Diagnostic assessment of a deep learning system for detecting atrial fibrillation in pulse waveforms

ObjectiveTo evaluate the diagnostic performance of a deep learning system for automated detection of atrial fibrillation (AF) in photoplethysmographic (PPG) pulse waveforms.MethodsWe trained a deep convolutional neural network (DCNN) to detect AF in 17 s PPG waveforms using a training data set of 149 048 PPG waveforms constructed from several publicly available PPG databases. The DCNN was validated using an independent test data set of 3039 smartphone-acquired PPG waveforms from adults at high risk of AF at a general outpatient clinic against ECG tracings reviewed by two cardiologists. Six established AF detectors based on handcrafted features were evaluated on the same test data set for performance comparison.ResultsIn the validation data set (3039 PPG waveforms) consisting of three sequential PPG waveforms from 1013 participants (mean (SD) age, 68.4 (12.2) years; 46.8% men), the prevalence of AF was 2.8%. The area under the receiver operating characteristic curve (AUC) of the DCNN for AF detection was 0.997 (95% CI 0.996 to 0.999) and was significantly higher than all the other AF detectors (AUC range: 0.924–0.985). The sensitivity of the DCNN was 95.2% (95% CI 88.3% to 98.7%), specificity was 99.0% (95% CI 98.6% to 99.3%), positive predictive value (PPV) was 72.7% (95% CI 65.1% to 79.3%) and negative predictive value (NPV) was 99.9% (95% CI 99.7% to 100%) using a single 17 s PPG waveform. Using the three sequential PPG waveforms in combination (<1 min in total), the sensitivity was 100.0% (95% CI 87.7% to 100%), specificity was 99.6% (95% CI 99.0% to 99.9%), PPV was 87.5% (95% CI 72.5% to 94.9%) and NPV was 100% (95% CI 99.4% to 100%).ConclusionsIn this evaluation of PPG waveforms from adults screened for AF in a real-world primary care setting, the DCNN had high sensitivity, specificity, PPV and NPV for detecting AF, outperforming other state-of-the-art methods based on handcrafted features.

Download Full-text

Development and Validation of a Score for Screening Suicide of Patients With Neuroendocrine Neoplasms

Frontiers in Psychiatry ◽

10.3389/fpsyt.2021.638152 ◽

2021 ◽

Vol 12 ◽

Author(s):

Lili Lu ◽

Yuru Shang ◽

Dietmar Zechner ◽

Christina Susanne Mullins ◽

Michael Linnebacher ◽

...

Keyword(s):

Life Expectancy ◽

Test Data ◽

Characteristic Curve ◽

Life Quality ◽

Training Data ◽

Neuroendocrine Neoplasm ◽

Neuroendocrine Neoplasms ◽

Validation Data ◽

Data Set ◽

Commit Suicide

Background: If the diagnosis of neuroendocrine neoplasm (NEN) increases the risk of patients to commit suicide has not been investigated so far. Identifying NEN patients at risk to commit suicide is important to increase their life quality and life expectancy.Methods and findings: Cancer cases were extracted from the Surveillance, Epidemiology, and End Results program and were divided into the NEN and the non-NEN cohorts. Subsequently, the NEN patients were randomly split into a training data set and a validation data set. Analyzing the training data set, we developed a score for assessing the risk to commit suicide for patients with NEN. In addition, we validated the score using the validation data set and evaluated, if this score could also be applied to other cancer entities by using the test data set, a non-NEN cohort. The odds ratio (OR) of suicide between NEN and non-NEN patients was determined. Moreover, the performance of a score was evaluated by the receiver operating characteristic curve and the area under the curve (AUC). Compared to non-NEN, NEN significantly increased the risk of suicide to 1.8-fold (NEN vs. non-NEN; OR, 1.832; P < 0.001). In addition, we observed that age, gender, race, marital status, tumor stage, histologic grade, surgery, and chemotherapy were associated with suicide among NEN patients; and a synthesized score based on these factors could significantly distinguish suicide individuals from non-suicide individuals in the training data set (AUC, 0.829; P < 0.001) and in the validation data set (AUC, 0.735; P < 0.001). This score also had a good performance when it was assessed by the test data set (AUC, 0.690; P < 0.001). This demonstrates that the score might also be applicable to other cancer entities.Conclusions: This population-based study suggests that NEN patients have a higher risk of suicide than non-NEN patients. In addition, this study provided a score, which can identify NEN patients at high-risk of committing suicide. Thus, this score in combination with current screening and prevention strategies for suicide may improve life quality and life expectancy of NEN patients.

