers involved in the designing and development of advanced systems for the detection of lung nodules to accomplish the goal of noninvasive diagnosis of lung cancer.
Collapse
|
4
|
Analysis of Smart Lung Tumour Detector and Stage Classifier Using Deep Learning Techniques with Internet of Things. COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE 2022; 2022:4608145. [PMID: 36148416 PMCID: PMC9489382 DOI: 10.1155/2022/4608145] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 06/04/2022] [Revised: 07/15/2022] [Accepted: 07/27/2022] [Indexed: 11/30/2022]
Abstract
The use of artificial intelligence (AI) and the Internet of Things (IoT), which is a developing technology in medical applications that assists physicians in making more informed decisions regarding patients' courses of treatment, has become increasingly widespread in recent years in the field of healthcare. On the other hand, the number of PET scans that are being performed is rising, and radiologists are getting significantly overworked as a result. As a direct result of this, a novel approach that goes by the name “computer-aided diagnostics” is now being investigated as a potential method for reducing the tremendous workloads. A Smart Lung Tumor Detector and Stage Classifier (SLD-SC) is presented in this study as a hybrid technique for PET scans. This detector can identify the stage of a lung tumour. Following the development of the modified LSTM for the detection of lung tumours, the proposed SLD-SC went on to develop a Multilayer Convolutional Neural Network (M-CNN) for the classification of the various stages of lung cancer. This network was then modelled and validated utilising standard benchmark images. The suggested SLD-SC is now being evaluated on lung cancer pictures taken from patients with the disease. We observed that our recommended method gave good results when compared to other tactics that are currently being used in the literature. These findings were outstanding in terms of the performance metrics accuracy, recall, and precision that were assessed. As can be shown by the much better outcomes that were achieved with each of the test images that were used, our proposed method excels its rivals in a variety of respects. In addition to this, it achieves an average accuracy of 97 percent in the categorization of lung tumours, which is much higher than the accuracy achieved by the other approaches.
Collapse
|
5
|
Bhattacharyya D, Thirupathi Rao N, Joshua ESN, Hu YC. A bi-directional deep learning architecture for lung nodule semantic segmentation. THE VISUAL COMPUTER 2022; 39:1-17. [PMID: 36097497 PMCID: PMC9453728 DOI: 10.1007/s00371-022-02657-1] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Accepted: 08/20/2022] [Indexed: 06/15/2023]
Abstract
Lung nodules are abnormal growths and lesions may exist. Both lungs may have nodules. Most lung nodules are harmless (not cancerous/malignant). Pulmonary nodules are rare in lung cancer. X-rays and CT scans identify the lung nodules. Doctors may term the growth a lung spot, coin lesion, or shadow. It is necessary to obtain properly computed tomography (CT) scans of the lungs to get an accurate diagnosis and a good estimate of the severity of lung cancer. This study aims to design and evaluate a deep learning (DL) algorithm for identifying pulmonary nodules (PNs) using the LUNA-16 dataset and examine the prevalence of PNs using DB-Net. The paper states that a new, resource-efficient deep learning architecture is called for, and it has been given the name of DB-NET. When a physician orders a CT scan, they need to employ an accurate and efficient lung nodule segmentation method because they need to detect lung cancer at an early stage. However, segmentation of lung nodules is a difficult task because of the nodules' characteristics on the CT image as well as the nodules' concealed shape, visual quality, and context. The DB-NET model architecture is presented as a resource-efficient deep learning solution for handling the challenge at hand in this paper. Furthermore, it incorporates the Mish nonlinearity function and the mask class weights to improve segmentation effectiveness. In addition to the LUNA-16 dataset, which contained 1200 lung nodules collected during the LUNA-16 test, the LUNA-16 dataset was extensively used to train and assess the proposed model. The DB-NET architecture surpasses the existing U-NET model by a dice coefficient index of 88.89%, and it also achieves a similar level of accuracy to that of human experts.
