Independent Test Dataset Research Articles

A Novel Deep Multi-Image Object Detection Approach for Detecting Alien Barleys in Oat Fields Using RGB UAV Images

Oat products are significant parts of a healthy diet. Pure oat is gluten-free, which makes it an excellent choice for people with celiac disease. Elimination of alien cereals is important not only in gluten-free oat production but also in seed production. Detecting gluten-rich crops such as wheat, rye, and barley in an oat production field is an important initial processing step in gluten-free food industries; however, this particular step can be extremely time consuming. This article demonstrates the potential of emerging drone techniques for identifying alien barleys in an oat stand. The primary aim of this study was to develop and assess a novel machine-learning approach that automatically detects and localizes barley plants by employing drone images. An Unbiased Teacher v2 semi-supervised object-detection deep convolutional neural network (CNN) was employed to detect barley ears in drone images with a 1.5 mm ground sample distance. The outputs of the object detector were transformed into ground coordinates by employing a photogrammetric technique. The ground coordinates were analyzed with the kernel density estimate (KDE) clustering approach to form a probabilistic map of the ground locations of barley plants. The detector was trained using a dataset from a reference data production site (located in Ilmajoki, Finland) and tested using a 10% independent test data sample from the same site and a completely unseen dataset from a commercial gluten-free oats production field in Seinäjoki, Finland. In the reference data production dataset, 82.9% of the alien barley plants were successfully detected; in the independent farm test dataset, 60.5% of the ground-truth barley plants were correctly recognized. Our results establish the usefulness and importance of the proposed drone-based ultra-high-resolution red–green–blue (RGB) imaging approach for modern grain production industries.

Identification of biomarkers associated with the invasion of nonfunctional pituitary neuroendocrine tumors based on the immune microenvironment.

The invasive behavior of nonfunctioning pituitary neuroendocrine tumors (NF-PitNEts) affects complete resection and indicates a poor prognosis. Cancer immunotherapy has been experimentally used for the treatment of many tumors, including pituitary tumors. The current study aimed to screen the key immune-related genes in NF-PitNEts with invasion. We used two cohorts to explore novel biomarkers in NF-PitNEts. The immune infiltration-associated differentially expressed genes (DEGs) were obtained based on high/low immune scores, which were calculated through the ESTIMATE algorithm. The abundance of immune cells was predicted using the ImmuCellAI database. WGCNA was used to construct a coexpression network of immune cell-related genes. Random forest analysis was used to select the candidate genes associated with invasion. The expression of key genes was verified in external validation set using quantitative real-time polymerase chain reaction (qRT‒PCR). The immune and invasion related DEGs was obtained based on the first dataset of NF-PitNEts (n=112). The immune cell-associated modules in NF-PitNEts were calculate by WGCNA. Random forest analysis was performed on 81 common genes intersected by immune-related genes, invasion-related genes, and module genes. Then, 20 of these genes with the highest RF score were selected to construct the invasion and immune-associated classification model. We found that this model had high prediction accuracy for tumor invasion, which had the largest area under the receiver operating characteristic curve (AUC) value in the training dataset from the first dataset (n=78), the self-test dataset from the first dataset (n=34), and the independent test dataset (n=73) (AUC=0.732/0.653/0.619). Functional enrichment analysis revealed that 8 out of the 20 genes were enriched in multiple signaling pathways. Subsequently, the 8-gene (BMP6, CIB2, FABP5, HOMER2, MAML3, NIN, PRKG2 and SIDT2) classification model was constructed and showed good efficiency in the first dataset (AUC=0.671). In addition, the expression levels of these 8 genes were verified by qRT‒PCR. We identified eight key genes associated with invasion and immunity in NF-PitNEts that may play a fundamental role in invasive progression and may provide novel potential immunotherapy targets for NF-PitNEts.

Novel algorithms improve arrhythmia detection accuracy in insertable cardiac monitors.

