scholarly journals Prognostic Model Development with Missing Labels

2019 ◽  
Vol 61 (3) ◽  
pp. 327-343
Author(s):  
Patrick Zschech ◽  
Kai Heinrich ◽  
Raphael Bink ◽  
Janis S. Neufeld
Author(s):  
Mihaela van der Schaar ◽  
Harry Hemingway

Machine learning offers an alternative to the methods for prognosis research in large and complex datasets and for delivering dynamic models of prognosis. Machine learning foregrounds the capacity to learn from large and complex data about the pathways, predictors, and trajectories of health outcomes in individuals. This reflects wider societal drives for data-driven modelling embedded and automated within powerful computers to analyse large amounts of data. Machine learning derives algorithms that can learn from data and can allow the data full freedom, for example, to follow a pragmatic approach in developing a prognostic model. Rather than choosing factors for model development in advance, machine learning allows the data to reveal which features are important for which predictions. This chapter introduces key machine learning concepts relevant to each of the four prognosis research types, explains where it may enhance prognosis research, and highlights challenges.


Author(s):  
Michael T. Koopmans ◽  
Irem Y. Tumer

Uncertainty assessment and management is becoming an increasingly essential aspect of good prognostic design for engineering complex systems. Uncertainty surrounding diagnostics, loads, and fault progression models is very real and propagating this uncertainty from component-level health estimates to the system-level remains difficult at best. In this work, a test stand is used to conduct real-time failure experiments aboard various aircraft platforms to collect failure response data, expanding the actuator knowledge base that forms the foundation of component health estimations. The research takes a step towards standardizing a test stand design to produce comparable and scalable failure data sets, fostering uncertainty reduction within the electromechanical actuator prognostic model. This paper specifically presents a method to optimize the actuator coupling for a commercially available actuator where a model was built to minimize the coupling deflection and estimate the coupling life. Using this model, researchers can rapidly develop their own electromechanical actuator test stands.


Author(s):  
Mohammad H. Jamal ◽  
Suhail A. Doi ◽  
Sarah AlYouha ◽  
Sulaiman Almazeedi ◽  
Mohannad Al-Haddad ◽  
...  

AbstractBackgroundCOVID19 is worldwide pandemic that is mild in the majority of patients but can result in a pneumonia like illness with progression to acute respiratory distress syndrome and death. Predicting the disease severity at time of diagnosis can be helpful in prioritizing hospital admission and resources.MethodsWe prospectively recruited 1096 consecutive patients with COVID19 from the Jaber Hospital, a COVID19 facility in Kuwait, between 24 February and 20 April 2020. The primary endpoint of interest was disease severity defined algorithmically. Predefined risk variables were collected at the time of PCR based diagnosis of the infection. Prognostic model development used 5-fold cross-validated regularized logit regression. The cohort was divided into a training and validation cohort and all model development proceeded on the training cohort.ResultsThere were 643 patients with clinical course data of whom 94 developed severe COVID19. In the final model, age, CRP, procalcitonin, lymphocyte and monocyte percentages and serum albumin were independent predictors of a more severe illness course. The final prognostic model demonstrated good discrimination, calibration and internal validity.ConclusionWe developed and validated a simple score calculated at time of diagnosis that can predict patients with severe COVID19 disease.


Blood ◽  
2011 ◽  
Vol 118 (21) ◽  
pp. 499-499 ◽  
Author(s):  
Theresa Hahn ◽  
Philip L. McCarthy ◽  
Jeanette Carreras ◽  
Mei-Jie Zhang ◽  
Hillard M. Lazarus ◽  
...  

