Non-Contact Molecular Chemical Imaging Assessment of Tissue Congestion: A Validation Study Compared to MRI
Saltman, A. E.; Ghosh, R. N.; Eckman, P. M.; Chappuis, M.; Collins, J. D.; Morely, P.; Mazis, H.; Schweitzer, R.; Enriquez-Sarano, M.; Tazik, S. K.; Gomer, H.; Nelson, M.; Treado, P. J.
Show abstract
BackgroundHeart failure (HF) is a healthcare problem burdening patients, care teams, and health care systems. A relentless, downward spiral of decompensation, hospital admission, and functional deterioration is challenging to break, partly because of the lack of an objective, inexpensive tool to aid the clinicians physical examination. We studied performance of a shortwave infrared (SWIR) molecular chemical imaging (MCI) tool to measure relative tissue congestion (TC) in HF patients shins, compared to MRI Dixon sequence TC measurements as ground truth. MethodsForty-seven (47) subjects underwent paired SWIR MCI and MRI measurements of their lower extremities. Thirty-six (36) subjects were hospitalized with decompensated HF while 11 healthy outpatients served as controls. A partial least squares (PLS) regression model was trained to ingest the SWIR MCI spectra and produce a CardioVerification Index (CVI) that mirrored MRI measurements. ResultsThe SWIR MCI model reflected MRI TC measurements accurately for all subjects with a 0.743 linear correlation coefficient. Surprisingly, the MRI results identified a significant fraction of non-HF subjects with elevated TC levels, so a sub-analysis was performed on "Low TC" subjects with low and undetectable levels of extremity edema. This subpopulations linear correlation between MRI and CVI was 0.674. A logistic classifier model differentiated HF from non-HF subjects; the area under the receiver operating characteristic curves was 0.906 for the entire subject group and 0.821 for the Low TC subpopulation. ConclusionSWIR MCI techniques can be used in a TC-measuring tool that replicates MRI TC measurements. Clinicians could use such tools to help guide therapy for HF patients during in-hospital management of acute HF decompensation and for outpatient monitoring. SWIR MCIs ability to identify elevated TC in otherwise normal subjects may also meet the unmet need for a novel, widely applicable HF screening tool.
Matching journals
The top 8 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Metabolomic profile for understanding heart failure classifications 95%
- Left ventricular mass and global wall thickness – prognostic utility and characterization of left ventricular hypertrophy 94%
- Improved evaluation of left ventricular hypertrophy using the spatial QRS-T angle by electrocardiography 94%
Similar papers in this journal
- The HeartMagic prospective observational study protocol - characterizing subtypes of heart failure with preserved ejection fraction 95%
- Non-Invasive Scale Measurement of Cardiac Output Compared with the Gold-Standard Direct Fick Method: A Feasibility Study 95%
- An International Longitudinal Natural History Study of Danon Disease Patients: Unique Cardiac Trajectories Identified Based on Sex and Heart Failure Outcomes 94%
Similar papers in this journal
- Hemodynamic profiles by non-invasive monitoring of cardiac index and vascular tone in acute heart failure patients in the emergency department: external validation and clinical outcomes 93%
- Metabolic Modulation Predicts Heart Failure Tests Performance 93%
- Predictors and outcomes of Cardiac Dyssynchrony among patients with heart failure attending Benjamin Mkapa Hospital in Dodoma, central Tanzania: A protocol of prospective-longitudinal study 93%
Similar papers in this journal
- Prognostic value of compact myocardial thinning in patients with left ventricular non-compaction 94%
- Investigating Electrocardiographic Abnormalities in Patients with Coronary Microvascular Dysfunction 93%
- A Multicenter Evaluation of the Impact of Procedural and Pharmacological Interventions on Deep Learning-based Electrocardiographic Markers of Hypertrophic Cardiomyopathy 93%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.