Independent Test Set Research Articles

Rapid Measurement of Total Saponins, Mannitol, and Naringenin in Dendrobium officinale by Near-Infrared Spectroscopy and Chemometrics

Dendrobium officinale has drawn increasing attention as a dual-use plant with herbal medicine and food applications. The efficient quality evaluation of D. officinale is essential to ensuring its nutritional and pharmaceutical value. Given that traditional analytical methods are generally time-consuming, expensive, and laborious, this study developed a rapid and efficient approach to assess the quality of D. officinale from different geographical origins by near-infrared (NIR) spectroscopy and chemometrics. Total saponins, mannitol, and naringenin were utilized as quality indicators. Two wavelength selection methods, namely, uninformative variable elimination and competitive adaptive reweighted sampling (CARS), were utilized to enhance the prediction accuracy of the quantification model. Moreover, multiple spectral pretreatment methods were applied for model optimization. Results indicated that the partial least squares (PLS) model constructed based on the wavelengths selected by CARS exhibited superior performance in predicting the contents of the quality indicators. The coefficient of determination (RP2) and root mean square error (RMSEP) in the independent test sets were 0.8949 and 0.1250 g kg−1 for total saponins, 0.9664 and 0.2192 g kg−1 for mannitol, and 0.8570 and 0.003159 g kg−1 for naringenin, respectively. This study revealed that NIR spectroscopy and the CARS-PLS model could be used as a rapid and accurate technique to evaluate the quality of D. officinale.

PermuteDDS: a permutable feature fusion network for drug-drug synergy prediction.

Drug combination therapies have shown promise in clinical cancer treatments. However, it is hard to experimentally identify all drug combinations for synergistic interaction even with high-throughput screening due to the vast space of potential combinations. Although a number of computational methods for drug synergy prediction have proven successful in narrowing down this space, fusing drug pairs and cell line features effectively still lacks study, hindering current algorithms from understanding the complex interaction between drugs and cell lines. In this paper, we proposed a Permutable feature fusion network for Drug-Drug Synergy prediction, named PermuteDDS. PermuteDDS takes multiple representations of drugs and cell lines as input and employs a permutable fusion mechanism to combine drug and cell line features. In experiments, PermuteDDS exhibits state-of-the-art performance on two benchmark data sets. Additionally, the results on independent test set grouped by different tissues reveal that PermuteDDS has good generalization performance. We believed that PermuteDDS is an effective and valuable tool for identifying synergistic drug combinations. It is publicly available at https://github.com/littlewei-lazy/PermuteDDS . First, this paper proposes a permutable feature fusion network for predicting drug synergy termed PermuteDDS, which extract diverse information from multiple drug representations and cell line representations. Second, the permutable fusion mechanism combine the drug and cell line features by integrating information of different channels, enabling the utilization of complex relationships between drugs and cell lines. Third, comparative and ablation experiments provide evidence of the efficacy of PermuteDDS in predicting drug-drug synergy.

PeSTo-Carbs: Geometric Deep Learning for Prediction of Protein-Carbohydrate Binding Interfaces.

The Protein Structure Transformer (PeSTo), a geometric transformer, has exhibited exceptional performance in predicting protein-protein binding interfaces and distinguishing interfaces with nucleic acids, lipids, small molecules, and ions. In this study, we introduce PeSTo-Carbs, an extension of PeSTo specifically engineered to predict protein-carbohydrate binding interfaces. We evaluate the performance of this approach using independent test sets and compare them with those of previous methods. Furthermore, we highlight the model's capability to specialize in predicting interfaces involving cyclodextrins, a biologically and pharmaceutically significant class of carbohydrates. Our method consistently achieves remarkable accuracy despite the scarcity of available structural data for cyclodextrins.

Identification of Multiclass Pediatric Low-Grade Neuroepithelial Tumor Molecular Subtype with ADC MR Imaging and Machine Learning.

