Citations

Clear

5212 citation(s) found for "Wang, Li-Ju".

Abstract 4999: Multi-omics insights into the molecular signature and prognosis of hypopharyngeal squamous cell carcinoma
Ren, Yanxin , Feng, Chun , Yu, Dan , Wang, Xiaoyan , Yu, Siting, Zhang, Hongjiang, Huo, Bo , Jiang, Honglu, Li, Zuli, Wang, Junlin , Zhong, Qi , Xiong, Wei, Yang, Qin , Zhang, Hongjiang, Huo, Bangyun, Su, Yu‐Xiong , Yang, Ping , Liao, Yong , Wang, Junwen

Mayo Clinic

Cancer Research | April 21, 2025 | Volume: 85 | Issue: 8_Supplement_1 | Pages: 4999-4999

10.1158/1538-7445.am2025-4999 Get It

Abstract Approximately two-thirds of hypopharyngeal squamous cell carcinoma (HPSCC) are diagnosed in advanced stage, with the worst prognosis among head and neck squamous cell carcinomas (HNSCCs). Identifying biomarkers for high-risk patients who need aggressive treatment is crucial. We present mutational, transcriptomic and proteomic studies of 103 Chinese HPSCC patients, observing a higher prevalence and poorer prognosis in males comparing to females. Genes in estrogen response pathways are up-regulated and proteins phosphorylated by protein kinase C (PKC) and Cyclin-dependent kinases (CDKs) are aberrantly regulated. We identify aberrant copy number regions including SOX2 (3q26.33), FGFR (8p11.23), CCND1 (11q13.3), CDKN2A/2B (9p21.3), and MYC (8q24.21). HPV status combined with highly mutated genes as predictive markers was assessed, including SYNE1 in HPV(-), MUC4 in HPV(+) patients. We confirm the presence of four known gene expression (GE) subtypes of HNSCC in our HPSCC cohort and present the genomic alterations, patient characteristics and survival status of each subtype. A predictive model involving clinical factors and transcriptomic expression of six genes was established and cross-site validated. The findings open new opportunities for stratifying high-risk patients and molecular targets for novel therapeutic strategies. Citation Format: Yanxin Ren, Wei Xiong, Chun Feng, Dan Yu, Xiaoyan Wang, Qing Yang, Siting Yu, Hongjiang Zhang, Bangyun Huo, Honglu Jiang, Zuli Li, Junlin Wang, Yu-Xiong Su, Ping Yang, Yong Liao, Qi Zhong, Junwen Wang. Multi-omics insights into the molecular signature and prognosis of hypopharyngeal squamous cell carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4999.
Abstract 2429: A multi-modal learning framework to integrate digital pathology image and multi-omics data in breast cancer
Prodduturi, Naresh, Weinshilboum, Richard, Klee, Eric W. , Wang, Liewei , Goetz, Matthew P. , Tang, Xiaojia , Tizhoosh, Hamid R. , Thompson, Kevin J. , Giridhar, Karthik V. , Boughey, Judy C. , Goetz, Matthew P. , Suman, Vera J. , Kalari, Krishna R.

Mayo Clinic

Cancer Research | April 21, 2025 | Volume: 85 | Issue: 8_Supplement_1 | Pages: 2429-2429