Download Full-text

Deep Learning Using Multiple Degrees of Maximum-Intensity Projection for PET/CT Image Classification in Breast Cancer

Tomography ◽

10.3390/tomography8010011 ◽

2022 ◽

Vol 8 (1) ◽

pp. 131-141

Author(s):

Kanae Takahashi ◽

Tomoyuki Fujioka ◽

Jun Oyama ◽

Mio Mori ◽

Emi Yamaga ◽

...

Keyword(s):

Breast Cancer ◽

Deep Learning ◽

Image Classification ◽

Test Data ◽

Characteristic Curve ◽

Maximum Intensity Projection ◽

Maximum Intensity ◽

Training Data ◽

Ct Image ◽

Pet Ct

Deep learning (DL) has become a remarkably powerful tool for image processing recently. However, the usefulness of DL in positron emission tomography (PET)/computed tomography (CT) for breast cancer (BC) has been insufficiently studied. This study investigated whether a DL model using images with multiple degrees of PET maximum-intensity projection (MIP) images contributes to increase diagnostic accuracy for PET/CT image classification in BC. We retrospectively gathered 400 images of 200 BC and 200 non-BC patients for training data. For each image, we obtained PET MIP images with four different degrees (0°, 30°, 60°, 90°) and made two DL models using Xception. One DL model diagnosed BC with only 0-degree MIP and the other used four different degrees. After training phases, our DL models analyzed test data including 50 BC and 50 non-BC patients. Five radiologists interpreted these test data. Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated. Our 4-degree model, 0-degree model, and radiologists had a sensitivity of 96%, 82%, and 80–98% and a specificity of 80%, 88%, and 76–92%, respectively. Our 4-degree model had equal or better diagnostic performance compared with that of the radiologists (AUC = 0.936 and 0.872–0.967, p = 0.036–0.405). A DL model similar to our 4-degree model may lead to help radiologists in their diagnostic work in the future.

Download Full-text

DNA Methylation Markers for Pan-Cancer Prediction by Deep Learning

Genes ◽

10.3390/genes10100778 ◽

2019 ◽

Vol 10 (10) ◽

pp. 778 ◽

Cited By ~ 6

Author(s):

Liu ◽

Pan ◽

Li ◽

Yang ◽

...

Keyword(s):

Dna Methylation ◽

Deep Learning ◽

Sensitivity And Specificity ◽

Test Data ◽

Data Sets ◽

Methylation Data ◽

Average Sensitivity ◽

Validation Data ◽

Data Set ◽

Cancer Types

For cancer diagnosis, many DNA methylation markers have been identified. However, few studies have tried to identify DNA methylation markers to diagnose diverse cancer types simultaneously, i.e., pan-cancers. In this study, we tried to identify DNA methylation markers to differentiate cancer samples from the respective normal samples in pan-cancers. We collected whole genome methylation data of 27 cancer types containing 10,140 cancer samples and 3386 normal samples, and divided all samples into five data sets, including one training data set, one validation data set and three test data sets. We applied machine learning to identify DNA methylation markers, and specifically, we constructed diagnostic prediction models by deep learning. We identified two categories of markers: 12 CpG markers and 13 promoter markers. Three of 12 CpG markers and four of 13 promoter markers locate at cancer-related genes. With the CpG markers, our model achieved an average sensitivity and specificity on test data sets as 92.8% and 90.1%, respectively. For promoter markers, the average sensitivity and specificity on test data sets were 89.8% and 81.1%, respectively. Furthermore, in cell-free DNA methylation data of 163 prostate cancer samples, the CpG markers achieved the sensitivity as 100%, and the promoter markers achieved 92%. For both marker types, the specificity of normal whole blood was 100%. To conclude, we identified methylation markers to diagnose pan-cancers, which might be applied to liquid biopsy of cancers.

Download Full-text

Deep learning for automated glaucomatous optic neuropathy detection from ultra-widefield fundus images

British Journal of Ophthalmology ◽

10.1136/bjophthalmol-2020-317327 ◽

2020 ◽

pp. bjophthalmol-2020-317327

Author(s):

Zhongwen Li ◽

Chong Guo ◽

Duoru Lin ◽

Danyao Nie ◽

Yi Zhu ◽

...