Molecular biomarker identification increasingly influences the treatment planning of pediatric low-grade neuroepithelial tumors (PLGNTs). We aimed to develop and validate a radiomics-based ADC signature predictive of the molecular status of PLGNTs. In this retrospective bi-institutional study, we searched the PACS for baseline brain MRIs from children with PLGNTs. Semiautomated tumor segmentation on ADC maps was performed using the semiautomated level tracing effect tool with 3D Slicer. Clinical variables, including age, sex, and tumor location, were collected from chart review. The molecular status of tumors was derived from biopsy. Multiclass random forests were used to predict the molecular status and fine-tuned using a grid search on the validation sets. Models were evaluated using independent and unseen test sets based on the combined data, and the area under the receiver operating characteristic curve (AUC) was calculated for the prediction of 3 classes: KIAA1549-BRAF fusion, BRAF V600E mutation, and non-BRAF cohorts. Experiments were repeated 100 times using different random data splits and model initializations to ensure reproducible results. Two hundred ninety-nine children from the first institution and 23 children from the second institution were included (53.6% male; mean, age 8.01 years; 51.8% supratentorial; 52.2% with KIAA1549-BRAF fusion). For the 3-class prediction using radiomics features only, the average test AUC was 0.74 (95% CI, 0.73-0.75), and using clinical features only, the average test AUC was 0.67 (95% CI, 0.66-0.68). The combination of both radiomics and clinical features improved the AUC to 0.77 (95% CI, 0.75-0.77). The diagnostic performance of the per-class test AUC was higher in identifying KIAA1549-BRAF fusion tumors among the other subgroups (AUC = 0.81 for the combined radiomics and clinical features versus 0.75 and 0.74 for BRAF V600E mutation and non-BRAF, respectively). ADC values of tumor segmentations have differentiative signals that can be used for training machine learning classifiers for molecular biomarker identification of PLGNTs. ADC-based pretherapeutic differentiation of the BRAF status of PLGNTs has the potential to avoid invasive tumor biopsy and enable earlier initiation of targeted therapy.

Posterior circulation ischemic stroke: radiomics-based machine learning approach to identify onset time from magnetic resonance imaging.

Posterior circulation ischemic stroke (PCIS) possesses unique features. However, previous studies have primarily or exclusively relied on anterior circulation stroke cases to build machine learning (ML) models for predicting onset time. To date, there is no research reporting the effectiveness and stability of ML in identifying PCIS onset time. We aimed to build diffusion-weighted imaging-based ML models to identify the onset time of PCIS patients. Consecutive PCIS patients within 24h of definite symptom onset were included (112 in the training set and 49 in the independent test set). Images were processed as follows: volume of interest segmentation, image feature extraction, and feature selection. Five ML models, naïve Bayes, logistic regression, tree ensemble, k-nearestneighbor, and random forest, were built based on the training set to estimate the stroke onset time (binary classification: ≤ 4.5h or > 4.5h). Relative standard deviations (RSD), receiver operating characteristic (ROC) curves, and the calibration plot was performed to evaluate the stability and performance of the five models. The random forest model had the best performance in the test set, with the highest area under the curve (AUC, 0.840; 95% CI: 0.706, 0.974). This model also achieved the highest accuracy, sensitivity, specificity, positive predictive value, and negative predictive value (83.7%, 64.3%, 91.4%, 75.0%, and 86.5%, respectively). Furthermore, the model had high stability (RSD = 0.0094). The PCIS case-based ML model was effective for estimating the symptom onset time and achieved considerably high specificity and stability.

Diagnostic performance of a deep-learning model using 18F-FDG PET/CT for evaluating recurrence after radiation therapy in patients with lung cancer.

We developed a deep learning model for distinguishing radiation therapy (RT)-related changes and tumour recurrence in patients with lung cancer who underwent RT, and evaluated its performance. We retrospectively recruited 308 patients with lung cancer with RT-related changes observed on 18F-fluorodeoxyglucose positron emission tomography-computed tomography (18F-FDG PET/CT) performed after RT. Patients were labelled as positive or negative for tumour recurrence through histologic diagnosis or clinical follow-up after 18F-FDG PET/CT. A two-dimensional (2D) slice-based convolutional neural network (CNN) model was created with a total of 3329 slices as input, and performance was evaluated with five independent test sets. For the five independent test sets, the area under the curve (AUC) of the receiver operating characteristic curve, sensitivity, and specificity were in the range of 0.98-0.99, 95-98%, and 87-95%, respectively. The region determined by the model was confirmed as an actual recurred tumour through the explainable artificial intelligence (AI) using gradient-weighted class activation mapping (Grad-CAM). The 2D slice-based CNN model using 18F-FDG PET imaging was able to distinguish well between RT-related changes and tumour recurrence in patients with lung cancer.

Identification of 6-methyladenosine sites using novel feature encoding methods and ensemble models.