Insertable cardiac monitors (ICMs) are commonly used to diagnose cardiac arrhythmias. False detections in the latest ICM systems remain an issue, primarily due to inaccurate R-wave sensing. New discrimination algorithms were developed and tested to reduce false detections of atrial fibrillation (AF), pause, and tachycardia episodes in ICMs. Stored electrograms (EGMs) of AF, pause, and tachycardia episodes detected by Abbott Confirm Rx™ ICMs were extracted from the Merlin.net™ Patient Care Network, and manually adjudicated to establish independent training and testing datasets. New discrimination algorithms were developed to reject false episodes due to inaccurate R-wave sensing, P-wave identification, and R-R interval patterns. The performance of these new algorithms was quantified by false positive reduction (FPR) and true positive maintenance (TPM), relative to the existing algorithms. The new AF detection algorithm was trained on 5911 EGMs from 744 devices, resulting in 66.9% FPR and 97.8% TPM. In the testing data set of 1354 EGMs from 119 devices, this algorithm achieved 45.8% FPR and 97.0% TPM. The new pause algorithm was trained on 7178 EGMs from 1490 devices, resulting in 70.9% FPR and 98.7% TPM. In the testing data set of 1442 EGMs from 340 devices, this algorithm achieved 74.4% FPR and 99.3% TPM. The new tachycardia algorithm was trained on 520 EGMs from 204 devices, resulting in 57.0% FPR and 96.6% TPM. In the testing data set of 459 EGMs from 237 devices, this algorithm achieved 57.9% FPR and 96.5% TPM. The new algorithms substantially reduced false AF, pause, and tachycardia episodes while maintaining the majority of true arrhythmia episodes detected by the Abbott ICM algorithms that exist today. Implementing these algorithms in the next-generation ICM systems may lead to improved detection accuracy, in-clinic efficiency, and device battery longevity.

An ensemble of stacking classifiers for improved prediction of miRNA–mRNA interactions

MicroRNAs (miRNAs) are small non-coding RNA molecules that play a crucial role in regulating gene expression at the post-transcriptional level by binding to potential target sites of messenger RNAs (mRNAs), facilitated by the Argonaute family of proteins. Selecting the conservative candidate target sites (CTS) is a challenging step, considering that most of the existing computational algorithms primarily focus on canonical site types, which is a time-consuming and inefficient utilization of miRNA target site interactions. We developed a stacking classifier algorithm that addresses the CTS selection criteria using feature-encoding techniques that generates feature vectors, including k-mer nucleotide composition, dinucleotide composition, pseudo-nucleotide composition, and sequence order coupling. This innovative stacking classifier algorithm surpassed previous state-of-the-art algorithms in predicting functional miRNA targets. We evaluated the performance of the proposed model on 10 independent test datasets and obtained an average accuracy of 79.77%, which is a significant improvement of 7.26 % over previous models. This improvement shows that the proposed method has great potential for distinguishing highly functional miRNA targets and can serve as a valuable tool in biomedical and drug development research.

Understanding structure-guided variant effect predictions using 3D convolutional neural networks.

Predicting pathogenicity of missense variants in molecular diagnostics remains a challenge despite the available wealth of data, such as evolutionary information, and the wealth of tools to integrate that data. We describe DeepRank-Mut, a configurable framework designed to extract and learn from physicochemically relevant features of amino acids surrounding missense variants in 3D space. For each variant, various atomic and residue-level features are extracted from its structural environment, including sequence conservation scores of the surrounding amino acids, and stored in multi-channel 3D voxel grids which are then used to train a 3D convolutional neural network (3D-CNN). The resultant model gives a probabilistic estimate of whether a given input variant is disease-causing or benign. We find that the performance of our 3D-CNN model, on independent test datasets, is comparable to other widely used resources which also combine sequence and structural features. Based on the 10-fold cross-validation experiments, we achieve an average accuracy of 0.77 on the independent test datasets. We discuss the contribution of the variant neighborhood in the model's predictive power, in addition to the impact of individual features on the model's performance. Two key features: evolutionary information of residues in the variant neighborhood and their solvent accessibilities were observed to influence the predictions. We also highlight how predictions are impacted by the underlying disease mechanisms of missense mutations and offer insights into understanding these to improve pathogenicity predictions. Our study presents aspects to take into consideration when adopting deep learning approaches for protein structure-guided pathogenicity predictions.

