0
selected
-
1.
A fusion decision system to identify and grade malnutrition in cancer patients: Machine learning reveals feasible workflow from representative real-world data.
Yin, L, Song, C, Cui, J, Lin, X, Li, N, Fan, Y, Zhang, L, Liu, J, Chong, F, Wang, C, et al
Clinical nutrition (Edinburgh, Scotland). 2021;(8):4958-4970
Abstract
BACKGROUND AND AIMS Most nutritional assessment tools are based on pre-defined questionnaires or consensus guidelines. However, it has been postulated that population data can be used directly to develop a solution for assessing malnutrition. This study established a machine learning (ML)-based, individualized decision system to identify and grade malnutrition using large-scale data from cancer patients. METHODS This was an observational, nationwide, multicenter cohort study that included 14134 cancer patients from five institutions in four different geographic regions of China. Multi-stage K-means clustering was performed to isolate and grade malnutrition based on 17 core nutritional features. The effectiveness of the identified clusters for reflecting clinical characteristics, nutritional status and patient outcomes was comprehensively evaluated. The study population was randomly split for model derivation and validation. Multiple ML algorithms were developed, validated and compared to screen for optimal models to implement the cluster prediction. RESULTS A well-nourished cluster (n = 8193, 58.0%) and a malnourished cluster with three phenotype-specific severity levels (mild = 2195, 15.5%; moderate = 2491, 17.6%; severe = 1255, 8.9%) were identified. The clusters showed moderate agreement with the Patient-Generated Subjective Global Assessment and the Global Leadership Initiative on Malnutrition criteria. The severity of malnutrition was negatively associated with the nutritional status, physical status, quality of life, and short-term outcomes, and was monotonically correlated with reduced overall survival. A multinomial logistic regression was found to be the optimal ML algorithm, and models built based on this algorithm showed almost perfect performance to predict the clusters in the validation data. CONCLUSIONS This study developed a fusion decision system that can be used to facilitate the identification and severity grading of malnutrition in patients with cancer. Moreover, the study workflow is flexible, and might provide a generalizable solution for the artificial intelligence-based assessment of malnutrition in a wider variety of scenarios.
-
2.
Impact of a Computerized Antithrombotic Risk Assessment Tool on the Prescription of Thromboprophylaxis in Atrial Fibrillation: Hospital Setting.
Pandya, E, Masood, N, Wang, Y, Krass, I, Bajorek, B
Clinical and applied thrombosis/hemostasis : official journal of the International Academy of Clinical and Applied Thrombosis/Hemostasis. 2018;(1):85-92
Abstract
The computerized antithrombotic risk assessment tool (CARAT) is an online decision-support algorithm that facilitates a systematic review of a patient's stroke risk, bleeding risk, and pertinent medication safety considerations, to generate an individualized treatment recommendation. The CARAT was prospectively applied across 2 hospitals in the greater Sydney area. Its impact on antithrombotics utilization for thromboprophylaxis in patients with nonvalvular atrial fibrillation was evaluated. Factors influencing prescribers' treatment selection were identified. The CARAT recommended a change in baseline therapy for 51.8% of patients. Among anticoagulant-eligible patients (ie, where the risk of stroke outweighed the risk of bleeding) using "nil therapy" or antiplatelet therapy at baseline, the CARAT recommended an upgrade to warfarin in 60 (30.8%) patients. For those in whom the bleeding risk outweighed the stroke risk, the CARAT recommended a downgrade from warfarin to safer alternatives (eg, aspirin) in 37 (19%) patients. Among the "most eligible" (ie, high stroke risk, low bleeding risk, no contraindications; n = 75), the CARAT recommended warfarin for all cases. Discharge therapy observed a marginal increase in anticoagulation prescription in eligible patients (n = 116; 57.8% vs 64.7%, P = .35) compared to baseline. Predictors of warfarin use (vs antiplatelets) included congestive cardiac failure, diabetes mellitus, and polypharmacy. The CARAT was able to optimize the selection of therapy, increasing anticoagulant use among eligible patients. With the increasing complexity of decision-making, such tools may be useful adjuncts in therapy selection in atrial fibrillation. Future studies should explore the utility of such tools in selecting therapies from within an expanded treatment armamentarium comprising the non-vitamin K antagonist oral anticoagulants.
-
3.
Quantitative contrast-enhanced harmonic EUS in differential diagnosis of focal pancreatic masses (with videos).