N6-methyladenosine (6mA) is the most common internal modification in eukaryotic mRNA. Mass spectrometry and site-directed mutagenesis, two of the most common conventional approaches, have been shown to be laborious and challenging. In recent years, there has been a rising interest in analyzing RNA sequences to systematically investigate mutated locations. Using novel methods for feature development, the current work aimed to identify 6mA locations in RNA sequences. Following the generation of these novel features, they were used to train an ensemble of models using methods such as stacking, boosting, and bagging. The trained ensemble models were assessed using an independent test set and k-fold cross validation. When compared to baseline predictors, the suggested model performed better and showed improved ratings across the board for key measures of accuracy.

Abstract CT203: Multi-omic analysis of serial biopsies to inform biomarkers of sensitivity to olaparib and durvalumab in patients with metastatic BRCA-wildtype triple negative breast cancer (mTNBC)

Abstract Background: The phase 2 Adaptive Multi-Drug Treatment of Evolving Cancers (AMTEC) trial (NCT03801369) evaluated the efficacy of the combination of the PARP inhibitor Olaparib (Ola) and the PD-L1 inhibitor Durvalumab (Durva) in participants with BRCA-wildtype mTNBC. The combination was found to be effective with a median progression free survival (mPFS) of 5.5 months (AACR 2022). Here we report on updated biomarker analyses from paired biopsies (Bx) pre- and on-Ola therapy from 14 AMTEC participants. Methods: AMTEC participants undergo a pre-Ola Bx (Bx1), then one (28-day) cycle of Ola monotherapy with a repeat on-Ola Bx (Bx2) before adding Durva to Ola. Multi-omic profiling of DNA, RNA, and protein signals in Bx1 and Bx2 using Deep candidate gene sequencing, RNAseq, Nanostring Digital Spatial Profiling (DSP), and multiplex immunohistochemistry (mIHC) was correlated with clinical outcomes to identify predictors of Ola + Durva sensitivity, and adaptive resistance to PARPi therapy. Results: We identified over 30 biomarkers, with the optimal predictive value arising from Bx2. Key potential biomarkers were confirmed in an independent test set. Markers of good prognosis: • Basal Immune Activated (BLIA)/Basal Immune Suppressed (BLIS subtype) - A BLIA subtype on Bx2 was associated with a mPFS of 7.36 months compared to 1.71 months for BLIS/Luminal androgen receptor subtypes (p=0.001) • Multi-omic immune composite - A positive immune consensus signature (RNA, DSP, mIHC) on Bx2 was associated with a mPFS of 7.36 months compared to 2.29 months (p=0.005) • B cell activation signature - We developed a novel B-cell RNA activation signature that was highly predictive of response in Bx2, with mPFS of 8.54 and 1.04 months in signature-positive and signature-negative tumors, respectively (p=0.0004) Markers of poor prognosis: • AKT/MAPK pathway - AKT pathway gene mutations in Bx2 with concomitant protein phosphorylation of pathway members correlated with a shorter mPFS of 2.4 months compared to 7.13 months for no pathway mutation/activation (p=0.03). MAPK pathway protein phosphorylation in Bx2 was a negative prognostic factor, with mPFS of 2.17 months (p=0.004) • Angiogenesis - An upregulated Hallmark of Cancer angiogenesis RNA signature on Bx2 was associated with a shorter mPFS of 1.94 months compared to 7.95 months with no angiogenesis upregulation (p=0.0009) The following markers did not show significant association with survival outcomes on AMTEC: • PD-L1 positivity by IHC (defined as >1% Tumor Proportion Score, 22C3 antibody) • Tumor mutational burden • HRD, as measured by Rad51 foci or by mutations in HR pathway members Conclusions: Findings highlight the value of paired Bxs to identify predictive biomarkers of PARPi + immune checkpoint inhibitor sensitivity. The striking predictive value of the BLIA/BLIS signature warrants evaluation in a larger trial. Emerging resistance mechanisms justify AMTEC trial expansion to include PARPi + MEKi or PARPi + AKTi in biomarker selected patients, which is now in progress. Citation Format: Zahi I. Mitri, Allison L. Creason, Jayne M. Stommel, Daniel Bottomly, Jeong Youn Lim, SMMART Clinical Trials Program, Christopher L. Corless, Shannon McWeeney, Gordon B. Mills. Multi-omic analysis of serial biopsies to inform biomarkers of sensitivity to olaparib and durvalumab in patients with metastatic BRCA-wildtype triple negative breast cancer (mTNBC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr CT203.

Detection and Grading of Radiographic Hand Osteoarthritis Using an Automated Machine Learning Platform.