10.1158/1538-7445.am2025-2429 Get It

Abstract Deep learning has advanced digital pathology and multi-omics data analysis. However, there is still a critical need to integrate these modalities to better understand complex diseases. Deep learning models, such as Prov-Gigapath, a state-of-the-art pathology foundation model, have been developed for analyzing whole-slide images (WSI). Our previously published deep-learning framework, OmicsFootPrint (OFP), transforms high-dimensional multi-omics data into intuitive genomically-arranged circular images and utilizes transfer learning to overcome the challenges of high-dimensional, low-sample-size data. In this study, we aimed to combine WSI and multi-omics data to develop a framework capable of distinguishing between invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC), which is clinically significant as it may help explain differential responses to therapies, such as tamoxifen versus letrozole. We enhanced the performance of our OFP model by dividing each multi-omics circular image (1024x1024x3 pixels) into 16 equal-sized patches (256x256x3 pixels). Each patch was trained using a separate EfficientNetV2 architecture with pre-trained ImageNet weights. The probability outcomes from these 16 models were subsequently combined through a meta-learner model using the AutoGluon framework. Concurrently, WSI images corresponding to the same patients were independently trained, maintaining consistent train/validation/test splits. The Prov-Gigapath model, which was trained using a dataset from the Providence Health System, was employed to extract patch-level embeddings from the WSI data. These embeddings were fine-tuned using a LongNet-based masked autoencoder to predict outcomes. Finally, predictions from the OFP model and the WSI-based Prov-Gigapath model were integrated through a meta-learner, preserving the original train/validation/test split. Model performances (AUC) were evaluated using the held-out test set. The gold standard used for the analysis was histological identification of IDC versus ILC, against which the performance of omics data alone, WSI data alone, and the multi-modal integration approach were compared. We evaluated our integrated model as a proof-of-concept using The Cancer Genome Atlas (TCGA) breast cancer cohort, comprising 92 ILC and 391 IDC samples. Multi-omics data, including gene expression, microRNA, copy number variation, and RPPA, were analyzed using the OFP framework, while WSI data were processed using the Prov-Gigapath method. The results demonstrated that the multi-modal integration approach significantly outperformed individual modalities, achieving an AUC of 0.98 compared to 0.87 for OFP (original), 0.95 for OFP (16-patches), and 0.92 for WSI. Integrating WSI and multi-omics data improves predictive accuracy, significantly improving the classification of breast cancer subtypes over histology alone. Citation Format: Naresh Prodduturi, Xiaojia Tang, Hamid R. Tizhoosh, Kevin J. Thompson, Richard Weinshilboum, Karthik V. Giridhar, Judy C. Boughey, Eric W. Klee, Liewei Wang, Matthew P. Goetz, Vera Suman, Krishna R. Kalari. A multi-modal learning framework to integrate digital pathology image and multi-omics data in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2429.
Sustained mTORC1 activation in activated T cells impairs vaccine responses in older individuals
Lin, Xiaorong , Du, Yanhua , Kan, Shuo , Chen, Junjie , Yin, Yunxue, Li, Linlin , Jiang, Wuhua , Cao, Wenqiang , Kim, Chulwoo , Chen, Liang , Wang, Shiwen , Goronzy, Jörg J. , Jin, Jun , Jiang, Wenrong , Cao, Wenqiang , Kim, Chulwoo , Chen, Liang , Wang, Shiwen , Jin, Jun , Goronzy, Jorg , Lin, Xiaorong , Du, Yanhua , Li, Linlin, Chen, Jingwen, Chen, Jingwen

Mayo Clinic

Science Advances | April 18, 2025 | Volume: 11 | Issue: 16 | Pages: No page available

10.1126/sciadv.adt4881 | PMID 40249803 Get It

T cell aging contributes to the lower vaccine efficacy in older adults, yet the molecular mechanism remains elusive. Here, we show the density of initially responding naïve CD4 + T cells is instructive in T follicular helper (TFH) cell fate decisions and declines with age. A lower number of initially responding cells did not affect TFH differentiation at peak responses after immunization but accounted for an increased contraction phase manifesting as a larger loss of CXCR5 expression. Mechanistically, cells activated at a lower initial density had more sustained mammalian target of rapamycin complex 1 (mTORC1) activities that impair CXCR5 maintenance. YAP-dependent regulation of SLC7A5 involved in the cell density–dependent regulation of mTORC1 activities and TFH loss. Old mice fed with a leucine-restricted diet after peak responses showed smaller TFH loss and improved humoral immune responses. Attenuating mTORC1 signaling after peak response is a strategy to boost vaccine responses in older individuals.
Intravascular Imaging-Guided Versus Angiography-Guided Percutaneous Coronary Intervention in Patients with Non-ST-Segment Elevation Myocardial Infarction in the United States: Results from Big Data Analysis
Krittanawong, Chayakrit , Maitra, Neil Sagar , Wang, Zhen , Jneid, Hani , Sharma, Samin K. , Ang, Song P , Sharma, Samin K , Krittanawong, Chayakrit , Ang, Song Peng , Alam, Mahboob