Keyword(s):

Deep Learning ◽

Optic Neuropathy ◽

Network Architecture ◽

Characteristic Curve ◽

False Negative ◽

Learning System ◽

Glaucomatous Optic Neuropathy ◽

Pathological Myopia ◽

Data Set ◽

Negative Results

Background/AimsTo develop a deep learning system for automated glaucomatous optic neuropathy (GON) detection using ultra-widefield fundus (UWF) images.MethodsWe trained, validated and externally evaluated a deep learning system for GON detection based on 22 972 UWF images from 10 590 subjects that were collected at 4 different institutions in China and Japan. The InceptionResNetV2 neural network architecture was used to develop the system. The area under the receiver operating characteristic curve (AUC), sensitivity and specificity were used to assess the performance of detecting GON by the system. The data set from the Zhongshan Ophthalmic Center (ZOC) was selected to compare the performance of the system to that of ophthalmologists who mainly conducted UWF image analysis in clinics.ResultsThe system for GON detection achieved AUCs of 0.983–0.999 with sensitivities of 97.5–98.2% and specificities of 94.3–98.4% in four independent data sets. The most common reasons for false-negative results were confounding optic disc characteristics caused by high myopia or pathological myopia (n=39 (53%)). The leading cause for false-positive results was having other fundus lesions (n=401 (96%)). The performance of the system in the ZOC data set was comparable to that of an experienced ophthalmologist (p>0.05).ConclusionOur deep learning system can accurately detect GON from UWF images in an automated fashion. It may be used as a screening tool to improve the accessibility of screening and promote the early diagnosis and management of glaucoma.

Download Full-text

Image classification of different clove (Syzygium aromaticum) quality using deep learning method with convolutional neural network algorithm

IOP Conference Series Earth and Environmental Science ◽

10.1088/1755-1315/905/1/012018 ◽

2021 ◽

Vol 905 (1) ◽

pp. 012018

Author(s):

I Y Prayogi ◽

Sandra ◽

Y Hendrawan

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

Test Data ◽

Training Data ◽

Syzygium Aromaticum ◽

Learning Method ◽

Validation Data ◽

Reading Accuracy

Abstract The objective of this study is to classify the quality of dried clove flowers using deep learning method with Convolutional Neural Network (CNN) algorithm, and also to perform the sensitivity analysis of CNN hyperparameters to obtain best model for clove quality classification process. The quality of clove as raw material in this study was determined according to SNI 3392-1994 by PT. Perkebunan Nusantara XII Pancusari Plantation, Malang, East Java, Indonesia. In total 1,600 images of dried clove flower were divided into 4 qualities. Each clove quality has 225 training data, 75 validation data, and 100 test data. The first step of this study is to build CNN model architecture as first model. The result of that model gives 65.25% reading accuracy. The second step is to analyze CNN sensitivity or CNN hyperparameter on the first model. The best value of CNN hyperparameter in each step then to be used in the next stage. Finally, after CNN hyperparameter carried out the reading accuracy of the test data is improved to 87.75%.

Download Full-text

Accurate prediction of somatic variants using deep learning model.

Journal of Clinical Oncology ◽

10.1200/jco.2020.38.15_suppl.e13659 ◽

2020 ◽

Vol 38 (15_suppl) ◽

pp. e13659-e13659

Author(s):

Peng Zhang ◽

Kai Wang ◽

Ming Yao ◽

Aodi Wang ◽

Lijuan Chen ◽

...

Keyword(s):