Automated machine learning (autoML) platforms allow health care professionals to play an active role in the development of machine learning (ML) algorithms according to scientific or clinical needs. The aim of this study was to develop and evaluate such a model for automated detection and grading of distal hand osteoarthritis (OA). A total of 13,690 hand radiographs from 2,863 patients within the Swiss Cohort of Quality Management (SCQM) and an external control data set of 346 non-SCQM patients were collected and scored for distal interphalangeal OA (DIP-OA) using the modified Kellgren/Lawrence (K/L) score. Giotto (Learn to Forecast [L2F]) was used as an autoML platform for training two convolutional neural networks for DIP joint extraction and subsequent classification according to the K/L scores. A total of 48,892 DIP joints were extracted and then used to train the classification model. Heatmaps were generated independently of the platform. User experience of a web application as a provisional user interface was investigated by rheumatologists and radiologists. The sensitivity and specificity of this model for detecting DIP-OA were 79% and 86%, respectively. The accuracy for grading the correct K/L score was 75%, with a κ score of 0.76. The accuracy per DIP-OA class differed, with 86% for no OA (defined as K/L scores 0 and 1), 71% for a K/L score of 2, 46% for a K/L score of 3, and 67% for a K/L score of 4. Similar values were obtained in an independent external test set. Qualitative and quantitative user experience testing of the web application revealed a moderate to high demand for automated DIP-OA scoring among rheumatologists. Conversely, radiologists expressed a low demand, except for the use of heatmaps. AutoML platforms are an opportunity to develop clinical end-to-end ML algorithms. Here, automated radiographic DIP-OA detection is both feasible and usable, whereas grading among individual K/L scores (eg, for clinical trials) remains challenging.

Deep Learning-Based Automated Detection of Retinal Breaks and Detachments on Fundus Photography.

The purpose of this study was to develop a deep learning algorithm, to detect retinal breaks and retinal detachments on ultra-widefield fundus (UWF) optos images using artificial intelligence (AI). Optomap UWF images of the database were annotated to four groups by two retina specialists: (1) retinal breaks without detachment, (2) retinal breaks with retinal detachment, (3) retinal detachment without visible retinal breaks, and (4) a combination of groups 1 to 3. The fundus image data set was split into a training set and an independent test set following an 80% to 20% ratio. Image preprocessing methods were applied. An EfficientNet classification model was trained with the training set and evaluated with the test set. A total of 2489 UWF images were included into the dataset, resulting in a training set size of 2008 UWF images and a test set size of 481 images. The classification models achieved an area under the receiver operating characteristic curve (AUC) on the testing set of 0.975 regarding lesion detection, an AUC of 0.972 for retinal detachment and an AUC of 0.913 for retinal breaks. A deep learning system to detect retinal breaks and retinal detachment using UWF images is feasible and has a good specificity. This is relevant for clinical routine as there can be a high rate of missed breaks in clinics. Future clinical studies will be necessary to evaluate the cost-effectiveness of applying such an algorithm as an automated auxiliary tool in a large practices or tertiary referral centers. This study demonstrates the relevance of applying AI in diagnosing peripheral retinal breaks in clinical routine in UWF fundus images.

Fully automated deep learning system for osteoporosis screening using chest computed tomography images.