Out-of-Sample R 2: Estimation and Inference

Out-of-sample prediction is the acid test of predictive models, yet an independent test dataset is often not available for assessment of the prediction error. For this reason, out-of-sample performance is commonly estimated using data splitting algorithms such as cross-validation or the bootstrap. For quantitative outcomes, the ratio of variance explained to total variance can be summarized by the coefficient of determination or in-sample R 2, which is easy to interpret and to compare across different outcome variables. As opposed to in-sample R 2, out-of-sample R 2 has not been well defined and the variability on out-of-sample R ̂ 2 has been largely ignored. Usually only its point estimate is reported, hampering formal comparison of predictability of different outcome variables. Here we explicitly define out-of-sample R 2 as a comparison of two predictive models, provide an unbiased estimator and exploit recent theoretical advances on uncertainty of data splitting estimates to provide a standard error for R ̂ 2 . The performance of the estimators for R 2 and its standard error are investigated in a simulation study. We demonstrate our new method by constructing confidence intervals and comparing models for prediction of quantitative Brassica napus and Zea mays phenotypes based on gene expression data. Our method is available in the R-package oosse.

Reconciling Pavement Condition Data from Connected Vehicles with the International Roughness Index from Standard Monitoring Equipment Using Physics-Integrated Machine Learning

This study investigates the possibility for augmenting infrequent standard pavement survey data with spatiotemporally continuous sensor data crowdsourced from connected vehicles. A framework is proposed to leverage physics-integrated machine learning (ML) models to reconcile the two disparate data sources. The proposed approach is demonstrated for the International Roughness Index (IRI). Using the quarter-car simulation, vertical acceleration data records were generated based on random stratified sampling from probability distributions of vehicle suspension properties at different speed levels on elevation profiles collected from various pavement sections. An ML model was pretrained to approximate the quarter-car simulation. Using transfer learning, the parameters of the outer layers of the model were finetuned to adapt model output to the measured values, while the parameters of the inner layers were fixed to preserve the embedded knowledge of the physical behavior of the suspension. The resulting physics-integrated ML model can predict the standard IRI value of a pavement section using speed, suspension properties, and features of the acceleration response from vehicles traversing that road section. The evaluation results on independent training and test datasets showed high accuracy, precision, and generalization capability, especially on smoother road profiles. This is attributed to the limited training data for rougher roads. After averaging the predicted IRI from all vehicles passing a certain road segment, a correction function is used to rectify the model’s under-predicted output and estimate the actual IRI value for that segment. High-frequency data resulting from the proposed approach can enhance pavement design, maintenance management, and asset management practices.

Image quality assessment using deep learning in high b-value diffusion-weighted breast MRI

The objective of this IRB approved retrospective study was to apply deep learning to identify magnetic resonance imaging (MRI) artifacts on maximum intensity projections (MIP) of the breast, which were derived from diffusion weighted imaging (DWI) protocols. The dataset consisted of 1309 clinically indicated breast MRI examinations of 1158 individuals (median age [IQR]: 50 years [16.75 years]) acquired between March 2017 and June 2020, in which a DWI sequence with a high b-value equal to 1500 s/mm2 was acquired. From these, 2D MIP images were computed and the left and right breast were cropped out as regions of interest (ROI). The presence of MRI image artifacts on the ROIs was rated by three independent observers. Artifact prevalence in the dataset was 37% (961 out of 2618 images). A DenseNet was trained with a fivefold cross-validation to identify artifacts on these images. In an independent holdout test dataset (n = 350 images) artifacts were detected by the neural network with an area under the precision-recall curve of 0.921 and a positive predictive value of 0.981. Our results show that a deep learning algorithm is capable to identify MRI artifacts in breast DWI-derived MIPs, which could help to improve quality assurance approaches for DWI sequences of breast examinations in the future.

IAMPCN: a deep-learning approach for identifying antimicrobial peptides and their functional activities.