Collapse
Affiliation(s)
- Debnath Bhattacharyya
- Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, Guntur, 522 502 India
| | - N. Thirupathi Rao
- Department of Computer Science and Engineering, Vignan’s Institute of Information Technology (A), Visakhapatnam, 530049 AP India
| | - Eali Stephen Neal Joshua
- Department of Computer Science and Engineering, Vignan’s Institute of Information Technology (A), Visakhapatnam, 530049 AP India
| | - Yu-Chen Hu
- Department of Computer Science and Information Management, Providence University, 200, Sec. 7, Taiwan Boulevard, Shalu Dist., Taichung City, 43301 Taiwan R.O.C
| |
Collapse
|
6
|
Tandon R, Agrawal S, Chang A, Band SS. VCNet: Hybrid Deep Learning Model for Detection and Classification of Lung Carcinoma Using Chest Radiographs. Front Public Health 2022; 10:894920. [PMID: 35795700 PMCID: PMC9251197 DOI: 10.3389/fpubh.2022.894920] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/12/2022] [Accepted: 04/22/2022] [Indexed: 11/25/2022] Open
Abstract
Detection of malignant lung nodules from Computed Tomography (CT) images is a significant task for radiologists. But, it is time-consuming in nature. Despite numerous breakthroughs in studies on the application of deep learning models for the identification of lung cancer, researchers and doctors still face challenges when trying to deploy the model in clinical settings to achieve improved accuracy and sensitivity on huge datasets. In most situations, deep convolutional neural networks are used for detecting the region of the main nodule of the lung exclusive of considering the neighboring tissues of the nodule. Although the accuracy achieved through CNN is good enough but this models performance degrades when there are variations in image characteristics like: rotation, tiling, and other abnormal image orientations. CNN does not store relative spatial relationships among features in scanned images. As CT scans have high spatial resolution and are sensitive to misalignments during the scanning process, there is a requirement of a technique which helps in considering spatial information of image features also. In this paper, a hybrid model named VCNet is proposed by combining the features of VGG-16 and capsule network (CapsNet). VGG-16 model is used for object recognition and classification. CapsNet is used to address the shortcomings of convolutional neural networks for image rotation, tiling, and other abnormal image orientations. The performance of VCNeT is verified on the Lung Image Database Consortium (LIDC) image collection dataset. It achieves higher testing accuracy of 99.49% which is significantly better than MobileNet, Xception, and VGG-16 that has achieved an accuracy of 98, 97.97, and 96.95%, respectively. Therefore, the proposed hybrid VCNet framework can be used for the clinical purpose for nodule detection in lung carcinoma detection.
Collapse
Affiliation(s)
- Ritu Tandon
- Institute of Advance Computing, SAGE University, Indore, India
- *Correspondence: Ritu Tandon
| | - Shweta Agrawal
- Institute of Advance Computing, SAGE University, Indore, India
- Shweta Agrawal
| | - Arthur Chang
- Bachelor Program in Interdisciplinary Studies, National Yunlin University of Science and Technology, Douliu, Taiwan
- Arthur Chang
| | - Shahab S. Band
- Future Technology Research Center, College of Future, National Yunlin University of Science and Technology, Douliu, Taiwan
- Shahab S. Band
| |
Collapse
|
7
|
Cloud-Based Lung Tumor Detection and Stage Classification Using Deep Learning Techniques. BIOMED RESEARCH INTERNATIONAL 2022; 2022:4185835. [PMID: 35047635 PMCID: PMC8763490 DOI: 10.1155/2022/4185835] [Citation(s) in RCA: 11] [Impact Index Per Article: 5.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 10/23/2021] [Revised: 11/30/2021] [Accepted: 12/07/2021] [Indexed: 02/01/2023]
Abstract
Artificial intelligence (AI), Internet of Things (IoT), and the cloud computing have recently become widely used in the healthcare sector, which aid in better decision-making for a radiologist. PET imaging or positron emission tomography is one of the most reliable approaches for a radiologist to diagnosing many cancers, including lung tumor. In this work, we proposed stage classification of lung tumor which is a more challenging task in computer-aided diagnosis. As a result, a modified computer-aided diagnosis is being considered as a way to reduce the heavy workloads and second opinion to radiologists. In this paper, we present a strategy for classifying and validating different stages of lung tumor progression, as well as a deep neural model and data collection using cloud system for categorizing phases of pulmonary illness. The proposed system presents a Cloud-based Lung Tumor Detector and Stage Classifier (Cloud-LTDSC) as a hybrid technique for PET/CT images. The proposed Cloud-LTDSC initially developed the active contour model as lung tumor segmentation, and multilayer convolutional neural network (M-CNN) for classifying different stages of lung cancer has been modelled and validated with standard benchmark images. The performance of the presented technique is evaluated using a benchmark image LIDC-IDRI dataset of 50 low doses and also utilized the lung CT DICOM images. Compared with existing techniques in the literature, our proposed method achieved good result for the performance metrics accuracy, recall, and precision evaluated. Under numerous aspects, our proposed approach produces superior outcomes on all of the applied dataset images. Furthermore, the experimental result achieves an average lung tumor stage classification accuracy of 97%-99.1% and an average of 98.6% which is significantly higher than the other existing techniques.
Collapse
|