Osteoporosis, a disease stemming from bone metabolism irregularities, affects approximately 200 million people worldwide. Timely detection of osteoporosis is pivotal in grappling with this public health challenge. Deep learning (DL), emerging as a promising methodology in the field of medical imaging, holds considerable potential for the assessment of bone mineral density (BMD). This study aimed to propose an automated DL framework for BMD assessment that integrates localization, segmentation, and ternary classification using various dominant convolutional neural networks (CNNs). In this retrospective study, a cohort of 2,274 patients underwent chest computed tomography (CT) was enrolled from January 2022 to June 2023 for the development of the integrated DL system. The study unfolded in 2 phases. Initially, 1,025 patients were selected based on specific criteria to develop an automated segmentation model, utilizing 2 VB-Net networks. Subsequently, a distinct cohort of 902 patients was employed for the development and testing of classification models for BMD assessment. Then, 3 distinct DL network architectures, specifically DenseNet, ResNet-18, and ResNet-50, were applied to formulate the 3-classification BMD assessment model. The performance of both phases was evaluated using an independent test set consisting of 347 individuals. Segmentation performance was evaluated using the Dice similarity coefficient; classification performance was appraised using the receiver operating characteristic (ROC) curve. Furthermore, metrics such as the area under the curve (AUC), accuracy, and precision were meticulously calculated. In the first stage, the automatic segmentation model demonstrated excellent segmentation performance, with mean Dice surpassing 0.93 in the independent test set. In the second stage, both the DenseNet and ResNet-18 demonstrated excellent diagnostic performance in detecting bone status. For osteoporosis, and osteopenia, the AUCs were as follows: DenseNet achieved 0.94 [95% confidence interval (CI): 0.91-0.97], and 0.91 (95% CI: 0.87-0.94), respectively; ResNet-18 attained 0.96 (95% CI: 0.92-0.98), and 0.91 (95% CI: 0.87-0.94), respectively. However, the ResNet-50 model exhibited suboptimal diagnostic performance for osteopenia, with an AUC value of only 0.76 (95% CI: 0.69-0.80). Alterations in tube voltage had a more pronounced impact on the performance of the DenseNet. In the independent test set with tube voltage at 100 kVp images, the accuracy and precision of DenseNet decreased on average by approximately 14.29% and 18.82%, respectively, whereas the accuracy and precision of ResNet-18 decreased by about 8.33% and 7.14%, respectively. The state-of-the-art DL framework model offers an effective and efficient approach for opportunistic osteoporosis screening using chest CT, without incurring additional costs or radiation exposure.

Bio-QSARs 2.0: Unlocking a new level of predictive power for machine learning-based ecotoxicity predictions by exploiting chemical and biological information

Practical, legal, and ethical reasons necessitate the development of methods to replace animal experiments. Computational techniques to acquire information that traditionally relied on animal testing are considered a crucial pillar among these so-called new approach methodologies. In this light, we recently introduced the Bio-QSAR concept for multispecies aquatic toxicity regression tasks. These machine learning models, trained on both chemical and biological information, are capable of both cross-chemical and cross-species predictions. Here, we significantly extend these models’ applicability. This was realized by increasing the quantity of training data by a factor of approximately 20, accomplished by considering both additional chemicals and aquatic organisms. Additionally, variable test durations and associated random effects were accommodated by employing a machine learning algorithm that combines tree-boosting with mixed-effects modeling (i.e., Gaussian Process Boosting). We also explored various biological descriptors including Dynamic Energy Budget model parameters, taxonomic distances, as well as genus-specific traits and investigated the inclusion of mode-of-action information. Through these efforts, we developed Bio-QSARs for fish and aquatic invertebrates with exceptional predictive power (R squared of up to 0.92 on independent test sets). Moreover, we made considerable strides to make models applicable for a range of use cases in environmental risk assessment as well as research and development of chemicals. Models were made fully explainable by implementing an algorithmic multicollinearity correction combined with SHapley Additive exPlanations. Furthermore, we devised novel approaches for applicability domain construction that take feature importance into account. We are hence confident these models, which are available via open access, will make a significant contribution towards the implementation of new approach methodologies and ultimately have the potential to support “Green Chemistry” and “Green Toxicology”.

Tumor conspicuity enhancement-based segmentation model for liver tumor segmentation and RECIST diameter measurement in non-contrast CT images

Background and objectiveLiver tumor segmentation (LiTS) accuracy on contrast-enhanced computed tomography (CECT) images is higher than that on non-contrast computed tomography (NCCT) images. However, CECT requires contrast medium and repeated scans to obtain multiphase enhanced CT images, which is time-consuming and cost-increasing. Therefore, despite the lower accuracy of LiTS on NCCT images, which still plays an irreplaceable role in some clinical settings, such as guided brachytherapy, ablation, or evaluation of patients with renal function damage. In this study, we intend to generate enhanced high-contrast pseudo-color CT (PCCT) images to improve the accuracy of LiTS and RECIST diameter measurement on NCCT images. MethodsTo generate high-contrast CT liver tumor region images, an intensity-based tumor conspicuity enhancement (ITCE) model was first developed. In the ITCE model, a pseudo color conversion function from an intensity distribution of the tumor was established, and it was applied in NCCT to generate enhanced PCCT images. Additionally, we design a tumor conspicuity enhancement-based liver tumor segmentation (TCELiTS) model, which was applied to improve the segmentation of liver tumors on NCCT images. The TCELiTS model consists of three components: an image enhancement module based on the ITCE model, a segmentation module based on a deep convolutional neural network, and an attention loss module based on restricted activation. Segmentation performance was analyzed using the Dice similarity coefficient (DSC), sensitivity, specificity, and RECIST diameter error. ResultsTo develop the deep learning model, 100 patients with histopathologically confirmed liver tumors (hepatocellular carcinoma, 64 patients; hepatic hemangioma, 36 patients) were randomly divided into a training set (75 patients) and an independent test set (25 patients). Compared with existing tumor automatic segmentation networks trained on CECT images (U-Net, nnU-Net, DeepLab-V3, Modified U-Net), the DSCs achieved on the enhanced PCCT images are both improved compared with those on NCCT images. We observe improvements of 0.696–0.713, 0.715 to 0.776, 0.748 to 0.788, and 0.733 to 0.799 in U-Net, nnU-Net, DeepLab-V3, and Modified U-Net, respectively, in terms of DSC values. In addition, an observer study including 5 doctors was conducted to compare the segmentation performance of enhanced PCCT images with that of NCCT images and showed that enhanced PCCT images are more advantageous for doctors to segment tumor regions. The results showed an accuracy improvement of approximately 3%–6%, but the time required to segment a single CT image was reduced by approximately 50 %. ConclusionsExperimental results show that the ITCE model can generate high-contrast enhanced PCCT images, especially in liver regions, and the TCELiTS model can improve LiTS accuracy in NCCT images.

