Development and validation of the AI-based diagnosis system for pathological findings in invasive front of colorectal cancer
Ontology highlight
ABSTRACT:
DISEASE(S): Colorectal Cancer
PROVIDER: 2653863 | ecrin-mdr-crc |
REPOSITORIES: ECRIN MDR
Similar Datasets
Project description:Pathology is the gold standard of clinical diagnosis. Artificial intelligence (AI) in pathology becomes a new trend, but it is still not widely used due to the lack of necessary explanations for pathologists to understand the rationale. Clinic-compliant explanations besides the diagnostic decision of pathological images are essential for AI model training to provide diagnostic suggestions assisting pathologists practice. In this study, we propose a new annotation form, PathNarratives, that includes a hierarchical decision-to-reason data structure, a narrative annotation process, and a multimodal interactive annotation tool. Following PathNarratives, we recruited 8 pathologist annotators to build a colorectal pathological dataset, CR-PathNarratives, containing 174 whole-slide images (WSIs). We further experiment on the dataset with classification and captioning tasks to explore the clinical scenarios of human-AI-collaborative pathological diagnosis. The classification tasks show that fine-grain prediction enhances the overall classification accuracy from 79.56 to 85.26%. In Human-AI collaboration experience, the trust and confidence scores from 8 pathologists raised from 3.88 to 4.63 with providing more details. Results show that the classification and captioning tasks achieve better results with reason labels, provide explainable clues for doctors to understand and make the final decision and thus can support a better experience of human-AI collaboration in pathological diagnosis. In the future, we plan to optimize the tools for the annotation process, and expand the datasets with more WSIs and covering more pathological domains.
| S-EPMC9908590 | biostudies-literature
Project description:Splice-switching oligonucleotides (SSOs) are antisense compounds that act directly on pre-mRNA to modulate alternative splicing (AS). This study demonstrates the value that artificial intelligence/machine learning (AI/ML) provides for the identification of functional, verifiable, and therapeutic SSOs. We trained XGboost tree models using splicing factor (SF) pre-mRNA binding profiles and spliceosome assembly information to identify modulatory SSO binding sites on pre-mRNA. Using Shapley and out-of-bag analyses we also predicted the identity of specific SFs whose binding to pre-mRNA is blocked by SSOs. This step adds considerable transparency to AI/ML-driven drug discovery and informs biological insights useful in further validation steps. We applied this approach to previously established functional SSOs to retrospectively identify the SFs likely to regulate those events. We then took a prospective validation approach using a novel target in triple negative breast cancer (TNBC), NEDD4L exon 13 (NEDD4Le13). Targeting NEDD4Le13 with an AI/ML-designed SSO decreased the proliferative and migratory behavior of TNBC cells via downregulation of the TGFβ pathway. Overall, this study illustrates the ability of AI/ML to extract actionable insights from RNA-seq data.
2024-04-25 | GSE215917 | GEO
Project description:Splice-switching oligonucleotides (SSOs) are antisense compounds that act directly on pre-mRNA to modulate alternative splicing (AS). This study demonstrates the value that artificial intelligence/machine learning (AI/ML) provides for the identification of functional, verifiable, and therapeutic SSOs. We trained XGboost tree models using splicing factor (SF) pre-mRNA binding profiles and spliceosome assembly information to identify modulatory SSO binding sites on pre-mRNA. Using Shapley and out-of-bag analyses we also predicted the identity of specific SFs whose binding to pre-mRNA is blocked by SSOs. This step adds considerable transparency to AI/ML-driven drug discovery and informs biological insights useful in further validation steps. We applied this approach to previously established functional SSOs to retrospectively identify the SFs likely to regulate those events. We then took a prospective validation approach using a novel target in triple negative breast cancer (TNBC), NEDD4L exon 13 (NEDD4Le13). Targeting NEDD4Le13 with an AI/ML-designed SSO decreased the proliferative and migratory behavior of TNBC cells via downregulation of the TGFβ pathway. Overall, this study illustrates the ability of AI/ML to extract actionable insights from RNA-seq data.