Antimicrobial peptides (AMPs) are short peptides that play crucial roles in diverse biological processes and have various functional activities against target organisms. Due to the abuse of chemical antibiotics and microbial pathogens' increasing resistance to antibiotics, AMPs have the potential to be alternatives to antibiotics. As such, the identification of AMPs has become a widely discussed topic. A variety of computational approaches have been developed to identify AMPs based on machine learning algorithms. However, most of them are not capable of predicting the functional activities of AMPs, and those predictors that can specify activities only focus on a few of them. In this study, we first surveyed 10 predictors that can identify AMPs and their functional activities in terms of the features they employed and the algorithms they utilized. Then, we constructed comprehensive AMP datasets and proposed a new deep learning-based framework, iAMPCN (identification of AMPs based on CNNs), to identify AMPs and their related 22 functional activities. Our experiments demonstrate that iAMPCN significantly improved the prediction performance of AMPs and their corresponding functional activities based on four types of sequence features. Benchmarking experiments on the independent test datasets showed that iAMPCN outperformed a number of state-of-the-art approaches for predicting AMPs and their functional activities. Furthermore, we analyzed the amino acid preferences of different AMP activities and evaluated the model on datasets of varying sequence redundancy thresholds. To facilitate the community-wide identification of AMPs and their corresponding functional types, we have made the source codes of iAMPCN publicly available at https://github.com/joy50706/iAMPCN/tree/master. We anticipate that iAMPCN can be explored as a valuable tool for identifying potential AMPs with specific functional activities for further experimental validation.

Biomarker Discovery in Rare Malignancies: Development of a miRNA Signature for RDEB-cSCC

Machine learning has been proven to be a powerful tool in the identification of diagnostic tumor biomarkers but is often impeded in rare cancers due to small patient numbers. In patients suffering from recessive dystrophic epidermolysis bullosa (RDEB), early-in-life development of particularly aggressive cutaneous squamous-cell carcinomas (cSCCs) represents a major threat and timely detection is crucial to facilitate prompt tumor excision. As miRNAs have been shown to hold great potential as liquid biopsy markers, we characterized miRNA signatures derived from cultured primary cells specific for the potential detection of tumors in RDEB patients. To address the limitation in RDEB-sample accessibility, we analyzed the similarity of RDEB miRNA profiles with other tumor entities derived from the Cancer Genome Atlas (TCGA) repository. Due to the similarity in miRNA expression with RDEB-SCC, we used HN-SCC data to train a tumor prediction model. Three models with varying complexity using 33, 10 and 3 miRNAs were derived from the elastic net logistic regression model. The predictive performance of all three models was determined on an independent HN-SCC test dataset (AUC-ROC: 100%, 83% and 96%), as well as on cell-based RDEB miRNA-Seq data (AUC-ROC: 100%, 100% and 91%). In addition, the ability of the models to predict tumor samples based on RDEB exosomes (AUC-ROC: 100%, 93% and 100%) demonstrated the potential feasibility in a clinical setting. Our results support the feasibility of this approach to identify a diagnostic miRNA signature, by exploiting publicly available data and will lay the base for an improvement of early RDEB-SCC detection.

Automated Interpretation of Clinical Electroencephalograms Using Artificial Intelligence

Electroencephalograms (EEGs) are a fundamental evaluation in neurology but require special expertise unavailable in many regions of the world. Artificial intelligence (AI) has a potential for addressing these unmet needs. Previous AI models address only limited aspects of EEG interpretation such as distinguishing abnormal from normal or identifying epileptiform activity. A comprehensive, fully automated interpretation of routine EEG based on AI suitable for clinical practice is needed. To develop and validate an AI model (Standardized Computer-based Organized Reporting of EEG-Artificial Intelligence [SCORE-AI]) with the ability to distinguish abnormal from normal EEG recordings and to classify abnormal EEG recordings into categories relevant for clinical decision-making: epileptiform-focal, epileptiform-generalized, nonepileptiform-focal, and nonepileptiform-diffuse. In this multicenter diagnostic accuracy study, a convolutional neural network model, SCORE-AI, was developed and validated using EEGs recorded between 2014 and 2020. Data were analyzed from January 17, 2022, until November 14, 2022. A total of 30 493 recordings of patients referred for EEG were included into the development data set annotated by 17 experts. Patients aged more than 3 months and not critically ill were eligible. The SCORE-AI was validated using 3 independent test data sets: a multicenter data set of 100 representative EEGs evaluated by 11 experts, a single-center data set of 9785 EEGs evaluated by 14 experts, and for benchmarking with previously published AI models, a data set of 60 EEGs with external reference standard. No patients who met eligibility criteria were excluded. Diagnostic accuracy, sensitivity, and specificity compared with the experts and the external reference standard of patients' habitual clinical episodes obtained during video-EEG recording. The characteristics of the EEG data sets include development data set (N = 30 493; 14 980 men; median age, 25.3 years [95% CI, 1.3-76.2 years]), multicenter test data set (N = 100; 61 men, median age, 25.8 years [95% CI, 4.1-85.5 years]), single-center test data set (N = 9785; 5168 men; median age, 35.4 years [95% CI, 0.6-87.4 years]), and test data set with external reference standard (N = 60; 27 men; median age, 36 years [95% CI, 3-75 years]). The SCORE-AI achieved high accuracy, with an area under the receiver operating characteristic curve between 0.89 and 0.96 for the different categories of EEG abnormalities, and performance similar to human experts. Benchmarking against 3 previously published AI models was limited to comparing detection of epileptiform abnormalities. The accuracy of SCORE-AI (88.3%; 95% CI, 79.2%-94.9%) was significantly higher than the 3 previously published models (P < .001) and similar to human experts. In this study, SCORE-AI achieved human expert level performance in fully automated interpretation of routine EEGs. Application of SCORE-AI may improve diagnosis and patient care in underserved areas and improve efficiency and consistency in specialized epilepsy centers.