Development and evaluation of a deep learning-based model for simultaneous detection and localization of rib and clavicle fractures in trauma patients’ chest radiographs

PurposeTo develop a rib and clavicle fracture detection model for chest radiographs in trauma patients using a deep learning (DL) algorithm.Materials and methodsWe retrospectively collected 56 145 chest X-rays (CXRs)...

Development and validation of a nomogram for predicting lymph node metastasis in ductal carcinoma in situ with microinvasion: A SEER population-based study.

Ductal carcinoma in situ with microinvasion (DCIS-MI) is a special type of breast cancer. It is an invasive lesion less than 1.0 mm in size related to simple ductal carcinoma in situ (DCIS). Lymph node metastasis (LNM) in DCIS-MI often indicates a poor prognosis. Therefore, the management of lymph nodes plays a vital role in the treatment strategy of DCIS-MI. Since DCIS-MI is often diagnosed by postoperative paraffin section and immunohistochemical detection, to obtain the best clinical benefits for such patients, we aim to establish and verify a nomogram to predict the possibility of lymph node metastasis in DCIS-MI patients and help preoperative or intraoperative clinical decision-making. A retrospective analysis of patients with DCIS-MI in the Surveillance, Epidemiology, and End Results (SEER) database from 2010 to 2019 was performed. The study cohort was randomly divided into a training cohort and a validation cohort at a ratio of 7:3. The risk factors were determined by univariate and multivariate logistic regression analyses in the training cohort, and a nomogram was constructed. The receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) were used to evaluate the nomogram in the training set and validation set. An independent data cohort was obtained from the Shanghai Jiao Tong University Breast Cancer Database (SJTU-BCDB) for external validation. This study included 3951 female patients from SEER with DCIS-MI, including 244 patients with regional lymph node metastasis, accounting for 6.18% of the total. An independent test set of 323 patients from SJTU-BCDB was used for external validation. According to the multifactorial logistic regression analysis results, age at diagnosis, ethnicity, grade, and surgical modality were included in the prediction model. The areas under the ROC curves (AUCs) were 0.739 (95% CI: 0.702~0.775), 0.732 (95% CI: 0.675~0.788), and 0.707 (95%CI: 0.607-0.807) in the training, validation and external test groups, suggesting that the column line graphs had excellent differentiation. The calibration curves slope was close to 1, and the model's predicted values were in good agreement with the actual values. The DCA curves showed good clinical utility. In this study, we constructed accurate and practical columnar maps with some clinical benefit to predict the likelihood of lymph node metastasis in patients with postoperatively diagnosed DCIS-MI and provide a reference value for specifying treatment strategies.