| S-SCDT-10_1038-S44320-024-00034-9 | biostudies-other
Project description:BackgroundDiagnosing underlying causes of nonneurogenic male lower urinary tract symptoms associated with bladder outlet obstruction (BOO) is challenging. Video-urodynamic studies (VUDS) and pressure-flow studies (PFS) are both invasive diagnostic methods for BOO. VUDS can more precisely differentiate etiologies of male BOO, such as benign prostatic obstruction, primary bladder neck obstruction, and dysfunctional voiding, potentially outperforming PFS.ObjectiveThese examinations' invasive nature highlights the need for developing noninvasive predictive models to facilitate BOO diagnosis and reduce the necessity for invasive procedures.MethodsWe conducted a retrospective study with a cohort of men with medication-refractory, nonneurogenic lower urinary tract symptoms suspected of BOO who underwent VUDS from 2001 to 2022. In total, 2 BOO predictive models were developed-1 based on the International Continence Society's definition (International Continence Society-defined bladder outlet obstruction; ICS-BOO) and the other on video-urodynamic studies-diagnosed bladder outlet obstruction (VBOO). The patient cohort was randomly split into training and test sets for analysis. A total of 6 machine learning algorithms, including logistic regression, were used for model development. During model development, we first performed development validation using repeated 5-fold cross-validation on the training set and then test validation to assess the model's performance on an independent test set. Both models were implemented as paper-based nomograms and integrated into a web-based artificial intelligence prediction tool to aid clinical decision-making.ResultsAmong 307 patients, 26.7% (n=82) met the ICS-BOO criteria, while 82.1% (n=252) were diagnosed with VBOO. The ICS-BOO prediction model had a mean area under the receiver operating characteristic curve (AUC) of 0.74 (SD 0.09) and mean accuracy of 0.76 (SD 0.04) in development validation and AUC and accuracy of 0.86 and 0.77, respectively, in test validation. The VBOO prediction model yielded a mean AUC of 0.71 (SD 0.06) and mean accuracy of 0.77 (SD 0.06) internally, with AUC and accuracy of 0.72 and 0.76, respectively, externally. When both models' predictions are applied to the same patient, their combined insights can significantly enhance clinical decision-making and simplify the diagnostic pathway. By the dual-model prediction approach, if both models positively predict BOO, suggesting all cases actually resulted from medication-refractory primary bladder neck obstruction or benign prostatic obstruction, surgical intervention may be considered. Thus, VUDS might be unnecessary for 100 (32.6%) patients. Conversely, when ICS-BOO predictions are negative but VBOO predictions are positive, indicating varied etiology, VUDS rather than PFS is advised for precise diagnosis and guiding subsequent therapy, accurately identifying 51.1% (47/92) of patients for VUDS.ConclusionsThe 2 machine learning models predicting ICS-BOO and VBOO, based on 6 noninvasive clinical parameters, demonstrate commendable discrimination performance. Using the dual-model prediction approach, when both models predict positively, VUDS may be avoided, assisting in male BOO diagnosis and reducing the need for such invasive procedures.
| S-EPMC11303901 | biostudies-literature
Project description:Non-invasive and cost-effective diagnosis of gastric cancer is essential to improve outcomes. Aim of the study was to establish a neural network model based on patient demographic data and serum biomarker panels to aid gastric cancer diagnosis. A total of 295 patients hospitalized in Nanjing Drum Tower hospital diagnosed with gastric cancer based on tissue biopsy, and 423 healthy volunteers were included in the study. Demographical information and tumor biomarkers were obtained from Hospital Information System (HIS) as original data. Pearson's correlation analysis was applied on 574 individuals' data (training set, 229 patients and 345 healthy volunteers) to analyze the relationship between each variable and the final diagnostic result. And independent sample t test was used to detect the differences of the variables. Finally, a neural network model based on 14 relevant variables was constructed. The model was tested on the validation set (144 individuals including 66 patients and 78 healthy volunteers). The predictive ability of the proposed model was compared with other common machine learning models including logistic regression and random forest. Tumor markers contributing significantly to gastric cancer screening included CA199, CA125, AFP, and CA242 were identified, which might be considered as important inspection items for gastric cancer screening. The accuracy of the model on validation set was 86.8% and the F1-score was 85.0%, which were better than the performance of other models under the same condition. A non-invasive and low-cost artificial neural network model was developed and proved to be a valuable tool to assist gastric cancer diagnosis.