Deep Learning ◽

Sequence Data ◽

Variant Calling ◽

Training Data ◽

Stochastic Gradient Descent ◽

Validation Data ◽

Data Set ◽

Candidate Mutation ◽

And Control ◽

Panel Sequencing

e13659 Background: Efficient and accurate identification of somatic variant is important for understanding the formation, progression, and treatment of cancer. It is necessary to conduct manual review by Integrative Genomic Viewer (IGV) in traditional variant calling process. However, the traditional manual is heavy workload when evaluating tumor with a high variant burden. In this study, a new convolutional neural network (CNN) method was created to train models for somatic mutation identification, which was suitable for Panel sequencing platform with different tumor purities. Methods: A total of 1000 tumor samples from next generation sequencing (NGS)-based genetic testing by a College of American Pathologists (CAP) accredited and Clinical Laboratory Improvement Amendments (CLIA) certified laboratory. Through variant calling program, like GATK, the candidate mutation locations were identified and standardized by manual confirmation. For each candidate mutation location, reads of both tumor and control tissue were extracted. A 2-dimensional feature matrix M of size (2k+1) * 32 in each candidate base was created. The rows of 2k+1 represented the length of candidate region, and the 32 columns included the reads coverage frequency, mapping quality messages, and genome local scores of different tumor and control tissues. CNN model, which includes nine convolutional layers structured by Temporal Convolutional Networks (TCN) but with a different structure to adapt to the proposed input matrix, was used for training. The training data set including manually validated sequence data was used as benchmark test, and optimized by Stochastic Gradient Descent (SGD) optimizer with a learning rate of 0.01 was used for training. Results: The validation data set included 15 mixed samples which were composed of different proportions of known cell lines and real mixed blood samples. The pooled DNA contained 2,359 somatic variants, with expected variant allele frequencies ranged from 3% to 97% in each pool. The overall sensitivity and positive predictive value (PPV) of single nucleotide variants (SNVs) were 99.3% and 99.8%, respectively. Conclusions: A novel and sensitive computational tool for somatic variation detection in DNA Panel sequencing was developed. Our result showed that the deep learning CNN model could call variant in Panel sequencing data.

Download Full-text

Predictive Model of Chemotherapy-induced Myelosuppression for Patients With Esophageal Cancer

10.21203/rs.3.rs-871270/v1 ◽

2021 ◽

Author(s):

Ziming Zheng ◽

Qilin Zhang ◽

Yong Han ◽

Tingting Wu ◽

Yu Zhang

Keyword(s):

Esophageal Cancer ◽

Prediction Model ◽

Characteristic Curve ◽

Predictive Ability ◽

Influential Factors ◽

Training Data ◽

Central China ◽

Ethical Guidelines ◽

Validation Data ◽

Data Set

Abstract Background: The influential factors of chemotherapy-induced myelosuppression in esophageal cancer in central China are unclear. This study aimed to develop a model for prediction of incidence of myelosuppression during chemotherapy among patients with esophageal cancer. Methods: A total of 1446 patients with esophageal cancer who underwent five different chemotherapy regimens between 2013 and 2020 at our institute were randomly assigned in a 7:3 ratio to training and validation data sets. Clinical and drug-related variables were used to develop the prediction model from the training data set by the machine learning method of random forest. Finally, this model were tested in the validation data set.Results: The prediction model were established with 16 indispensable variables selected from 46 variables. The model obtained an area under the receiver-operating characteristic curve of 0.883 and accompanied by prediction accuracy of 80.0%, sensitivity of 77.8% and specificity of 81.8%. Conclusion: This new prediction model showed excellent predictive ability of incidence of myelosuppression in turn providing preventative measures for patients with esophageal cancer during chemotherapy.Trial registration: This study protocol was approved by the institutional ethics board of the Union Hospital of Huazhong University of Science and Technology (retrospectively registered No. 2018S333). This study was performed in accordance with the ethical guidelines of the Declaration of Helsinki and Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).

Download Full-text

The application of deep learning for the classification of correct and incorrect SNP genotypes from whole-genome DNA sequencing pipelines

Journal of Applied Genetics ◽

10.1007/s13353-020-00586-0 ◽

2020 ◽

Vol 61 (4) ◽

pp. 607-616

Author(s):

Krzysztof Kotlarz ◽

Magda Mielczarek ◽

Tomasz Suchocki ◽

Bartosz Czech ◽

Bernt Guldbrandtsen ◽

...