Mayo Clinic

Journal of Cardiovascular Development and Disease | April 17, 2025 | Volume: 12 | Issue: 4 | Pages: 161

10.3390/jcdd12040161 | PMID 40278220 Get It

Non-ST-segment elevation myocardial infarction (NSTEMI) can be managed by ischemia guide strategies or early invasive strategies. Here, we present the findings of an updated contemporary analysis regarding the use of intracoronary imaging (ICI)-guided PCI versus angiography-guided PCI and in-hospital mortality in patients with NSTEMI in the United States using the NIS database from 2016 to 2021. ICI use increased by nearly threefold between 2016 and 2021, without a significant difference in in-hospital mortality, though interestingly, mortality rates compared with angiography guidance were similar and relatively low. In this study, the use of ICI was associated with lower adjusted odds of in-hospital mortality, cardiogenic shock, and cardiac arrest, but with a longer length of stay and cost of hospitalization.
Improved Survival and Prognostication in Melanoma Patients With Brain Metastases: An Update of the Melanoma Graded Prognostic Assessment
Sperduto, Paul W. , Marqueen, Kathryn E. , Chang, Enoch , Li, Jing , Davies, Michael A. , Ebner, Daniel K. , Breen, William G. , Lamba, Nayan , Shih, Helen A. , Mahal, Amandeep R. , Rahman, Rifaquat , Ankrah, Nii, Boggs, Drexell H. , Lewis, Calvin D. , Hyer, Daniel E. , Johri, Fasila, Soliman, Hany , Masucci, Laura , Chiang, Veronica , Phuong, Christina , Dajani, Salah, Wan, Zihan, Niedzwiecki, Donna , Vaios, Eugene , Pasetsky, Jared , Wang, Tony J. C. , Kotecha, Rupesh , Ross, Richard B. , Rusthoven, Chad G. , Nakano, Toshimichi , Tawbi, Hussein A. , Edwards, Donna , Kim, Michelle M , Rahman, Rifaquat , Boggs, Drexell H , Lewis, Calvin , Hyer, Daniel , Buatti, John M , Aneja, Sanjay , Chiang, Veronica , Phuong, Christina , Braunstein, Steve , Sachdev, Sean , Niedzwiecki, Donna , Vaios, Eugene , Pasetsky, Jared , Wang, Tony J C , Kotecha, Rupesh , Ross, Richard B , Rusthoven, Chad G , Nakano, Toshimichi , Tawbi, Hussein , Mehta, Minesh P , Shih, Helen A , Kim, Michelle M. , Buatti, John M. , Roberge, David , Braunstein, Steve , Kirkpatrick, John P. , Kütük, Tuğçe , Mehta, Minesh P.

Mayo Clinic

Journal of Clinical Oncology | April 17, 2025 | Volume: 43 | Issue: 16 | Pages: 1910-1919