A machine-learning model that incorporates CD45 surface expression predicts hematopoietic progenitor cell recovery after freeze–thaw

Background aimsSufficient doses of viable CD34+ (vCD34) hematopoietic progenitor cells (HPCs) are crucial for engraftment. Additional-day apheresis collections can compensate for potential loss during cryopreservation but incur high cost and additional risk. To aid predicting such losses for clinical decision support, we developed a machine-learning model using variables obtainable on the day of collection. MethodsIn total, 370 consecutive autologous HPCs, apheresis-collected since 2014 at the Children's Hospital of Philadelphia, were retrospectively reviewed. Flow cytometry was used to assess vCD34% on fresh products and thawed quality control vials. The ratio of vCD34% thawed to fresh, which we call “post-thaw index,” was used as an outcome measure, with a “poor” post-thaw index defined as <70%. HPC CD45 normalized mean fluorescence intensity (MFI) was calculated by dividing CD45 MFI of HPCs to the CD45 MFI of lymphocytes in the same sample. We trained XGBoost, k-nearest neighbor and random forest models for the prediction and calibrated the best model to minimize falsely-reassuring predictions. ResultsIn total, 63 of 370 (17%) products had a poor post-thaw index. The best model was XGBoost, with an area under the receiver operator curve of 0.83 evaluated on an independent test data set. The most important predictor for a poor post-thaw index was the HPC CD45 normalized MFI. Transplants after 2015, based on the lower of the two vCD34% values, showed faster engraftment than older transplants, which were based on fresh vCD34% only (average 10.6 vs 11.7 days, P = 0.0006). ConclusionsTransplants taking into account post-thaw vCD34% improved engraftment time in our patients; however, it came at the cost of unnecessary multi-day collections. The results from applying our predictive algorithm retrospectively to our data suggest that more than one-third of additional-day collections could have been avoided. Our investigation also identified CD45 nMFI as a novel marker for assessing HPC health post-thaw.

Development of gene model combined with machine learning technology to predict for advanced atherosclerotic plaques

BackgroundAtherosclerosis, as a major cause of stroke, is responsible for a quarter of deaths worldwide. In particular, rupture of late-stage plaques in large vessels such as the carotid artery can lead to serious cardiovascular disease. The aim of our study was to establish a genetic model combined with machining leaning techniques to screen out gene signatures and predict for advanced atherosclerosis plaques. MethodsThe microarray dataset GSE28829 and GSE43292 which were publicly obtained from the Gene Expression Omnibus database were utilized to screen for potential predictive genes. Differentially expressed genes (DEGs) were identified by using the “limma” R package. Gene Ontology (GO) and Kyoto Encyclopedia of Genes Genomes (KEGG) analyses of these DEGs were performed by Metascape. Later, Random Forest (RF) algorithm was applied to further screen out top-30 genes which contribute the most. The expression data of top 30-DEGs were converted into a “Gene Score”. Finally, we developed a model based on artificial neural network (ANN) to predict advanced atherosclerotic plaques. The model later was validated in an independent test dataset GSE104140. ResultsA total of 176 DEGs were identified in the training datasets. GO and KEGG enrichment analysis revealed that these genes were enriched in leukocyte-mediated immune response, cytokine- cytokine interactions, and immunoinflammatory signaling. Further, top-30 genes (including 25 upregulated and 5 downregulated DEGs) were screened as predictors by RF algorithm. The predictive model was developed with a significantly predictive value (AUC = 0.913) in the training datasets, and was validated with an independent dataset GSE104140 (AUC = 0.827). ConclusionIn present study, our prediction model was established and showed satisfactory predictive power in both training and test datasets. In addition, this is the first study adopted bioinformatics methods combined with machine learning techniques (RF and ANN) to explore and predict for the advanced atherosclerotic plaques. However, further investigations were needed to verify the screened DEGs and predictive effectiveness of this model.