Keyword(s):

Deep Learning ◽

Next Generation Sequencing ◽

Training Data ◽

Next Generation ◽

Validation Data ◽

Data Set ◽

Genotype Information ◽

Naive Algorithm ◽

Generation Sequencing

Abstract A downside of next-generation sequencing technology is the high technical error rate. We built a tool, which uses array-based genotype information to classify next-generation sequencing–based SNPs into the correct and the incorrect calls. The deep learning algorithms were implemented via Keras. Several algorithms were tested: (i) the basic, naïve algorithm, (ii) the naïve algorithm modified by pre-imposing different weights on incorrect and correct SNP class in calculating the loss metric and (iii)–(v) the naïve algorithm modified by random re-sampling (with replacement) of the incorrect SNPs to match 30%/60%/100% of the number of correct SNPs. The training data set was composed of data from three bulls and consisted of 2,227,995 correct (97.94%) and 46,920 incorrect SNPs, while the validation data set consisted of data from one bull with 749,506 correct (98.05%) and 14,908 incorrect SNPs. The results showed that for a rare event classification problem, like incorrect SNP detection in NGS data, the most parsimonious naïve model and a model with the weighting of SNP classes provided the best results for the classification of the validation data set. Both classified 19% of truly incorrect SNPs as incorrect and 99% of truly correct SNPs as correct and resulted in the F1 score of 0.21 — the highest among the compared algorithms. We conclude the basic models were less adapted to the specificity of a training data set and thus resulted in better classification of the independent, validation data set, than the other tested models.

Download Full-text

Evaluation of automated cephalometric analysis based on the latest deep learning method

The Angle Orthodontist ◽

10.2319/021220-100.1 ◽

2021 ◽

Author(s):

Hye-Won Hwang ◽

Jun-Ho Moon ◽

Min-Gyu Kim ◽

Richard E. Donatelli ◽

Shin-Jae Lee

Keyword(s):

Deep Learning ◽

Test Data ◽

Training Data ◽

Superior Performance ◽

Cephalometric Analysis ◽

Data Sets ◽

Learning Method ◽

Test Results ◽

Classification Rate ◽

Data Set

ABSTRACT Objectives To compare an automated cephalometric analysis based on the latest deep learning method of automatically identifying cephalometric landmarks (AI) with previously published AI according to the test style of the worldwide AI challenges at the International Symposium on Biomedical Imaging conferences held by the Institute of Electrical and Electronics Engineers (IEEE ISBI). Materials and Methods This latest AI was developed by using a total of 1983 cephalograms as training data. In the training procedures, a modification of a contemporary deep learning method, YOLO version 3 algorithm, was applied. Test data consisted of 200 cephalograms. To follow the same test style of the AI challenges at IEEE ISBI, a human examiner manually identified the IEEE ISBI-designated 19 cephalometric landmarks, both in training and test data sets, which were used as references for comparison. Then, the latest AI and another human examiner independently detected the same landmarks in the test data set. The test results were compared by the measures that appeared at IEEE ISBI: the success detection rate (SDR) and the success classification rates (SCR). Results SDR of the latest AI in the 2-mm range was 75.5% and SCR was 81.5%. These were greater than any other previous AIs. Compared to the human examiners, AI showed a superior success classification rate in some cephalometric analysis measures. Conclusions This latest AI seems to have superior performance compared to previous AI methods. It also seems to demonstrate cephalometric analysis comparable to human examiners.

Download Full-text

A novel ferroptosis related gene signature is associated with prognosis in patients with ovarian serous cystadenocarcinoma

Scientific Reports ◽

10.1038/s41598-021-90126-5 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Zhixiang Yu ◽

Haiyan He ◽

Yanan Chen ◽

Qiuhe Ji ◽

Min Sun

Keyword(s):

Cox Regression ◽

Expression Profiles ◽

Critical Role ◽

Gene Signature ◽

Cancer Genome ◽

Training Data ◽

Hub Genes ◽

Validation Data ◽

Data Set ◽

Cox Regression Analysis

AbstractOvarian cancer (OV) is a common type of carcinoma in females. Many studies have reported that ferroptosis is associated with the prognosis of OV patients. However, the mechanism by which this occurs is not well understood. We utilized Genotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) to identify ferroptosis-related genes in OV. In the present study, we applied Cox regression analysis to select hub genes and used the least absolute shrinkage and selection operator to construct a prognosis prediction model with mRNA expression profiles and clinical data from TCGA. A series of analyses for this signature was performed in TCGA. We then verified the identified signature using International Cancer Genome Consortium (ICGC) data. After a series of analyses, we identified six hub genes (DNAJB6, RB1, VIMP/ SELENOS, STEAP3, BACH1, and ALOX12) that were then used to construct a model using a training data set. The model was then tested using a validation data set and was found to have high sensitivity and specificity. The identified ferroptosis-related hub genes might play a critical role in the mechanism of OV development. The gene signature we identified may be useful for future clinical applications.

Download Full-text