1235 Using Supervised Machine Learning to Validate a Novel Scoring System for the Prediction of Disease Remission of Functional Adenomas Following Transsphenoidal Resection

INTRODUCTION: Functional pituitary adenomas are common primary central nervous system tumors with potent systemic endocrinological effects. Postoperatively, remission rates range from 65-98%. METHODS: 392 adult patients who underwent primary resection of a functional pituitary adenoma at the three Mayo Clinic between 2006-2022 were retrospectively reviewed. Multivariate logistic regression was performed to identify variables predictive of disease remission. Variables found statistically significant on multivariate analysis were incorporated into our proposed pituitary-SCHEME score. Machine learning models were implemented to compare the accuracy of the pituitary-SCHEME score against the multivariate models. After training (314 patients) and cross-validation of the machine learning models, an independent testing set of 78 patients was performed. SPSS V27 and R4.2.1 were used for statistical analyses. RESULTS: 261 (66.6%) patients achieved post-operative biochemical remission. On multivariate analysis gross-total resection and Knosp-Grade 0,I,II adenomas were predictive of disease remission following resection, while macroadenomas (adenoma > 10mm), male sex, mammosomatotroph and Mixed Growth-Hormone + Prolactin adenomas were predictive of continued disease. In machine learning models, without Pituitary-SCHEME score, the K-Nearest Neighbors (KNN) model achieved the highest accuracy at 75.6% followed by Naive Bayes model. An increase in model sensitivity was achieved with inclusion of Pituitary-SCHEME score with the Linear-Discriminant-Analysis (LDA) model achieving the highest accuracy at 86.9%, followed by the Classification and Regression Trees (CART) model. The random forest model had the largest AUC-ROC among models with and without pituitary-SCHEME score. Model prediction accuracy (with vs. without pituitary-SCHEME score) were found to be statistically different based on Wilcoxon rank sum testing (p < 0.0001). CONCLUSIONS: The novel pituitary-SCHEME score, which incorporates perioperative and endocrinological measures, has promise as a clinical tool to predict patient outcomes following surgical resection of functional adenomas.

Machine Learning Estimation of Maximum Vertical Velocity from Radar

Abstract The quantification of storm updrafts remains unavailable for operational forecasting despite their inherent importance to convection and its associated severe weather hazards. Updraft proxies, like overshooting top area from satellite images, have been linked to severe weather hazards but only relate to a limited portion of the total storm updraft. This study investigates if a machine learning model, namely, U-Nets, can skillfully retrieve maximum vertical velocity and its areal extent from three-dimensional gridded radar reflectivity alone. The machine learning model is trained using simulated radar reflectivity and vertical velocity from the National Severe Storm Laboratory’s convection permitting Warn-on-Forecast System (WoFS). A parametric regression technique using the sinh–arcsinh–normal distribution is adapted to run with U-Nets, allowing for both deterministic and probabilistic predictions of maximum vertical velocity. The best models after hyperparameter search provided less than 50% root mean squared error, a coefficient of determination greater than 0.65, and an intersection over union (IoU) of more than 0.45 on the independent test set composed of WoFS data. Beyond the WoFS analysis, a case study was conducted using real radar data and corresponding dual-Doppler analyses of vertical velocity within a supercell. The U-Net consistently underestimates the dual-Doppler updraft speed estimates by 50%. Meanwhile, the area of the 5 and 10 m s−1 updraft cores shows an IoU of 0.25. While the above statistics are not exceptional, the machine learning model enables quick distillation of 3D radar data that is related to the maximum vertical velocity, which could be useful in assessing a storm’s severe potential. Significance Statement All convective storm hazards (tornadoes, hail, heavy rain, straight line winds) can be related to a storm’s updraft. Yet, there is no direct measurement of updraft speed or area available for forecasters to make their warning decisions from. This paper addresses the lack of observational data by providing a machine learning solution that skillfully estimates the maximum updraft speed within storms from only the radar reflectivity 3D structure. After further vetting the machine learning solutions on additional real-world examples, the estimated storm updrafts will hopefully provide forecasters with an added tool to help diagnose a storm’s hazard potential more accurately.

Clinical Utility of a CT-based AI Prognostic Model for Segmentectomy in Non-Small Cell Lung Cancer.