10.1200/jco-24-01351 | PMID 40245362 Get It

PURPOSE Survival for patients with melanoma has recently improved. The propensity of melanoma to metastasize to the brain remains a common and serious feature of this disease. The purposes of this study were to evaluate prognostic factors for patients with newly diagnosed melanoma brain metastases (MBMs) in a large cohort treated with modern multimodal therapies, compare those results with those in prior eras, and update the Melanoma Graded Prognostic Assessment (GPA). METHODS Univariable and multivariable (MVA) analyses of prognostic factors and treatments associated with survival were performed on 1,796 patients with newly diagnosed MBM treated between January 01, 2015, and December 31, 2021, using a multi-institutional retrospective database. Multiple imputation was used to address missingness of potential predictors. Significant variables in combined MVA were used to update the Melanoma GPA. Comparisons were made with legacy cohorts. RESULTS Median survivals for cohorts A (1985-2007, n = 481), B (2006-2015, n = 823), and C (2015-2021, n = 1,796) were 6.7, 9.8, and 16.6 months and median follow-up times were 40.1, 43.6, and 48.8 months, respectively. In combined MVA, significant prognostic factors for survival were higher Karnofsky Performance Status, fewer MBMs, absence of extracranial metastases, lower serum lactate dehydrogenase, and no immunotherapy before MBM. These factors were incorporated into the updated Melanoma GPA. The combined median and 3-year survivals for patients with GPA 0-1, 1.5-2, and 2.5-4.0 were 5.4, 13.2, and 43.2 months and 12.4%, 28.8%, and 51.6%, respectively. CONCLUSION Prognostic factors have changed and survival has improved for patients with MBM but varies widely by GPA. The updated Melanoma GPA calculator (BrainMetGPA), available free online, can be used to estimate survival, individualize treatment, stratify clinical trials, guide surveillance, and augment clinical trial eligibility. Multidisciplinary treatment is essential. Trials are needed to elucidate the optimal sequencing of various therapeutic modalities.
Peripheral Iridectomy With Goniosynechialysis and Goniotomy vs Trabeculectomy for Advanced PACG
Lin, Fengbin , Lv, Aiguo , Li, Fei , Song, Yunhe , Xie, Lin , Zhu, Xiaomin , Tang, Li , Zhang, Yao , Huang, Jingjing , Zhang, Hengli , Lŭ, Lan, Xiao, Meichun , Xu, Jiangang , Nie, Xin , Liao, Mengfei, Yang, Yangfan , Gao, Kai , Yuan, Huiping , Song, Wulian , Zuo, Chengguo , Yan, Xiaowei , Wang, Zhenyu , Jin, Ling, Peng, Yu-Ying, Chen, Weirong , Park, Ki Ho , Aung, Tin, Lam, Dennis S.C. , Fan, Sujie , group, TVG study, Liu, Yuhong , Li, Xiaoyan , Zhou, Fengqi, Tang, Guangxian , Nie, Xin, Zuo, Chengguo , Lu, Ping, Zhou, Fengqi , Barton, Keith , Aung, Tin , Weinreb, Robert Neal , Wang, Ning Li , Tham, Clement Chee Yung , Zhang, Xiulan , group, TVG study, Liang, Xiaohong