| S-EPMC9758153 | biostudies-literature
Project description:ObjectiveTo develop a technique for non-invasive prenatal diagnosis of spinal muscular atrophy and validate its performance.Study designPregnant women with 1 copy of SMN1 and male fetuses were enrolled. Seventeen women were included in test set A, and 10 of them were selected into test set B randomly and blinded. The two sets were tested independently by two different researchers blinded to fetal genotypes. Fetal DNA fractions were calculated based on the relative proportion of mapped chromosome Y sequencing reads. An algorithm was developed to decide fetal SMN1 copy numbers.ResultsThe concordance rate with the results of MLPA testing of amniocyte DNA was 94.12% in test set A and 90% in set B. For all tests with a classifiable result, the percent of agreement with the results of MLPA testing of amniocyte DNA was up to 100% (25/25).ConclusionWe have developed a direct, rapid, and low-cost technique, which has a potential to be utilized for first-trimester non-invasive prenatal diagnosis and screening for spinal muscular atrophy with considerable reliability and feasibility.
| S-EPMC7031576 | biostudies-literature
Project description:The goal of the study was to identify miRNAs differentially expressed in the invasive front of early and more advanced colorectal cancers, and to determine wheter miRNA profiles could differentiate between colorectal cancers with and without early metasttic progression. miRNA array data was used for analyses of associations beteen metastatic status and for associations to histopathological risk factors for metastatic progression.
2018-08-01 | GSE110402 | GEO
Project description:BackgroundSkeletal dysplasia (SD) is one of the most common inherited neonatal disorders worldwide, where the recurrent pathogenic mutations in the FGFR2, FGFR3, COL1A1, COL1A2 and COL2A1 genes are frequently reported in both non-lethal and lethal SD. The traditional prenatal diagnosis of SD using ultrasonography suffers from lower accuracy and performed at latter gestational stage. Therefore, it remains in desperate need of precise and accurate prenatal diagnosis of SD in early pregnancy. With the advancements of next-generation sequencing (NGS) technology and bioinformatics analysis, it is feasible to develop a NGS-based assay to detect genetic defects in association with SD in the early pregnancy.MethodsAn ampliseq-based targeted sequencing panel was designed to cover 87 recurrent hotspots reported in 11 common dominant SD and run on both Ion Proton and NextSeq550 instruments. Thirty-six cell-free and 23 genomic DNAs were used for assay developed. Spike-in DNA prepared from standard sample harboring known mutation and normal sample were also employed to validate the established SD workflow. Overall performances of coverage, uniformity, and on-target rate, and the detecting limitations on percentage of fetal fraction and read depth were evaluated.ResultsThe established targeted-seq workflow enables a single-tube multiplex PCR for library construction and shows high amplification efficiency and robust reproducibility on both Ion Proton and NextSeq550 platforms. The workflow reaches 100% coverage and both uniformity and on-target rate are > 96%, indicating a high quality assay. Using spike-in DNA with different percentage of known FGFR3 mutation (c.1138 G > A), the targeted-seq workflow demonstrated the ability to detect low-frequency variant of 2.5% accurately. Finally, we obtained 100% sensitivity and 100% specificity in detecting target mutations using established SD panel.ConclusionsAn expanded panel for rapid and cost-effective genetic detection of SD has been developed. The established targeted-seq workflow shows high accuracy to detect both germline and low-frequency variants. In addition, the workflow is flexible to be conducted in the majority of the NGS instruments and ready for routine clinical application. Taken together, we believe the established panel provides a promising diagnostic or therapeutic strategy for prenatal genetic testing of SD in routine clinical practice.