Preserving privacy in surgical video analysis using a deep learning classifier to identify out-of-body scenes in endoscopic videos

Surgical video analysis facilitates education and research. However, video recordings of endoscopic surgeries can contain privacy-sensitive information, especially if the endoscopic camera is moved out of the body of patients and out-of-body scenes are recorded. Therefore, identification of out-of-body scenes in endoscopic videos is of major importance to preserve the privacy of patients and operating room staff. This study developed and validated a deep learning model for the identification of out-of-body images in endoscopic videos. The model was trained and evaluated on an internal dataset of 12 different types of laparoscopic and robotic surgeries and was externally validated on two independent multicentric test datasets of laparoscopic gastric bypass and cholecystectomy surgeries. Model performance was evaluated compared to human ground truth annotations measuring the receiver operating characteristic area under the curve (ROC AUC). The internal dataset consisting of 356,267 images from 48 videos and the two multicentric test datasets consisting of 54,385 and 58,349 images from 10 and 20 videos, respectively, were annotated. The model identified out-of-body images with 99.97% ROC AUC on the internal test dataset. Mean ± standard deviation ROC AUC on the multicentric gastric bypass dataset was 99.94 ± 0.07% and 99.71 ± 0.40% on the multicentric cholecystectomy dataset, respectively. The model can reliably identify out-of-body images in endoscopic videos and is publicly shared. This facilitates privacy preservation in surgical video analysis.

XGBoost framework with feature selection for the prediction of RNA N5-methylcytosine sites

5-methylcytosine (m5C) is indeed a critical post-transcriptional alteration that is widely present in various kinds of RNAs and is crucial to the fundamental biological processes. By correctly identifying the m5C-methylation sites on RNA, clinicians can more clearly comprehend the precise function of these m5C-sites in different biological processes. Due to their effectiveness and affordability, computational methods have received greater attention over the last few years for the identification of methylation sites in various species. To precisely identify RNA m5C locations in five different species including Homo sapiens, Arabidopsis thaliana, Mus musculus, Drosophila melanogaster, and Danio rerio, we proposed a more effective and accurate model named m5C-pred. To create m5C-pred, five distinct feature encoding techniques were combined to extract features from the RNA sequence, and then we used SHapley Additive exPlanations to choose the best features among them, followed by XGBoost as a classifier. We applied the novel optimization method called Optuna to quickly and efficiently determine the best hyperparameters. Finally, the proposed model was evaluated using independent test datasets, and we compared the results with the previous methods. Our approach, m5C- pred, is anticipated to be useful for accurately identifying m5C sites, outperforming the currently available state-of-the-art techniques.

A convolutional network and attention mechanism-based approach to predict protein–RNA binding residues

Protein–RNA interactions play a key role in various biological cellular processes, and many experimental and computational studies have been initiated to analyze their interactions. However, experimental determination is quite complex and expensive. Therefore, researchers have worked to develop efficient computational tools to detect protein–RNA binding residues. The accuracy of existing methods is limited by the features of the target and the performance of the computational models; there remains room for improvement. To solve the problem of the accurate detection of protein–RNA binding residues, we propose a convolutional network model named PBRPre based on improved MobileNet. First, by extracting the position information of the target complex and the 3-mer amino acid feature data, the position-specific scoring matrix (PSSM) is improved by using spatial neighbor smoothing processing and discrete wavelet transform to fully exploit the spatial structure information of the target and enrich the feature dataset. Second, the deep learning model MobileNet is used to integrate and optimize the potential features in the target complexes; then, by introducing the Vision Transformer (ViT) network classification layer, the deep-level information of the target is mined to enhance the processing ability of the model for global information and to improve the detection accuracy of the classifiers. The results show that the AUC value of the model can reach 0.866 in the independent testing dataset, which shows that PBRPre can effectively realize the detection of protein–RNA binding residues. All datasets and resource codes of PBRPre are available at https://github.com/linglewu/PBRPre for academic use.