Mayo Clinic

JAMA Ophthalmology | April 17, 2025 | Volume: 143 | Issue: 6 | Pages: 472

10.1001/jamaophthalmol.2025.0757 | PMID 40244579 Get It

ImportanceThe safety and effectiveness of combining surgical peripheral iridectomy (SPI) with goniosynechialysis (GSL) and goniotomy (GT) vs trabeculectomy for intraocular pressure (IOP) reduction remains unknown.ObjectiveTo investigate the safety and effectiveness at 1 year of SPI + GSL + GT vs trabeculectomy in advanced primary angle-closure glaucoma (PACG) without cataract.Design, Setting, and ParticipantsThis noninferiority randomized clinical trial was conducted at 8 tertiary eye centers in China. A total of 88 Chinese patients (88 eyes) with advanced PACG without cataract were enrolled from January 2022 to July 2023. Data were analyzed from August 2024 to September 2024. Mean (SD) patient age was 60.3 (7.3) years, and 52 patients (59.1%) were female. Forty-three patients were randomized to SPI + GSL + GT and 45 were randomized to trabeculectomy; 86 patients (97.7%) completed the 12-month follow-up.InterventionsParticipants were randomized 1:1 to receive SPI + GSL + GT or trabeculectomy.Main Outcomes and MeasuresThe primary outcome was IOP at 12 months (noninferior margin: 4 mm Hg). Secondary outcomes included surgical success (IOP: 5-18 mm Hg, ≥20% reduction from baseline, with or without antiglaucoma medications); postoperative complications and interventions, including bleb massage, suture lysis, or releasable sutures; and number of antiglaucomatous medications prescribed.ResultsAt 12 months, the SPI + GSL + GT group had a mean (SD) IOP of 15.6 (4.0) mm Hg vs 14.9 (4.2) mm Hg in the trabeculectomy group (difference, 0.5 mm Hg; 95% CI, −1.2 to 2.2; P = .55), which was within the 4-mm Hg noninferiority margin. Qualified success rates were 38 of 43 participants (88.4%) for SPI + GSL + GT and 42 of 45 participants (93.3%) for trabeculectomy (difference, −5.0%; 95% CI, −19.6% to 8.5%; P = .48). However, complete success rates were lower in the SPI + GSL + GT group (26 participants [60.5%]) vs the trabeculectomy group (37 participants [82.2%]; difference, −21.8%; 95% CI, −40.2% to −2.4%; P = .03). Postoperative complications were present for 8 participants (18.6%) in the SPI + GSL + GT group vs 9 participants (20.0%) in the trabeculectomy group (difference, −1.4%; 95% CI, −17.9% to 15.1%; P = .71). Postoperative interventions were lower in the SPI + GSL + GT group (3 participants [7.0%] vs 25 participants [55.6%]; difference, 48.6%; 95% CI, 32.2%-65.0%; P < .001). Median (IQR) numbers of medications used decreased from 2 (0-3) to 0 (0-1) in the SPI + GSL + GT group and from 2 (2-3) to 0 (0-0) in the trabeculectomy group (difference, −0.81; 95% CI, −1.36 to −0.26; P = .004).Conclusions and RelevanceIn this randomized clinical trial among patients with advanced PACG without cataract, SPI + GSL + GT demonstrated noninferiority (4-mm Hg margin) to trabeculectomy for IOP at 12 months, with fewer interventions (including bleb massage, suture lysis, or releasable sutures) but no difference in postoperative medication use. This suggests SPI + GSL + GT as a potential alternative to trabeculectomy for similar cases, pending validation in larger sample sizes with smaller noninferiority margins.Trial RegistrationClinicalTrials.gov Identifier: NCT05163951
Two-Year Outcomes of Phacogoniotomy vs Phacotrabeculectomy for Advanced Primary Angle-Closure Glaucoma With Cataract
Song, Yunhe , Fan, Sujie , Tang, Li , Lin, Fengbin , Li, Fei , Lv, Aiguo , Li, Xiaoyan , Wen, Tingli, Lŭ, Lan, Xiao, Meichun , Xie, Lin , Zhang, Hengli , Yan, Xiaowei , Yuan, Huiping , Song, Wulian , Yang, Yangfan , Xu, Jiangang , Wang, Zhenyu , Jin, Ling, Zhou, Minwen , Zhao, Xiaohuan , Chen, Weirong , Park, Ki-Ho, Aung, Tin, Lam, Dennis S.C. , Group, PVP Study, Liu, Yuhong , Fang, Zige , Zhou, Fengqi, Wen, Tingli, Zhu, Xiaomin , Tang, Guangxian , Zhou, Fengqi , Wang, Zhenyu, Liang, Xiaohong , Park, Ki-Ho, Barton, Keith , Aung, Tin , Tham, Clement Chee Yung , Weinreb, Robert Neal , Wang, Ning Li , Zhang, Xiulan , Group, PVP Study