Abstract Abstract 499 AHCT is standard therapy for relapsed or refractory HL. Published prognostic models for HL patients based on factors measured at the time of AHCT have been limited by small sample sizes. HL prognostic models based on information from diagnosis may be difficult to use for AHCT outcomes since diagnostic information is often not available to the tertiary transplant center or the tests were not uniformly performed by multiple referring physicians. Our goal was to develop a new prognostic model for PFS post-AHCT based on factors available at time of AHCT. We analyzed a cohort of 728 relapsed or refractory HL patients receiving an AHCT between 1996–2007, reported to the CIBMTR by 162 centers, who had complete data for all significant factors previously reported in prognostic models. Patient characteristics at diagnosis: 40% male, 52% stage III-IV, 57% B symptoms, 34% extranodal disease. Patient characteristics at AHCT: median (range) age 33 (7–74) years; 74% KPS≥90 pre-AHCT; 40% had ≥3 prior chemotherapy regimens; 36% chemo-sensitive relapse 27% CR2, 19% PR1, 12% chemo-resistant relapse, 6% primary refractory/resistant; median (range) time from diagnosis to AHCT 22 (3–368) months. Histologic types were: 74% nodular sclerosis, 14% mixed cellularity, 7% lymphocyte rich, 1% lymphocyte depleted, 4% other/unknown. High dose therapy regimens were primarily BEAM (71%) or CBV (13%). For the entire cohort, 3-year estimates of PFS and OS were 60% and 73%, respectively. Multivariate models for treatment failure (1-PFS) were built using a forward step-wise procedure with p<0.05 to enter the model. The following variables were considered: number of prior chemotherapy regimens; KPS; histology; B symptoms at diagnosis; disease status at AHCT; chemo-sensitivity at AHCT; serum LDH at AHCT; extranodal involvement any time prior to AHCT; size of largest mass prior to AHCT; time from diagnosis to AHCT. A random subset of patients was used for model development (n=337) and the model was validated in the remaining cases (n= 391). The final model is shown in the TableRisk FactorRR (95% CI)PScore# of prior chemotherapy regimens: (3,4,5) vs (0,1,2)1.80 (1.31–2.47)0.00032Extranodal involvement any time prior to AHCT: Yes vs No1.77 (1.24–2.53)0.00182KPS prior to AHCT: 0–80% vs 90–100%1.47 (1.04–2.07)0.02751HL chemo-sensitivity at AHCT: Resistant vs Sensitive1.45 (1.01–2.07)0.04401 Patients were assigned a risk group based on the prognostic score: High risk, (score = 4, 5, or 6); Intermediate risk, (score = 1, 2, or 3); and Low risk, (score = 0). Figure 1 shows the PFS curves for the model development, model verification and combined groups, respectively. This CIBMTR Prognostic Model identifies patients at low, intermediate and high risk for treatment failure (progression or death). These risk groups discriminate patients with good post-AHCT outcomes and those who may benefit from other therapies, such as allogeneic HCT. Prospective evaluation of different treatment strategies based on this prognostic model are needed on a national or international level. Disclosures: Hahn: Novartis: stock. Montoto:Genentech: Research Funding; Roche: Honoraria.


2020 ◽  
Author(s):  
Mike Domenik Rinderknecht ◽  
Yannick Klopfenstein

As the COVID-19 pandemic is challenging healthcare systems worldwide, early identification of patients with a high risk of complication is crucial. We present a prognostic model predicting critical state within 28 days following COVID-19 diagnosis trained on data from US electronic health records (IBM Explorys), including demographics, comorbidities, symptoms, insurance types, and hospitalization. Out of 15816 COVID-19 patients, 2054 went into critical state or deceased. Random, stratified train-test splits were repeated 100 times and lead to a ROC AUC of 0.872 [0.868, 0.877] and a precision-recall AUC of 0.500 [0.488, 0.509] (median and interquartile range). The model was well-calibrated, showing minor tendency to over-forecast probabilities above 0.5. The interpretability analysis confirmed evidence on major risk factors (e.g., older age, higher BMI, male gender, diabetes, and cardiovascular disease) in an efficient way compared to clinical studies, demonstrating the model validity. Such personalized predictions could enable fine-graded risk stratification for optimized care management.