Radiomics for therapy-specific head and neck squamous cell carcinoma survival prognostication (part I)

BackgroundTreatment plans for squamous cell carcinoma of the head and neck (SCCHN) are individually decided in tumor board meetings but some treatment decision-steps lack objective prognostic estimates. Our purpose was to explore the potential of radiomics for SCCHN therapy-specific survival prognostication and to increase the models’ interpretability by ranking the features based on their predictive importance.MethodsWe included 157 SCCHN patients (male, 119; female, 38; mean age, 64.39 ± 10.71 years) with baseline head and neck CT between 09/2014 and 08/2020 in this retrospective study. Patients were stratified according to their treatment. Using independent training and test datasets with cross-validation and 100 iterations, we identified, ranked and inter-correlated prognostic signatures using elastic net (EN) and random survival forest (RSF). We benchmarked the models against clinical parameters. Inter-reader variation was analyzed using intraclass-correlation coefficients (ICC).ResultsEN and RSF achieved top prognostication performances of AUC = 0.795 (95% CI 0.767–0.822) and AUC = 0.811 (95% CI 0.782–0.839). RSF prognostication slightly outperformed the EN for the complete (ΔAUC 0.035, p = 0.002) and radiochemotherapy (ΔAUC 0.092, p < 0.001) cohort. RSF was superior to most clinical benchmarking (p ≤ 0.006). The inter-reader correlation was moderate or high for all features classes (ICC ≥ 0.77 (± 0.19)). Shape features had the highest prognostic importance, followed by texture features.ConclusionsEN and RSF built on radiomics features may be used for survival prognostication. The prognostically leading features may vary between treatment subgroups. This warrants further validation to potentially aid clinical treatment decision making in the future.

Progressively refined deep joint registration segmentation (ProRSeg) of gastrointestinal organs at risk: Application to MRI and cone-beam CT.

Adaptive radiation treatment (ART) for locally advanced pancreatic cancer (LAPC) requires consistently accurate segmentation of the extremely mobile gastrointestinal (GI) organs at risk (OAR) including the stomach, duodenum, large and small bowel. Also, due to lack of sufficiently accurate and fast deformable image registration (DIR), accumulated dose to the GI OARs is currently only approximated, further limiting the ability to more precisely adapt treatments. Develop a 3-D Progressively refined joint Registration-Segmentation (ProRSeg) deep network to deformably align and segment treatment fraction magnetic resonance images (MRI)s, then evaluate segmentation accuracy, registration consistency, and feasibility for OAR doseaccumulation. ProRSeg was trained using five-fold cross-validation with 110 T2-weighted MRI acquired at five treatment fractions from 10 different patients, taking care that same patient scans were not placed in training and testing folds. Segmentation accuracy was measured using Dice similarity coefficient (DSC) and Hausdorff distance at 95th percentile (HD95). Registration consistency was measured using coefficient of variation (CV) in displacement of OARs. Statistical comparison to other deep learning and iterative registration methods were done using the Kruskal-Wallis test, followed by pair-wise comparisons with Bonferroni correction applied for multiple testing. Ablation tests and accuracy comparisons against multiple methods were done. Finally, applicability of ProRSeg to segment cone-beam CT (CBCT) scans was evaluated on a publicly available dataset of 80 scans using five-foldcross-validation. ProRSeg processed 3D volumes (128 × 192 × 128) in 3 s on a NVIDIA Tesla V100 GPU. It's segmentations were significantly more accurate ( ) than compared methods, achieving a DSC of 0.94 ±0.02 for liver, 0.88±0.04 for large bowel, 0.78±0.03 for small bowel and 0.82±0.04 for stomach-duodenum from MRI. ProRSeg achieved a DSC of 0.72±0.01 for small bowel and 0.76±0.03 for stomach-duodenum from public CBCT dataset. ProRSeg registrations resulted in the lowest CV in displacement (stomach-duodenum : 0.75%, : 0.73%, and : 0.81%; small bowel : 0.80%, : 0.80%, and : 0.68%; large bowel : 0.71%, : 0.81%, and : 0.75%). ProRSeg based dose accumulation accounting for intra-fraction (pre-treatment to post-treatment MRI scan) and inter-fraction motion showed that the organ dose constraints were violated in four patients for stomach-duodenum and for three patients for small bowel. Study limitations include lack of independent testing and ground truth phantom datasets to measure dose accumulationaccuracy. ProRSeg produced more accurate and consistent GI OARs segmentation and DIR of MRI and CBCTs compared to multiple methods. Preliminary results indicates feasibility for OAR dose accumulation usingProRSeg.