Mayo Clinic

JAMA Ophthalmology | April 17, 2025 | Volume: 143 | Issue: 6 | Pages: 462

10.1001/jamaophthalmol.2025.0685 | PMID 40244620 Get It

ImportanceIntraocular pressure (IOP) reduction with phacogoniotomy (phacoemulsification plus goniosynechialysis plus goniotomy) was not less than that of phacotrabeculectomy for advanced primary angle-closure glaucoma (PACG) with cataract at 1-year follow-up, but longer-term outcomes are needed.ObjectiveTo investigate if phacogoniotomy is noninferior to phacotrabeculectomy for advanced PACG with cataract at 2 years.Design, Setting, and ParticipantsThis multicenter, noninferiority, randomized clinical trial took place in 7 ophthalmology centers in China. The trial started May 31, 2021, and 2-year follow-up ended May 31, 2024. Included in this analysis were patients with advanced PACG and cataract. Study data were analyzed from September 2024 to January 2025.InterventionsRandom assignment (1:1) to phacogoniotomy or phacotrabeculectomy.Main Outcomes and MeasuresThe primary outcome measure was reduction in IOP from baseline to the 2-year visit with a noninferiority margin of 4 mm Hg.ResultsA total of 124 participants (124 eyes) were randomized (mean [SD] age, 66.4 [8.6] years; 67 female [54.0%]), 65 (52.4%) to the phacogoniotomy group and 59 (47.6%) to the phacotrabeculectomy group. A total of 59 patients (90.7%) in the phacogoniotomy group and 52 patients (88.1%) in the phacotrabeculectomy group completed 2-year visits. All participants were Chinese. Mean (SD) IOP reduction was −25.6 (10.2) mm Hg and −24.7 (9.4) mm Hg in the phacogoniotomy and phacotrabeculectomy groups, respectively, and the upper boundary of the CI for difference in change between groups was lower than the 4-mm Hg noninferiority margin (mean difference, −0.5 mm Hg; 97.5% CI, −1.7 mm Hg to 0.8 mm Hg; P = .42). The mean difference for complete success for phacogoniotomy vs phacotrabeculectomy was −6.7% (95% CI, −21.4% to 8.8%; P = .47) and for qualified success was 1.4% (95% CI, −11.0% to 14.3%, P = .30). Median (IQR) number of antiglaucomatous medication was 0 (0) vs 0 (0; Hodges-Lehmann estimate of location shift, 0; 95% CI, 0; P =.12) with phacogoniotomy vs phacotrabeculectomy, respectively (mean difference, 0.13; 95% CI, −0.36 to 0.63; P = .60).Conclusions and RelevanceMean IOP reduction with phacogoniotomy was noninferior to phacotrabeculectomy for advanced PACG and cataract at 2-year follow-up with no differences detected in complete or qualified success or mean number of antiglaucomatous medications. These findings support phacogoniotomy as an alternative to phacotrabeculectomy for patients with advanced PACG and cataract.Trial registrationClinicalTrials.gov Identifier: NCT04878458
Outcomes of Percutaneous Coronary Intervention in Nonagenarians in the United States
Mahmood, Arhum, Qadeer, Yusuf Kamran , Wang, Zhen , Jneid, Hani, Sharma, Samin, Krittanawong, Chayakrit , Ang, Song P , Ang, Song Peng , Alam, Mahboob , Jneid, Hani , Sharma, Samin

Mayo Clinic

Catheterization and Cardiovascular Interventions | April 17, 2025 | Volume: 106 | Issue: 1 | Pages: 250-257