2021 ◽  
Vol 2021 ◽  
pp. 1-14
Author(s):  
Xisong Liang ◽  
Zeyu Wang ◽  
Ziyu Dai ◽  
Hao Zhang ◽  
Quan Cheng ◽  
...  

Malignant neoplasms are characterized by poor therapeutic efficacy, high recurrence rate, and extensive metastasis, leading to short survival. Previous methods for grouping prognostic risks are based on anatomic, clinical, and pathological features that exhibit lower distinguishing capability compared with genetic signatures. The update of sequencing techniques and machine learning promotes the genetic panels-based prognostic model development, especially the RNA-panel models. Gliomas harbor the most malignant features and the poorest survival among all tumors. Currently, numerous glioma prognostic models have been reported. We systematically reviewed all 138 machine-learning-based genetic models and proposed novel criteria in assessing their quality. Besides, the biological and clinical significance of some highly overlapped glioma markers in these models were discussed. This study screened out markers with strong prognostic potential and 27 models presenting high quality. Conclusively, we comprehensively reviewed 138 prognostic models combined with glioma genetic panels and presented novel criteria for the development and assessment of clinically important prognostic models. This will guide the genetic models in cancers from laboratory-based research studies to clinical applications and improve glioma patient prognostic management.


BMJ Open ◽  
2019 ◽  
Vol 9 (7) ◽  
pp. e029813 ◽  
Author(s):  
David J Keene ◽  
Karan Vadher ◽  
Keith Willett ◽  
Dipesh Mistry ◽  
Matthew L Costa ◽  
...  

ObjectiveTo predict functional outcomes 6 months after ankle fracture in people aged ≥60 years using post-treatment and 6-week follow-up data to inform anticipated recovery, and identify people who may benefit from additional monitoring or rehabilitation.DesignPrognostic model development and internal validation.Setting24 National Health Service hospitals, UK.MethodsParticipants were the Ankle Injury Management clinical trial cohort (n=618) (ISRCTN04180738), aged 60–96 years, 459/618 (74%) female, treated surgically or conservatively for unstable ankle fracture. Predictors were injury and sociodemographic variables collected at baseline (acute hospital setting) and 6-week follow-up (clinic). Outcome measures were 6-month postinjury (primary) self-reported ankle function, using the Olerud and Molander Ankle Score (OMAS), and (secondary) Timed Up and Go (TUG) test by blinded assessor. Missing data were managed with single imputation. Multivariable linear regression models were built to predict OMAS or TUG, using baseline variables or baseline and 6-week follow-up variables. Models were internally validated using bootstrapping.ResultsThe OMAS baseline data model included: alcohol per week (units), postinjury EQ-5D-3L visual analogue scale (VAS), sex, preinjury walking distance and walking aid use, smoking status and perceived health status. The baseline/6-week data model included the same baseline variables, minus EQ-5D-3L VAS, plus five 6-week predictors: radiological malalignment, injured ankle dorsiflexion and plantarflexion range of motion, and 6-week OMAS and EQ-5D-3L. The models explained approximately 23% and 26% of the outcome variation, respectively. Similar baseline and baseline/6 week data models to predict TUG explained around 30% and 32% of the outcome variation, respectively.ConclusionsPredictive accuracy of the prognostic models using commonly recorded clinical data to predict self-reported or objectively measured ankle function was relatively low and therefore unlikely to be beneficial for clinical practice and counselling of patients. Other potential predictors (eg, psychological factors such as catastrophising and fear avoidance) should be investigated to improve predictive accuracy.Trial registration numberISRCTN04180738; Post-results.


Author(s):  
Arta Hoesseini ◽  
Nikki van Leeuwen ◽  
Aniel Sewnaik ◽  
Ewout W. Steyerberg ◽  
Robert Jan Baatenburg de Jong ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document