A convolutional neural network-based, quantitative complete blood count scattergram-mapping framework promptly screens acute promyelocytic leukemia with high sensitivity.

Acute promyelocytic leukemia (APL) is a subtype of acute myeloid leukemia (AML) characterized by its rapidly progressive and fatal clinical course if untreated, although it is curable if treated in a timely manner. Promptly screening patients who have results that are suspicious for APL is vital to overcome early death. The authors developed an innovative framework consisting of ResNet-18, a convolutional neural network architecture, with the objective of quantitatively mapping a complete blood count (CBC) scattergram to quickly and robustly indicate a probable susceptibility to APL. Three hundred and twenty scattergrams of the white blood cell differential channel from 51 patients with APL, 510 scattergrams from 105 patients who had non-APL AML, and 320 scattergrams from 320 healthy controls were randomly stratified at a ratio of 4:1 and split into training and testing data sets to accomplish five-fold cross-validation. Both the area under the curve and the average precision of >0.99 were achieved in each fold. Three hundred four of the 320 APL scattergrams (95%) were correctly flagged by the model, which outcompeted the CBC review rules recommended by the International Society of Laboratory Hematology (all p<.001). External validation based on an independent testing data set that included 56 scattergrams from 31 patients with APL, 56 scattergrams from 55 patients with non-APL AML, and 64 scattergrams from 64 healthy controls also confirmed the sensitivity and specificity of the framework. To the authors' knowledge, their convolutional neural network-based framework is the first to use scattergram output from routine CBC analysis to map suspicious APL early with outstanding sensitivity, specificity, and precision. The authors also describe a new CBC workflow incorporating this framework upstream of the morphologic review, which would provide the earliest flag for APL. The authors propose an innovative way to visualize complete blood counts (CBCs) by mapping the difference in white blood cell counts using automated CBC analysis to identify potential acute promyelocytic leukemia (APL) using a convolutional neural network (CNN), which can eliminate the potential pitfalls of manual observation. Analyses of an unprecedented, realistic data set validated that the quantitative relationship between the CBC scattergram and an APL abnormality is highly consistent. This is the first study to date focusing on screening for APL using scattergrams of the difference in white blood cell counts from routine CBC tests and has significant clinical relevance. The authors recommend using this method even before analyzing cell images, which could provide the earliest way to screen for APL in a sensitive and accurateway.

A contingency approach for time-cost trade-off in construction projects based on machine learning techniques

PurposeTime-cost trade-off is normal conduct in construction projects when projects are expectedly late for delivery. Existing research on time-cost trade-off strategic management mostly focused on the technical calculation towards the optimal combination of activities to be accelerated, while the managerial aspects are mostly neglected. This paper aims to understand the managerial efforts necessary to prepare construction projects ready for an upcoming trade-off implementation.Design/methodology/approachA preliminary list of critical factors was first identified from the literature and verified by a Delphi survey. Quantitative data was then collected by a questionnaire survey to first shortlist the preliminary factors and quantify the predictive model with different machine learning algorithms, i.e. k-nearest neighbours (kNN), radial basis function (RBF), multiplayer perceptron (MLP), multinomial logistic regression (MLR), naïve Bayes classifier (NBC) and Bayesian belief networks (BBNs).FindingsThe model's independent variable importance ranking revealed that the top challenges faced were the realism of contractual obligation, contractor planning and control and client management and monitoring. Among the tested machine learning algorithms, multilayer perceptron was demonstrated to be the most suitable in this case. This model accuracy reached 96.5% with the training dataset and 95.6% with an independent test dataset and could be used as the contingency approach for time-cost trade-offs.Originality/valueThe identified factor list contributed to the theoretical explanation of the failed implementation in general and practical managerial improvement to better avoid such failure. In addition, the established predictive model provided an ad-hoc early warning and diagnostic tool to better ensure time-cost implementation success.