10.1002/ccd.31538 | PMID 40241276 Get It

ABSTRACT Background Although Percutaneous Coronary Intervention (PCI) is the cornerstone treatment acute myocardial infarction (AMI), its use in the elderly, specifically nonagenarians patients, is not well studied. This study sought to compare the outcomes and complications of nonagenarian patients who experienced AMI between those who underwent PCI and those who underwent medical treatment only. Methods We evaluated 301,440 nonagenarian (ages 90−99) patients who presented to the hospital with AMI who were listed in the National Inpatient Sample from 2016 to 2021. AMI was defined according to the ICD‐10 Diagnostic Codes. Multivariable logistic regression analysis was used to examine the association of PCI with primary outcomes of mortality and secondary outcomes. The temporal trend of both the incidence of PCI in nonagenarian patients as well as the mortality rate between 2016 and 2021 were expressed as percentages over time. Results Of the total ( n = 301,440) nonagenarian patients with AMI, 33,035 patients underwent PCI while 268,406 did not undergo PCI and rather, just utilized optimized medical therapy (OMT). Of these, 3290 (9.96%) died in the PCI group, and 43580 (16.24%) died in the OMT group. All of the secondary outcomes were significantly different between the PCI and OMT groups. Comparing the two groups, the PCI group was associated with decreased mortality (OR 0.63 [95% CI, 0.58−0.69]; p < 0.001), acute heart failure (OR 0.88 [95% CI, 0.82−0.95] p < 0.001), and AKI (OR 0.75 [95% CI, 0.70−0.79]; p < 0.001), and increased cardiogenic shock (OR 3.06 [95% CI, 2.77−3.38]. The temporal of PCI in nonagenarian patients showed an increase in frequency from about 8.3 in 2016 to about 13.7% in 2021. Furthermore, comparing the mortality between the PCI and OMT groups showed a significant difference with a decreased mortality in the PCI group. Conclusions Nonagenarian patients experiencing AMI who underwent PCI is associated with a significant mortality decrease compared to those who underwent OMT only. The PCI group was also associated with a significant decrease in multiple secondary complications including acute heart failure, AKI, acute stroke, and an increase in cardiogenic shock. Temporally, we have seen an increase in PCI being used in nonagenarian patients over the interval.
Enhancing Lupus Anticoagulant Testing with Machine Learning: Deep Neural Networks Match Expert Performance without the Need for Feature Selection
Leger, Rachel, Chen, Dong , Seheult, Jansen N. , Chen, Dong , Seheult, Jansen, Seheult, Jansen N , Wang, Jeffrey Y.

Mayo Clinic

The Journal of Applied Laboratory Medicine | April 16, 2025 | Volume: 10 | Issue: 4 | Pages: 911-923

10.1093/jalm/jfaf039 | PMID 40238795 Get It

Abstract Background The lupus anticoagulant (LAC) is an important laboratory criterion in the diagnosis of antiphospholipid antibody syndrome. LAC testing at the Mayo Clinic Special Coagulation Laboratory includes up to 13 tests, which are interpreted by trained physicians to identify the presence of LAC and rule out anticoagulant interferences. This feasibility study explored the use of two deep neural network (DNN) architectures for multilabel classification of LAC profiles as a first step toward automating interpretation. Methods Seven thousand two hundred and two retrospective cases were randomly split (64:16:20) for training, validation, and test, respectively. LAC positivity by dilute Russell's viper venom time (LAC-DRVVT) and activated partial thromboplastin time (LAC-APTT) and the presence of warfarin (WAR) and heparin (HEP) were adjudicated by one expert. DNN architectures included: single-output DNNs using domain-knowledge input feature selection and a single-column multioutput DNN using all 13 inputs. Results Domain-knowledge-naïve multioutput DNN achieved similar or improved performance for all 4 label prediction tasks compared with domain knowledge optimized DNNs: F1 scores of 0.977 vs 0.968 for LAC-DRVVT, 0.954 vs 0.945 for LAC-APTT, 0.961 vs 0.957 for HEP, and 0.995 vs 0.977 for WAR, respectively. Conclusions The comparable performance of the 4 domain knowledge optimized DNNs and the multioutput DNN using all 13 input features suggests that the DNN may learn feature importance or mapping to a task without explicit input selection. Given its relative simplicity and versatility, the multioutput DNN is the preferred choice for implementation in a clinical laboratory to standardize LAC diagnosis.