| S-EPMC8600686 | biostudies-literature
Project description:Background: Whether girls with central precocious puberty (CPP) should undergo routine brain magnetic resonance imaging (MRI) to identify potential intracranial pathologies is controversial. Aims: To evaluate the brain MRI results of girls with CPP and identify the clinical and endocrine predictors of brain abnormalities. Methods: This retrospective study obtained data from pediatric endocrine clinics at Chang Gung Children's Hospital. From 1997 and 2017, 403 girls were consecutively diagnosed with CPP. After the exclusion of patients with a history of central nervous system (CNS) insults or associated neuropsychiatric symptom or signs, we studied the prevalence of brain abnormalities in 251 girls with CPP who received detailed MRI examinations of the hypothalamus and pituitary area. We also recorded the demographic data of the participants, including the onset of puberty; initial pubertal status; height; weight; uterus and ovary sizes; and basal luteinizing hormone (LH), follicle-stimulating hormone (FSH), estradiol levels, and the response to GnRH stimulation test. Results: Among the 251 girls with CPP, we observed no brain alterations in 190 (75.70%), abnormalities in the hypothalamic-pituitary (H-P) area in 54 (21.51%), and abnormalities in the non-H-P area in 7 (2.79%). Among the 54 patients that had H-P lesions, we noted pathological findings related to CPP (hypothalamic hamartoma) in only one (0.4%) girl aged below 6 years old. We also identified incidental findings in the other girls with H-P lesions, including non-functioning pituitary microadenomas (12.35%), pituitary pars intermedia cysts (4.38%), Rathke's pouch cysts (1.99%), pituitary hypoplasia (1.59%), and pineal gland cysts (0.8%). The patients that had non-H-P lesions were found to have arachnoid cysts (1.59%), Chiari I malformation (0.4%), prepontine nodule (0.4%), and choroidal fissure cyst (0.4%). Of all the patients with brain lesions, 45 (73.77%) underwent regular MRI follow-up. While none of the H-P and non-H-P lesions showed progression, 19.67% of these regressed during the follow-up. None of the participants exhibited other hormonal abnormalities or underwent surgery. Conclusion: The prevalence of true pathological brain lesions related to CPP in girls without prior symptoms or signs of CNS lesions was low (0.4%). None of the girls with intracranial lesions required further intervention besides the GnRH agonist treatment. These data question the routine use of brain MRI in all girls with CPP, especially in those who are healthy without neurologic symptoms.
| S-EPMC7214687 | biostudies-literature
Project description:BackgroundThis study aimed to develop and validate an AI (artificial intelligence)-aid method in myocardial perfusion imaging (MPI) to differentiate ischemia in coronary artery disease.MethodsWe retrospectively selected 599 patients who had received gated-MPI protocol. Images were acquired using hybrid SPECT-CT systems. A training set was used to train and develop the neural network and a validation set was used to test the predictive ability of the neural network. We used a learning technique named "YOLO" to carry out the training process. We compared the predictive accuracy of AI with that of physician interpreters (beginner, inexperienced, and experienced interpreters).ResultsTraining performance showed that the accuracy ranged from 66.20% to 94.64%, the recall rate ranged from 76.96% to 98.76%, and the average precision ranged from 80.17% to 98.15%. In the ROC analysis of the validation set, the sensitivity range was 88.9 ~ 93.8%, the specificity range was 93.0 ~ 97.6%, and the AUC range was 94.1 ~ 96.1%. In the comparison between AI and different interpreters, AI outperformed the other interpreters (most P-value < 0.05).ConclusionThe AI system of our study showed excellent predictive accuracy in the diagnosis of MPI protocols, and therefore might be potentially helpful to aid radiologists in clinical practice and develop more sophisticated models.
| S-EPMC10273563 | biostudies-literature