Background Currently, no tool exists for risk stratification in patients undergoing segmentectomy for non-small cell lung cancer (NSCLC). Purpose To develop and validate a deep learning (DL) prognostic model using preoperative CT scans and clinical and radiologic information for risk stratification in patients with clinical stage IA NSCLC undergoing segmentectomy. Materials and Methods In this single-center retrospective study, transfer learning of a pretrained model was performed for survival prediction in patients with clinical stage IA NSCLC who underwent lobectomy from January 2008 to March 2017. The internal set was divided into training, validation, and testing sets based on the assignments from the pretraining set. The model was tested on an independent test set of patients with clinical stage IA NSCLC who underwent segmentectomy from January 2010 to December 2017. Its prognostic performance was analyzed using the time-dependent area under the receiver operating characteristic curve (AUC), sensitivity, and specificity for freedom from recurrence (FFR) at 2 and 4 years and lung cancer-specific survival and overall survival at 4 and 6 years. The model sensitivity and specificity were compared with those of the Japan Clinical Oncology Group (JCOG) eligibility criteria for sublobar resection. Results The pretraining set included 1756 patients. Transfer learning was performed in an internal set of 730 patients (median age, 63 years [IQR, 56-70 years]; 366 male), and the segmentectomy test set included 222 patients (median age, 65 years [IQR, 58-71 years]; 114 male). The model performance for 2-year FFR was as follows: AUC, 0.86 (95% CI: 0.76, 0.96); sensitivity, 87.4% (7.17 of 8.21 patients; 95% CI: 59.4, 100); and specificity, 66.7% (136 of 204 patients; 95% CI: 60.2, 72.8). The model showed higher sensitivity for FFR than the JCOG criteria (87.4% vs 37.6% [3.08 of 8.21 patients], P = .02), with similar specificity. Conclusion The CT-based DL model identified patients at high risk among those with clinical stage IA NSCLC who underwent segmentectomy, outperforming the JCOG criteria. © RSNA, 2024 Supplemental material is available for this article.

Enhancing Convolutional Neural Networks via separately trained kernels for digit recognition

This paper introduces a novel methodology for pre-training neural networks. Instead of the traditional approach of running a single classifier for all objects, the method used in the research constructs individual classifiers for each object to extract unique features. These classifiers, in essence, act as a series of binary classifiers, each pre-training a distinct set of convolutional kernels. These individually trained kernels are then combined and processed by a comprehensive classifier. The methodology leverages the power of individual feature extraction and collaborative processing to enhance the overall performance of the classifier. The results demonstrate that this innovative pre-training approach leads to a significant improvement in classification performance, yielding higher accuracy compared to conventional methods. It underscores the benefits of incorporating individual object feature extraction and combined processing in the pre-training phase of neural network-based classification tasks. The study opens new paths for improving pre-training methods in neural networks, with potential applications in various fields that require high-accuracy object recognition. Future work will delve deeper into the potential challenges associated with model complexity and overfitting, and will also include a more comprehensive evaluation using extensive training, cross-validation, and independent test sets. This will further validate the effectiveness of our proposed pre-training methodology.

Novel peripheral blood mononuclear cell mRNA signature for IFN-beta therapy responsiveness prediction in multiple sclerosis

Interferon-beta (IFN- β ) is one of the classical drugs for immunomodulatory therapy in relapsing-remitting multiple sclerosis (RRMS) patients, but the drug responsiveness of different patients varies. Currently, there is no valid model to predict IFN- β responsiveness. This research attempted to develop an IFN- β responsiveness prediction model based on mRNA expression in RRMS patient peripheral blood mononuclear cells. Peripheral blood mononuclear cell mRNA expression datasets including 50 RRMS patients receiving IFN- β treatment were obtained from GEO. Among the datasets, 24 cases from GSE24427 were included in a training set, and 18 and 9 cases from GSE19285 and GSE33464, respectively, were adopted as two independent test sets. In the training set, blood samples were collected immediately before first, second, month 1, 12, and 24 IFN- β injection, and the mRNA expression data at four time points, namely, two days, one month, one year and two years after the onset of IFN- β treatment, were compared with pre-treatment data to identify IFN-stimulated genes (ISGs). The ISGs at the one-month time point were used to construct the drug responsiveness prediction model. Next, the drug responsiveness model was verified in the two independent test sets to examine the performance of the model in predicting drug responsiveness. Finally, we used CIBERSORTx to estimate the content of cell subtypes in samples and evaluated whether differences in the proportions of cell subtypes were related to differences in IFN- β responsiveness. Among the four time points, one month was the time point when the training set GSE24427 and test set GSE33464 had the highest number of ISGs. Functional analysis showed that these one-month ISGs were enriched in biological functions such as the innate immune response, type-I interferon signalling pathway, and other IFN- β -associated functions. Based on these ISGs, we obtained a four-factor prediction model for IFN- β responsiveness including MX1, MX2, XAF1, and LAMP3. In addition, the model demonstrated favourable predictive performance within the training set and two external test sets. A higher proportion of activated NK cells and lower naive CD4/total CD4 ratio might indicate better drug responsiveness. This research developed a polygene-based biomarker model that could predict RRMS patient IFN- β responsiveness in the early treatment period. This model could probably help doctors screen out patients who would not benefit from IFN- β treatment early and determine whether a current treatment plan should be continued.