Săftoiu, A, Vilmann, P, Dietrich, CF, Iglesias-Garcia, J, Hocke, M, Seicean, A, Ignee, A, Hassan, H, Streba, CT, Ioncică, AM, et al
Gastrointestinal endoscopy. 2015;(1):59-69
Abstract
BACKGROUND The role of EUS with contrast agents can be expanded through the use of time-intensity curve (TIC) analysis and computer-aided interpretation. OBJECTIVE To validate the use of parameters derived from TIC analysis in an artificial neural network (ANN) classification model designed to diagnose pancreatic carcinoma (PC) and chronic pancreatitis (CP). SETTING Prospective, multicenter, observational trial-endoscopy units from Romania, Denmark, Germany, and Spain. PATIENTS A total of 167 consecutive patients with PC or CP. INTERVENTIONS Contrast-enhanced harmonic EUS (CEH-EUS) and EUS-guided FNA (EUS-FNA), TIC analysis, and ANN processing. MAIN OUTCOME MEASUREMENTS Sensitivity, specificity, positive and negative predictive values (PPV, NPV) for EUS-FNA, CEH-EUS, and the ANN. RESULTS After excluding all of the recordings that did not meet the technical and procedural criteria, 112 cases of PC and 55 cases of CP were included. EUS-FNA was performed in 129 patients, and the diagnosis was confirmed by surgery (n = 15) or follow-up (n = 23) in the remaining cases. Its sensitivity and specificity were 84.82% and 100%, respectively, whereas the PPV and NPV were 100% and 76.63%, respectively. The sensitivity of real-time quantitative assessment of CEH-EUS was 87.5%, specificity 92.72%, PPV 96.07%, and NPV 78.46%. Peak enhancement, wash-in area under the curve, wash-in rate, and the wash-in perfusion index were significantly different between the groups. No significant differences were found between rise time, mean transit time, and time to peak. For the ANN, sensitivity was 94.64%, specificity 94.44%, PPV 97.24%, and NPV 89.47%. LIMITATIONS Only PC and CP lesions were included. CONCLUSION Parameters obtained through TIC analysis can differentiate between PC and CP cases and can be used in an automated computer-aided diagnostic system with good diagnostic results. ( CLINICAL TRIAL REGISTRATION NUMBER NCT01315548.).
-
4.
Pacemaker-based analysis of atrioventricular conduction and atrial tachyarrhythmias in patients with primary sinus node dysfunction.
Stockburger, M, Trautmann, F, Nitardy, A, Just-Teetzmann, M, Schade, S, Celebi, O, Krebs, A, Dietz, R
Pacing and clinical electrophysiology : PACE. 2009;(5):604-13
Abstract
BACKGROUND Most patients with symptomatic sinus node disease (SND) receive DDDR pacemakers (PM) in order to cover SND and atrioventricular (AV) block from the outset. But the concern about adverse effects of right ventricular pacing (RVP) is increasing. So far, data on the incidence of AV block in SND are based on clinical events. The study undertakes to assess and appraise AV block and atrial tachyarrhythmias (AT) from memory and electrograms of a dual-chamber PM set to an AAIR-DDDR switch mode (AAISafeR). METHODS A dual-chamber PM incorporating the AAISafeR mode was implanted in 58 patients (70 +/- 10 years, 28 males) with SND, but without AV block >I. AV block and AT episodes were retrieved from the PM memory and validated from electrograms. AV block episodes were classified potentially relevant while comprising AV block III or AV block I/II during exercise. RESULTS The patients experienced a median of 90 (interquartile range 7-1,084) commutations. Possibly relevant AV block occurred in 32 patients (55%). Validation revealed high-quality PM-based categorization. The RVP prevalence was 0% (0-16%). The median AT prevalence was 0.03 (0-26) min/day. RVP was the only multivariate predictor of AT (P = 0.001). CONCLUSIONS Potentially relevant AV block occurs frequently in patients with SND. Nonetheless, the RVP prevalence is kept low through the AAISafeR mode. The protection of SND patients with demand-actuated ventricular pacing appears reasonable. The AT prevalence is low in SND patients treated by the AAISafeR mode. Even low RVP proportions appear to favor AT. Prospective evaluation is needed.
-
5.
In vivo 31P MR spectral patterns and reproducibility in cancer patients studied in a multi-institutional trial.
Arias-Mendoza, F, Payne, GS, Zakian, KL, Schwarz, AJ, Stubbs, M, Stoyanova, R, Ballon, D, Howe, FA, Koutcher, JA, Leach, MO, et al
NMR in biomedicine. 2006;(4):504-12
Abstract
The standardization and reproducibility of techniques required to acquire anatomically localized 31P MR spectra non-invasively while studying tumors in cancer patients in a multi-institutional group at 1.5 T are reported. This initial group of patients was studied from 1995 to 2000 to test the feasibility of acquiring in vivo localized 31P MRS in clinical MR spectrometers. The cancers tested were non-Hodgkin's lymphomas, sarcomas of soft tissue and bone, breast carcinomas and head and neck carcinomas. The best accrual and spectral quality were achieved with the non-Hodgkin's lymphomas. The initial analysis of the spectral values of the sum of phosphoethanolamine plus phosphocholine normalized by the content of nucleotide triphosphates in a homogeneous sample of 32 NHL patients studied by in vivo (31)P MRS showed good reproducibility among different institutions. No statistical differences were found between the institution with the largest number of cases accrued and the rest of the multi-institutional NHL data (2.28 +/- 0.64, mean +/- standard error; n = 17, vs 2.08 +/- 0.14, n = 15). The preliminary data reported demonstrate that the institutions involved in this trial are obtaining reproducible 31P MR spectroscopic data non-invasively from human tumors. This is a fundamental prerequisite for the international cooperative group to be able to demonstrate the clinical value of the normalized determination of phosphoethanolamine plus phosphocholine by 31P MRS as predictor for treatment response in cancer patients.